» MAKE: NOISE — Discuss this article

Showing messages 1 through 221 of 221.

• thanksgiving power up
  You must be logged in to reply.
  
  
  I powered up the amp for the first time today. Sounds awesome. Thanks Ross
  

  Posted by simcain on November 24, 2011 at 16:13:48 Pacific Time

• thanksgiving power up
  You must be logged in to reply.
  
  
  You're welcome, and thank you for letting us know you were successful. I hope you continue to like the sound. I listen to one of the prototypes for several hours a day and haven't gotten tired of it.

  Posted by BAUHAUSLER on November 24, 2011 at 17:00:54 Pacific Time

• battery power
  You must be logged in to reply.
  
  
  Almost done building my amp per instructions, and as i ordered enough to build 2 more, was thinking about a jobsite radio. Using 18 volt cordless drill motor battery's. Would that work?

  Posted by simcain on November 06, 2011 at 14:13:54 Pacific Time

• battery power
  You must be logged in to reply.
  
  
  That's a great idea. The amp should draw very little current under most circumstances so the batts should last a long time. +/- 18VDC is a good operating voltage for this circuit. You would need 2 18V batteries to make up the bipolar supply.
  You can delete the rectifier diodes and the 2 big filter capacitors on the circuit board. Keep the ground point 'loop' of wire there for the audio circuit grounds to attach to. You can solder wires to the board at the positive and negative filter capacitor positions and retain all of the board wiring to these points. Your batteries would connect to the ground wire loop (0V), the upper capacitor position (+18V) and the lower capacitor position (-18V). You probably already 'get' this but I'm laying it out for other readers.
  The datasheet for the chip specifies operating voltages as low as +/- 8VDC or as high as +/- 30VDC so there's a wide range of batteries that would work as long as the supply is bipolar.

  Posted by BAUHAUSLER on November 06, 2011 at 14:28:28 Pacific Time

• Almost Finished...
  You must be logged in to reply.
  
  
  Hey. Just about finished the build on my Squelette. Modified the design a little. Just need to add the plexi top and it's done. Fired up the first time like a champ and sounds great. All in all a great first electronics project!
  
  Thanks for the article!
  
  Now on to an LM4780 setup for my outdoor speakers...
  
  http://www.flickr.com/photos/69320502@N03/sets/72157628032354800/

  Posted by miller.desu on November 01, 2011 at 20:21:27 Pacific Time

• Almost Finished...
  You must be logged in to reply.
  
  
  Very nice implementation. I love the casework. Those heat-pipe heat sinks are the bomb. I have a stack of them from Cisco servers waiting for future chip amp projects.
  I'm not sure if you have an insulating tab between your chip and heat sink. You should know that the chip tab is electrically connected to the -ve power rail internally. This means if it's directly connected to the heat sink the heat sink metal body will have -18VDC on it. Potentially this could cause a short if it comes into contact with ground. That could blow the fuse, but 19V isn't going to hurt anyone.
  Again, nice job. I hope you enjoy the sound.

  Posted by BAUHAUSLER on November 02, 2011 at 19:35:10 Pacific Time

• Almost Finished...
  You must be logged in to reply.
  
  
  Thanks! Yeah, no insulators on the chips - figured I didn't need them as it wasn't a grounded case. Building another one for a friend now (auditioned it for him and he loved it) - i'll order another set and get em' on there. Thanks again!

  Posted by miller.desu on November 08, 2011 at 20:51:25 Pacific Time

• Capacitor Choice
  You must be logged in to reply.
  
  
  I've read mixed reviews online about electrolytic vs. film capacitors. You specify a few of each in the project. Are there any downfalls to switching a film cap for an electrolytic as long as the values and voltage ratings are the same?

  Posted by j_bush on September 20, 2011 at 16:58:08 Pacific Time

• Capacitor Choice
  You must be logged in to reply.
  
  
  This circuit uses film capacitors where their clean signal handling is needed and electrolytics where their compact size and low inductance are advantages. There's really nowhere you could replace an electrolytic with a film cap in this without causing more problems than you tried to solve. Recall that compact size is key to this circuit's stability and substituting much larger film caps into the build would defeat that. The chips' critical local power rail bypass caps are a combo of film and electrolytic to utilize the properties of both. As for the 'sound' of the circuit, each chip contains 46 transistors and a number of other components. The IC's circuit topology, which you cannot change, determines the performance to a much greater extent than any external passive component.

  Posted by BAUHAUSLER on September 20, 2011 at 20:03:31 Pacific Time

• Capacitor Choice
  You must be logged in to reply.
  
  
  Thanks! For the bypass caps you mentioned, is it important for both chips to have their own bypass network? Or can I use the bypass network built for chip_1 to supply chip_2 as well?
  
  Thanks again

  Posted by j_bush on September 21, 2011 at 12:56:55 Pacific Time

• Capacitor Choice
  You must be logged in to reply.
  
  
  The purpose of the bypass caps is to ensure the lowest inductance in the DC supply to the chip. This supports high frequency stability. To accomplish this the bypass caps must be physically close to the chip's pins. It would be very difficult to orient the two chips so they are close enough to share bypasses and still far enough apart not to share heat on the heat sink. If you're interested in reading more, the datasheet on the chip has quite a bit of detail on implementation constraints like bypassing. I think I posted a link to the datasheet elsewhere on the page. It's not too technical. Or Google LM1875 datasheet.

  Posted by BAUHAUSLER on September 21, 2011 at 17:31:13 Pacific Time

• Signal Testing
  You must be logged in to reply.
  
  
  This has been a lot of fun so far but I just got stuck on the signal testing step 2d. I don't understand how to connect the signal source to the amp. For the signal source I downloaded and burned a 1000Hz sample tone to a CD but I'm not sure what to do next. Do I just connect RCA cables to the output of the CD player and then alligator clip the center pin part of each RCA cable to I8 and AD8? Do I need to also clip the outer metal ring of each RCA cable to something? I believe that's the ground. Or am I totally off? Any help would be greatly appreciated.
  
  Thanks.

  Posted by jagup on September 03, 2011 at 15:49:02 Pacific Time

• Signal Testing
  You must be logged in to reply.
  
  
  Thanks for the question. You have it exactly right. The signal leads of the source component connect to I8 and AD8. One of the grounds (rings) of the RCA plugs connects to the central ground on the board. The RCA grounds are electrically identical so connecting one is sufficient. I recommended using a portable CD player because they have a variable volume output. If you use a CD player's full level output it might drive the amp too hard because the signal is so high. I'd use a source where you can turn the output signal down initially and raise it as you start the test.

  Posted by BAUHAUSLER on September 03, 2011 at 18:24:06 Pacific Time

• Recommendation
  You must be logged in to reply.
  
  
  Well, I finally got this amp together. Sounds OK! Thanks for the article. Where can I learn more? Are there any books/sources you'd recommend for a noobish hobbyist? I built the circuit but I don't have a solid understanding of how and why all the parts were chosen and placed where they were in the circuit.

  Posted by Mossen on July 18, 2011 at 21:45:47 Pacific Time

• Recommendation
  You must be logged in to reply.
  
  
  Well done! I hope it gives you many hours of music pleasure. A good way to understand this chip is to read its data sheet. If that's too jargon-y and technical you need a website with a basic explanation of solid state amplification. The internals of the chip are quite complex, with each chip containing over 40 transistors and many diodes, so the chip circuit is unlikely to be intuitive. To get the very basics of how solid state amps work I'd recommend studying some of the one-transistor single ended MOSFET amps related to Nelson Pass' Zen amps. They're super basic so it's easy to see how the parts work. There's a whole enthusiast group working on these simple but inefficient circuits. You could Google around for "MOSFET Zen" or start at Rod Elliot's site SOUND.WESTHOST.COM and read the Death of Zen article for starters. I think it's fascinating stuff.

  Posted by BAUHAUSLER on July 19, 2011 at 05:52:42 Pacific Time

• Another question about grounding
  You must be logged in to reply.
  
  
  Is the ground of the power supply connected to the ground of the amplifier circuit?

  Posted by Mossen on June 12, 2011 at 22:49:42 Pacific Time

• Another question about grounding
  You must be logged in to reply.
  
  
  There is a star ground point at the power supply capacitors' zero volts node. It's the loop of bare wire at the back of the circuit board. All audio circuit grounds, the chassis and the power transformer center tap connect separately to this ground point.

  Posted by BAUHAUSLER on June 13, 2011 at 06:21:55 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  I am working on a project based on the Squelette. It is a well-documented project, but I have some questions about the heatsink design, since I will be building an enclosure of my own design and want to make sure it has adequate heat transfer. Can you explain, or provide any resources for how you determined that the aluminum case design for the original Squelette had low enough thermal resistance to carry heat away from the LM1975s?
  
  I have found lots of information about how to do thermal design, and how to compute junction temperatures based on known thermal resistance, but most of it depends on using the rated thermal resistance on the datasheet of a manufactured heatsink, and I can find little about determining how a piece of aluminum of arbitrary size (and shape) will perform under the same conditions.
  
  I would really appreciate any direction you can provide.
  
  Thank you very much,
  Andrew

  Posted by schamp on June 11, 2011 at 11:53:04 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  Good questions. I'm glad you're working out the technical details in advance. Thermal design for audio components is somewhat difficult because all of the documentation tends to deal with steady-state power levels. Audio signals are not steady state, which makes traditional analysis of heat design not very useful. For squelette I used a cut/try approach. I took one circuit board and attached it to a thin piece of alu 6" X 3" and ran it at various power levels. I found it would easily take 5 watts/chip continuous sine - roughly equal to an audio signal at the point where peaks clip. The squelette cabinet has more than 5 times as much area as that test heat sink so I deemed it safe.
  The datasheet for the chip (Google & download) has some guidelines for thermal management. Note that the heat developed in the chip is proportional to the average current output and not the average power output. It will heat up more producing 5W into 4 Ohms (1.12A) than it will delivering 5W into 8 Ohms (0.8A).

  Posted by BAUHAUSLER on June 11, 2011 at 13:28:06 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  As far as estimating the dissipation capacity of sheet aluminum, I would think the surface area is the dominant factor. A heat sink should perform roughly the same as plain aluminum with the same exposed surface, assuming the thickness of the sheet is adequate to allow quick heat conduction to all parts.
  My brother is a materials scientist who develops thermoelectric devices and creates his own heat testing machines. I'll ask him and if I'm badly wrong I'll post a correction here.

  Posted by BAUHAUSLER on June 11, 2011 at 13:33:30 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  Here's what my brother, who designs thermoelectric systems had to say about heat sinks:
  Re. heat sinks: you have to take into account the thermal resistance of the aluminum itself - heat transport isn't perfect, the point where the heat enters is always at a higher temperature than the point where it exits the aluminum. It's analogous to, and governed by the same equation as, electrical resistance causing a voltage drop. The thermal conductivity of aluminum is about 200 watts per (meter * degree C); you use the cross-sectional area and the distance the heat travels to find out how many degrees C the source rises per watt of heat transport. Hence your qualifier "of adequate thickness".
  
  Convection is pretty simple, here's a good site:
  http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html
  This describes the temperature difference between the ambient air and the surface of the aluminum. The coefficient for free (not forced by a fan) convection is 5 to 25 watts per (square meter * degree C). I haven't been able to determine what puts you on the high side or the low side of that range; the humidity, the orientation of the surface with respect to gravity, and other parameters are in it. BTW, if you use a heatsink in free convection, it's really important to orient the fins vertically, not horizontally, so the warm air can rise out of it. Elsewhere, you can find "heat sink equations" to determine size, fin density, etc. for your application.
  
  The above isn't going to completely answer his question. If the need for a heatsink was marginal in the first place, and the component isn't very sensitive (losing performance or lifetime operating at high temperature), then you can go for it. Otherwise, play it safe.

  Posted by BAUHAUSLER on June 11, 2011 at 16:18:57 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  Thanks for the reply! It's been very helpful. I am sorry for the delay in responding. I've been working through a lot of this stuff, and am preparing a more detailed write-up about it, which I'll post when I can wrap it up (and, possibly, test my build).

  Posted by schazamp on June 28, 2011 at 11:50:24 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  So I think I'm getting pretty close here. Here's my reasoning about thermal resistance of the heatsink. For a passive heatsink of given size and material, the heat dissipation will be determined by the worse of the convection heat transfer or the thermal resistance of the material. According to that convection calculator page you posted, using Newton's Law of Cooling, I can determine the temperature rise from ambient for a given heatsink size and power output, and therefore the effective thermal resistance (dT / power). Using this slab thermal resistance calculator ( http://www.novelconceptsinc.com/calculators-slab-thermal-resistance-constriction.cgi ), given a heatsink size and material and package size, I can determine the thermal resistance of a heatsink passing heat to another material (not counting convection or thermal emissivity).
  
  It seems to me, therefore, that no matter how low the thermal resistance of my heatsink material, it will be bounded by the amount of heat that can be dissipated by convection. Conversely, no matter how large my heatsink (and therefore the amount I can dissipate by convection), if the thermal resistance of my heatsink is too bad, then the heatsink will not draw heat away from the IC fast enough to dissipate it.
  
  So, therefore, the thermal resistance of my heatsink to ambient will effectively be the greater of what I can dissipate by convection and the thermal resistance of the heatsink material itself.
  
  Most of the resources I have found (this one was particularly helpful: http://sound.westhost.com/heatsinks.htm ) seem consider emissivity important, but generally dominated by convection, and I have not yet been able to find good information to help determine the effective heat dissipation by thermal radiation. However, that can only serve to lower the total thermal resistance I'm dealing with, so if I can find a design that seems effective in the worst case (lowest convective heat transfer coefficient, no emissivity effect), then any other factors (better heat transfer coefficient, or emissivity) can only serve to increase the headroom I have to work with.
  
  I'm still planning a more detailed write up with the calculations, but I wanted to put this out there first. Does this reasoning seem correct? Or is there something I'm failing to account for?

  Posted by schazamp on June 29, 2011 at 07:21:46 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  I follow you and I think you're on the right track. If you're going to model the system mathematically, don't forget the interface between the chip and the heat sink. Different materials have different thermal resistances. You can mount the chip directly to the sink for best conductivity but then the heat sink is electrically connected to the -ve supply rail and must be carefully insulated electrically from everything else.
  I built a small prototype in a radio shack project box using only the thin metal bottom. That unit would run at 100% output, both channels continuous before the temp rose enough for the chips' auto temp protection to shut them down. For permanent use I installed a tiny 25mm 12V 'muffin' fan to move air through it and the working temp dropped significantly.
  Hit this link and scroll to the bottom for a pic: http://www.boingboing.net/2010/08/02/my-favorite-makers-a.html
  
  Here are some pix of the internals:
  http://www.flickr.com/photos/ross-hershberger/4111722344/in/photostream/

  Posted by BAUHAUSLER on June 29, 2011 at 16:25:08 Pacific Time

• Heatsink Design
  You must be logged in to reply.
  
  
  I am including the thermal resistance of the insulators, which are Wakefield 173-7-220P. The datasheet claims that the TO-3 package ones have an Rth of 0.33 degC / Watt. I figured the TO-220 ones would be something close to that.

  Posted by schazamp on June 30, 2011 at 05:34:20 Pacific Time

• Heat Sink Design
  You must be logged in to reply.
  
  
  Well, I've completed my analysis, and here's the write up:
  http://schazamp.wordpress.com/2011/07/08/schamplifier-heat-sink-design/
  
  I ended up determining the thermal resistance by conduction of the heat sink material itself, and the thermal resistance by convection from the heat sink to ambient.
  
  I'm still looking for feedback about whether this analysis is sound, but I think that at the very least, I'm starting on the right track.
  
  Thank you very much for your support and advice!
  
  Andrew

  Posted by schazamp on July 11, 2011 at 04:37:16 Pacific Time

• Channel connections
  You must be logged in to reply.
  
  
  I'm a little confused by the schematic. The pictures in the article clearly show two 1875's, the schematic has only one. I assumed the schematic was the same for both channels, but how then does the volume pot fit in? Surely there aren't two volume knobs, but I don't understand how to hook it up so that the one pot controls both channels. Also, how does the power supply connect? Does Vcc+ and Vcc- just feed both 1875s?

  Posted by Mossen on May 29, 2011 at 21:28:51 Pacific Time

• Channel connections
  You must be logged in to reply.
  
  
  VCC+ and Vcc- do connect to both channels. the circuit board has one bipolar power supply in teh center and two identical audio channels, on the left and right sides. Each channel has one chip. The schematic has only one channel shown for clarity.
  The volume pot is a stereo part. It has two completely separate circuits inside controlled by one knob/shaft. The stereo pot has 6 terminals. 3 of them are used for one channel and 3 for the other. In this way the volume for both channels is controlled by one knob. Same thing for the input switch.

  Posted by BAUHAUSLER on May 29, 2011 at 21:47:33 Pacific Time

• What is the ground?
  You must be logged in to reply.
  
  
  This is the first circuit I've built using A/C. Reading through the article I noticed everything connects back to the "star" ground. But what does the star ground connect to? The chassis? Do all the components really just get wired together like that? I'm used to building DC circuits where everything eventually goes back to the negative battery terminal so this is a little confusing to me...

  Posted by Mossen on May 29, 2011 at 21:11:45 Pacific Time

• What is the ground?
  You must be logged in to reply.
  
  
  Good questions. It's a bipolar power supply with both negative and positive rails of the same voltage and current capacity. The star ground is the zero in between them. The star ground point does connect to the chassis but this is not required for the circuit to work. The chassis connection ensures that the fuse will blow if there's a power supply short to the chassis.
  The various 'ground' nodes in the audio and power supply circuits are all taken to the star point separately. This is so that their ground currents will not mix in the same wires. If, for instance, the power supply ground current shared a wire with the feedback circuit ground then the capacitor charging current would cause a small signal in the wire tha twould get fed into the feedback circuit, to the input of the chip and would be heard as noise. Think of all wires as having a small amount of inductance and resistance, and a small noise voltage caused by the current running through them. Sometimes those noises can become significant and should be separated. the datasheet for the chip mentions this specifically and calls for separate ground wires.

  Posted by BAUHAUSLER on May 29, 2011 at 21:20:51 Pacific Time

• What is the ground?
  You must be logged in to reply.
  
  
  OK, that makes sense. Thanks!

  Posted by Mossen on May 29, 2011 at 21:47:06 Pacific Time

• Speaker recommendation
  You must be logged in to reply.
  
  
  I came across a pair of KEF Cresta II bookshelf speakers for $100 and these may be the best match I've found for the Squelette. They're bookshelf sized - almost too big for my '50s era desk - but they produce lots of deep bass. Excellent balance and clarity without the edgy shrillness of a lot of modern speakers. And at 92 dbs they're efficient enough to rock a 900 square foot room with 10 watts. They're not common but are worth looking for if you want a taste of real high end sound for a bargain price.

  Posted by BAUHAUSLER on May 29, 2011 at 20:40:23 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  I read the paragraph about adding a headphone jack, but as a total noob, do not follow it. Can you provide more details about what parts are needed and what is connected to what, where? Thank you. Also, here's the link to Parts Express page where the tO200 insulators are found http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=055-400

  Posted by Eulipian on May 27, 2011 at 15:06:45 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Those silicone insulators from Pants Excess will work fine and that's a fair price for them. The headphones are just like speakers. Instead of wiring up the headphone jack with wires directly from the circuit board you connect a 150 Ohm resistor for each channel in series in place of the wire so each signal, left and right, goes through 150 Ohms.
  The metal body of the headphone jack serves as the ground for the left and right headphone channels. The jack ground (body) tab can be wired directly to the central power supply star ground connection at the capacitors on the board. For each channel, you connect the audio out signal through a 150 Ohm resistor to the jack. The signal can come either from the output position for the channel on the circuit board, or if it's more convenient, from the red speaker binding post for that channel. I have the prototype circuit board here on my desk but the letter/number card fell off so I'm not sure of the hole positions for the output. It's where the positive speaker wire connects.

  Posted by BAUHAUSLER on May 29, 2011 at 20:03:38 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  "Pants Excess" Ha! read it four times before I saw it.

  Posted by Eulipian on June 18, 2011 at 14:38:37 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Our little joke. We love Parts Express. They're a great source for audio hobbyist parts, and enthusiastically support us. Decent prices, fast shipping and a discussion forum full of experienced users. They're a great resource.

  Posted by BAUHAUSLER on June 19, 2011 at 07:07:49 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  The headphone jack will have three connections: ground, left signal and right signal. The left signal tab is the one that will contact the 'tip' of the headphone plug. The right signal connects to the metal 'ring' on the plug and the ground connects to the 'body' of the plug.

  Posted by BAUHAUSLER on May 29, 2011 at 20:07:15 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Thanks Ross. Am going to start this in a couple weeks as a Summer project with my son. I think the detail and thought you've put into this project are remarkable. Was happy to see so many Makers, including you, pictured at the Motor City Maker Faire. Hope the weather left you alone yesterday. After we build your Squelete, we're going to move back in time to a tube version. Here's the page for the
  Jonokuchi Thanks again for your kind reply.

  Posted by Eulipian on May 31, 2011 at 01:10:36 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Good luck with your Squelette build. It takes a bit of time and patience, but builders are finding that the sound quality is worth the effort.
  Motor City Maker Faire was a blast. I'm an unabashed fan of the magazine so it thrilled me to see a publication spawning a grassroots movement. And I got Mark Frauenfelder's autograph.
  The tube amp link you posted is to Pete Millett's site. Good choice. Pete's a technology guy who creates effective, rational designs. The Edcor transformers he specifies are the best value I've ever found, and I've tried everything available over the last 20 years. Plan for a long lead time on the transformers. Edcor makes to order and has a large backlog. My last order took over a month.
  Pete Millett and Max Robinson (Tubecad Journal) take an engineering approach to tube amp design. Much of what's on the internet on tube amps, and most of what's on discussion groups, relies on magical parts and subjectivity rather than math, measurement and conventional design practice.

  Posted by BAUHAUSLER on May 31, 2011 at 06:30:05 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Tube amps are my primary area of interest, and I've scratch built a number of them. Here's a link to a type 45 triode amp that I built about 13 years ago that had a layout similar to the Millett 13EM7 amp and a similar power output. It sounded great:http://www.flickr.com/photos/ross-hershberger/2548233850/in/photostream/

  Posted by BAUHAUSLER on May 31, 2011 at 06:42:19 Pacific Time

• [Headphone Jack
  You must be logged in to reply.
  
  
  Beautiful work Ross. I'm working my way toward the other end of the tube realm to dirty guitar amps, specifically the Fender 5F6A. Even as a commoner with no tech skills, I know what you mean about the science vs. magic part threads. I'm a big fan of facts. Here are a couple sites from the dirty side of the rails you might enjoy. is very informative with lots of builders deeply informed and tolerant of noobs like myself and science based. And, Prof. Errede at Champaign-Urbana is a theoretical nuclear physicist working with the Hadron and Fermi teams, who loves to rock old tube amps and has his undergrad students build or redo them according to science. Check out some of his /> for a grin. Mark was certainly enthused about the Detroit Maker Faire and it sounds like he had some time in the Henry Ford, which we've taken our kids to whenever we visit. One of the coolest places on the planet.

  Posted by Eulipian on June 01, 2011 at 00:11:19 Pacific Time

• [Headphone Jack
  You must be logged in to reply.
  
  
  Beautiful work Ross. I'm working my way toward the other end of the tube realm to dirty guitar amps, specifically the Fender 5F6A. Even as a commoner with no tech skills, I know what you mean about the science vs. magic part threads. I'm a big fan of facts. Here are a couple sites from the dirty side of the rails you might enjoy. Doug Hoffman Ampsis very informative with lots of builders deeply informed and tolerant of noobs like myself and science based. And, Prof. Errede at Champaign-Urbana is a theoretical nuclear physicist working with the Hadron and Fermi teams, who loves to rock old tube amps and has his undergrad students build or redo them according to science. Check out some of his student projects for a grin. Mark was certainly enthused about the Detroit Maker Faire and it sounds like he had some time in the Henry Ford, which we've taken our kids to whenever we visit. One of the coolest places on the planet. FIXED HTML TAGS.

  Posted by Eulipian on June 01, 2011 at 00:18:43 Pacific Time

• Headphone Jack
  You must be logged in to reply.
  
  
  Sorry, "tO220" insulators.

  Posted by Eulipian on May 27, 2011 at 15:08:13 Pacific Time

• Power input?
  You must be logged in to reply.
  
  
  I'm a little unclear on the chip schematic. What is the power input on the 3PDT switch? Is it from the transformer - 25.2V AC?

  Posted by Mossen on May 23, 2011 at 22:24:30 Pacific Time

• Power input?
  You must be logged in to reply.
  
  
  Oh, that is the signal input, right? If so, why is a switch needed?

  Posted by Mossen on May 23, 2011 at 23:00:20 Pacific Time

• Power input?
  You must be logged in to reply.
  
  
  No switch is needed on the input. I conceived this as a desktop amp so a computer and a portable player could be hooked up with the input switch to select between them. You can omit the switch. Just wire the signal and ground of the input jacks (3 wires - right signal, left signal and ground both) to the positions on the PCB where the switch is supposed to connect.

  Posted by BAUHAUSLER on May 24, 2011 at 07:19:34 Pacific Time

• improper AC reading across load resistor
  You must be logged in to reply.
  
  
  Hi,
  
  I'm in the testing stages of the circuit and I'm able to get .05V from across the load resistors. However when I wire up a signal source I only get wild AC readings. Its connected to a sansa mp3 player and its possible I don't have it all hooked up correctly. I took the ground off of AD8 & I8 and connected one to left channel and one to right. Then I grounded the signal source on the star ground connector. I wasn't sure if I had to leave AD8/I8 connected to each other but whether I leave them connected or no I still get crazy AC readings. Any thoughts would be appreciated.
  
  ~Mark

  Posted by huitlacoche on May 13, 2011 at 12:51:33 Pacific Time

• improper AC reading across load resistor
  You must be logged in to reply.
  
  
  well something did occur to me. In the directions when he indicates 'signal generator' I guess he means something thats putting off a clean tone and not...some rock mp3s.
  Would that be a reason for the wild AC readings?
  

  Posted by huitlacoche on May 13, 2011 at 13:07:00 Pacific Time

• improper AC reading across load resistor
  You must be logged in to reply.
  
  
  The signal source should be a steady state tone. This can be from a signal generator, but most people don't have one. You can download a sig gen program like sigjenny for the computer and create a MP3 file of a steady state 1000 hz tone to play through the amp. This can be played through the computer or burned to a CDR and played on a CD player (or other device). The output signal should be about 20 times as large as the input signal. This is because the ratio of the two resistors in the feedback network is about 20:1. A music signal will have a varying output level and will not give a coherent AC voltage reading at the output. I have found that a typical portable player headphone output level is about 0.060 volts. If this was used as an input, the output of the amp would be around 1.2 volts. Hope this helps.

  Posted by BAUHAUSLER on May 13, 2011 at 21:02:44 Pacific Time

• Worked Fine; Now Fuse Blowing?
  You must be logged in to reply.
  
  
  I recently had the pleasure of enjoying my LM1875 amp for about 24 hours! It sounded fantastic and all of the measurements looked good.
  
  I left the amp unplugged overnight (I do not have a power on/off switch installed) and didn't move or change anything with the position of the amp, wires, cables, or anything. When I plugged it in in the morning, to my surprise, the fuse blew!
  
  Should the fuses I'm using be slow blow or fast blow? Could the surge when I'm plugging the amp in be causing this?

  Posted by tommy_o on May 12, 2011 at 12:20:50 Pacific Time

• Worked Fine; Now Fuse Blowing?
  You must be logged in to reply.
  
  
  Oddly, the LM1875 chips are no longer insulated from the heatsink. Not sure how it would have worked before, but I'm going to switch these SilPads to mica tabs.

  Posted by tommy_o on May 19, 2011 at 14:37:19 Pacific Time

• Worked Fine; Now Fuse Blowing?
  You must be logged in to reply.
  
  
  Note that some meters will read a short between the chip metal tab and ground. the metal tab is internally connected to the (-) power supply pin. The power supply has several capacitors between it and ground. Meters that read resistance with AC signal may see signal flow through the capacitors as a short, so treat these resistance readings with caution in a complete circuit.

  Posted by BAUHAUSLER on May 29, 2011 at 20:50:26 Pacific Time

• Worked Fine; Now Fuse Blowing?
  You must be logged in to reply.
  
  
  Scratch that -- mica insulators installed and no continuity between IC tabs and heatsink, but when I turn power up it just blows the fuse. Maybe I'll DIY a PCB and redo the whole shibang.

  Posted by tommy_o on May 19, 2011 at 20:40:37 Pacific Time

• Worked Fine; Now Fuse Blowing?
  You must be logged in to reply.
  
  
  Scratch that [again] -- checked the power supply schematic and saw they should be slow blow. Swapped them out and it's running like a dream! I'll have to take some photos to show it off.

  Posted by tommy_o on May 28, 2011 at 22:32:31 Pacific Time

• Higher Voltage
  You must be logged in to reply.
  
  
  Hi, I'm at the first testing point (testing the big capacitors) and instead of 18V I'm getting 19.75V. Will this be an issue? I've followed all the wiring instructions closely, but I jumpered the designated connections with 22 gauge wire because there was no specific size called out. I wonder if this is the reason for the higher voltage? Thanks.

  Posted by jagup on April 25, 2011 at 19:41:52 Pacific Time

• Higher Voltage
  You must be logged in to reply.
  
  
  That's at the high end of the expected voltage but there are reasons for that. The power transformer is a bit of an over-achiever in terms of winding ratio. The voltage will drop down under load due to the resistance of the transformer winding wire. Ideally it would be 12.6 X 1.4 = 17.6VDC all of the time. But because the resistance causes the voltage to 'sag' when lots of current is drawn the designers of the transformer build in a margin of extra voltage so it doesn't drop too low at full output. It's also possible that your line voltage is a bit above 120VAC, which isn't unusual.
  If you're getting a positive and negative voltage between 17 and 20 volts and within about 0.75VDC of each other you're good to go. Transformers like this are typically specced with +/- 15% tolerances on voltage.

  Posted by BAUHAUSLER on April 25, 2011 at 20:07:33 Pacific Time

• Higher Voltage
  You must be logged in to reply.
  
  
  Maybe a silly question... Speaking of wire gauge, is there significance of using a thinner wire for the parts that call for 24 gauge or would it be fine to do everything with 18?

  Posted by soulman2050 on June 24, 2011 at 18:47:27 Pacific Time

• Higher Voltage
  You must be logged in to reply.
  
  
  I specced 24ga because it's adequate to carry the current and is small enough to fit through the holes. Larger gauge (lower number) will work fine but may be more awkward to route due to the size.

  Posted by BAUHAUSLER on June 25, 2011 at 07:34:38 Pacific Time

• T0-220 insulators
  You must be logged in to reply.
  
  
  I searched most of the major electronics sites, and none of them had t0-220 insulators . Where did you get yours.

  Posted by jcoster on April 19, 2011 at 19:16:43 Pacific Time

• T0-220 insulators
  You must be logged in to reply.
  
  
  I found insulators for to-220 devices on Mouser, Digikey and Newark. make sure you're not searching for t0-220 but to-220 or to220. Two letters, three digits.

  Posted by BAUHAUSLER on April 19, 2011 at 21:27:43 Pacific Time

• T0-220 insulators
  You must be logged in to reply.
  
  
  So I was just spelling it wrong. Thanks!

  Posted by jcoster on April 20, 2011 at 04:26:47 Pacific Time

• solder to nothing
  You must be logged in to reply.
  
  
  I am a newbie. in direction 1h, it says to solder h15 to g15. the 1K resistor is in h15 but noting is in G15. What should I solder to?

  Posted by Sickness on March 27, 2011 at 08:46:57 Pacific Time

• solder to nothing
  You must be logged in to reply.
  
  
  G15 should have one of the chip pins in it. Direction 1h indicates to install the chip at this point. See the illustrations for the correct holes to put the chip pins in. The chip pins go into G and E 15 and D, F and H 13.

  Posted by BAUHAUSLER on March 27, 2011 at 09:03:15 Pacific Time

• Upgraded Squelette gainclone
  You must be logged in to reply.
  
  
  I decided to use a hammond enclosure and upgraded the transformer to a torroid, installed headphone jack and an additional input for iphone, and used a nice big ol' billet cnc'd knob. I was so inspired, I went on to build a set of speakers too. Thanks for the inspiration bauhausler. This was my first audio project and the setup sound great.
  
  http://i1112.photobucket.com/albums/k488/monoroto1/Picture006.jpg
  http://i1112.photobucket.com/albums/k488/monoroto1/Picture046.jpg
  http://i1112.photobucket.com/albums/k488/monoroto1/Picture272.jpg
  http://i1112.photobucket.com/albums/k488/monoroto1/Picture272.jpg
  http://i1112.photobucket.com/albums/k488/monoroto1/Picture303.jpg

  Posted by DJStugots on March 10, 2011 at 23:51:59 Pacific Time

• Upgraded Squelette gainclone
  You must be logged in to reply.
  
  
  Could you post the schematics or a diagram and the model of the Harmony box you used? Thanks!

  Posted by Eulipian on May 06, 2011 at 13:34:37 Pacific Time

• Upgraded Squelette gainclone
  You must be logged in to reply.
  
  
  That's a really nice build. I can see you've done this kind of thing before. I hope others considering this project take a look at your pix for inspiration. Your style of cabinet construction is probably a little simpler to execute due to the parts all being in one box with a simple lid. I like the IEC power cord socket as well. Nicely planned and executed.

  Posted by BAUHAUSLER on March 11, 2011 at 07:32:08 Pacific Time

• Finished!
  You must be logged in to reply.
  
  
  Finally got around to finishing up mine! I still need to stain the wood and fill a couple small spots.
  
  http://img.photobucket.com/albums/v117/twmegee/P1000386.jpg
  http://img.photobucket.com/albums/v117/twmegee/P1000385.jpg

  Posted by Chankster on March 06, 2011 at 14:23:36 Pacific Time

• Finished!
  You must be logged in to reply.
  
  
  Nice packaging job. Very compact and efficient. I like the look of the wooden box. Those heat sinks should be more than adequate for this amp's power output.

  Posted by BAUHAUSLER on March 06, 2011 at 16:34:26 Pacific Time

• Breaking the perfboard
  You must be logged in to reply.
  
  
  This sounds like a really easy question, but I can't seem to get the perfboard to break right. Exactly how am I supposed to cut it? I've already wasted ten dollars at radioshack on perfboard after perfboard, and My allowance doesn't come in till next week, so I want to make this one count.

  Posted by jcoster on February 23, 2011 at 18:15:44 Pacific Time

• Breaking the perfboard
  You must be logged in to reply.
  
  
  I do it by scoring the board deeply on both sides with a stiff, sharp blade, then placing it on a table edge and levering a piece off. I'm scoring it along a line of holes, where it's weakest anyway. Some people score both sides deeply and then make the break with a wire cutter, working from the edge. To make straight, deep scores, use a straight edge or metal ruler as a guide and make multiple passes with the blade. You should be able to cut nearly all the way through by this method, making the right break a sure thing. Good luck. After cutting I finish the rough edges with a file or sanding disk.

  Posted by BAUHAUSLER on February 23, 2011 at 21:11:47 Pacific Time

• it works!!!
  You must be logged in to reply.
  
  
  I started a message thread a few weeks ago about my amp having audio issues. Turns out it was a wiring error :(
  What I did wrong was solder one end of the feedback resistors (22k&220k) to the wrong lead on the IC chip. After I fixed that, let me say that I was shocked at how amazing this amp sounds!! I tested with some mp3s off my phone and ran it to some pretty cheap speakers and it still sounds absolutely incredible.
  Thank you so much for this fun project, I can't wait for my next build.
  
  and for those interested, here are a couple of trashy pictures of the amp being used along with an iPad running a guitar amp+pedal simulator - a nice little practice amp for my apartment life guitar needs.
  http://i83.photobucket.com/albums/j298/breakneck_rust/amp_02.jpg
  
  http://i83.photobucket.com/albums/j298/breakneck_rust/amp_01.jpg

  Posted by jerkdrive on February 19, 2011 at 21:31:24 Pacific Time

• it works!!!
  You must be logged in to reply.
  
  
  Sorry if I missed the post about your issues. I intend to follow up on every comment but there have been a lot of them and some get overlooked. I'm glad you figured out the issue.
  As far as the sound of the amp, yeah it's pretty astonishing how clean and musical they are. And when you consider it's a cheap IC with a few supporting parts it sort of makes you recalibrate your ideas about high fidelity sound.
  Thanks for the pics. I really like your cabinet. If you have a source and part number for the aluminum cab that you modded to make your case, please post in a response. It looks like it would be useful for a lot of small scale electronics builds. Good casing is difficult to come by, hence all of the cigar box and Altoids tin builds.

  Posted by BAUHAUSLER on February 20, 2011 at 07:43:24 Pacific Time

• it works!!!
  You must be logged in to reply.
  
  
  Yea, I figured my reply just got lost in the mix.
  As for the enclosure, this is the one I bought:
  
  http://www.frys.com/product/1656457?site=sr:SEARCH:MAIN_RSLT_PG
  
  If you have a Fry's near, you can find the boxes in the components isle near the pc and bread boards. They have a wide variety of shapes and sizes of plastic and aluminum.
  And then, for the plexiglass, I bought that at local hardware store.

  Posted by jerkdrive on February 20, 2011 at 08:39:15 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Could this amplifier be modified for use in a car? All I really want is an auxiliary input, a power switch, and a volume knob to power the two speakers in my car, so something like this would be perfect.
  
  Seems that replacing the power supply circuitry with something to up the car's 12 Vdc to 18 (?) Vdc would do the trick, but I'm not quite sure what that'd entail.
  
  I've barely touched a soldering iron since high school, but if this is doable, it seems like it'd be a good way back in.

  Posted by hWonderfill on February 09, 2011 at 20:53:31 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Or.. YOu could just invest in an inexpensive inverter to put in front of this amp and use this exact amp. Easy Peasy...

  Posted by Sarek98 on June 01, 2011 at 20:30:21 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Here's one chip that will work off of a car power supply. It's similar to the LM1875 but the circuit has to be different in order to work off of a single supply voltage rather than a bipolar supply. If you Google around you can probably find some hobbyist builds using this chip to read as a guide:
  http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00000123.pdf

  Posted by BAUHAUSLER on February 09, 2011 at 21:49:20 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Looks like the rest of the URL got lost. Search the ST Microelectronics site for the TDA2003 datasheet.

  Posted by BAUHAUSLER on February 09, 2011 at 21:50:57 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Great, thanks for the tip. I'll do some poking around. Looks like there's a good deal of info out there.

  Posted by hWonderfill on February 10, 2011 at 16:49:11 Pacific Time

• Car stereo?
  You must be logged in to reply.
  
  
  Good question. the minimum power supply voltage for this chip is 16V, or +/- 8VDC. You have only 13.8V to work with in a car. To run this chip you would have to synthesize a higher voltage from the 13.8V, which would be difficult to do.
  I would recommend instead looking into chips specifically designed for car stereo use. There are a number of them, as most commercial car stereo components use chip amps. National Semiconductor would be a good place to start.

  Posted by BAUHAUSLER on February 09, 2011 at 21:13:40 Pacific Time

• Finished my amp
  You must be logged in to reply.
  
  
  ...and the sound is excellent! Better than I expected. And this was my first electronics project! <iframe title ="Preview" scrolling="no" marginheight="0" marginwidth="0" frameborder="0" style="width:320px;height:228px;padding:0;background-color:#fcfcfc;" src="http://cid-6231a897ea598e87.photos.live.com/embedphoto.aspx/MB^_Squelette/Amp^_01.jpg"></iframe>

  Posted by emm_bee on February 09, 2011 at 18:08:01 Pacific Time

• Finished my amp
  You must be logged in to reply.
  
  
  Another attempt to link:
  
  http://cid-6231a897ea598e87.skydrive.live.com/redir.aspx?page=play&resid=6231A897EA598E87!115&type=5&Bsrc=EMSHOO&Bpub=SN.Notifications
  
  Sorry, I'm not adept at posting the link, but I think if you paste this into your browser it will take you there.
  

  Posted by emm_bee on February 09, 2011 at 18:15:39 Pacific Time

• Finished my amp
  You must be logged in to reply.
  
  
  Well I messed up that link. Will try to figure it out.

  Posted by emm_bee on February 09, 2011 at 18:09:10 Pacific Time

• Finished my amp
  You must be logged in to reply.
  
  
  Let's see if this link works:
  
  http://cid-6231a897ea598e87.photos.live.com/embedphoto.aspx/MB%5E_Squelette/Amp%5E_01.jpg
  
  It looks very nice. I like the wood cabinet. Glad you like the sound. These little amp chips are something special.

  Posted by BAUHAUSLER on February 09, 2011 at 21:07:51 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  I ran into a minor snag. When I ordered from Digikey, I accidentally only ordered one of the two 1 ohm resistors and now I can't find them locally. Has anyone put this together without it? While the NatSemi datasheet lists the Zobel on there, several other DIY sites have done without it. Would this be safe to attempt?

  Posted by tommy_o on February 08, 2011 at 19:20:25 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  You could take 4, 4ohm quarter watts and put them in parallel. :)

  Posted by Sarek98 on June 05, 2011 at 16:20:10 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  The circuit will work without the Zobel. It's not part of the amplification circuits. The resistor is there as a 'seat belt' to stop the chip from amplifying stray high frequency signals and trying your tweeters. The chip itself has a bandwidth far beyond audio, right up into the RF bands. If a radio signal or other HF noise gets picked up it would be amplified and sent to the tweeter. The Zobel shunts out of band HF signals to ground and keeps the chip stable as well so it doesn't oscillate. You can build without it but the results depend on environment and other details of assembly.

  Posted by BAUHAUSLER on February 08, 2011 at 20:11:40 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  Looks like I also forgot mounting hardware so ill be ordering anyway. What type of TO-220 insulator is appropriate? Im having trouble telling the difference in all of these at digikey. Would graphite based with a shoulder washer be all that is needed? Thank you for all this help, it is invaluable.

  Posted by tommy_o on February 09, 2011 at 08:24:19 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  I missed the part about Graphite. I think that's a non-insulating type of thermal mounting. The common types of insulating thermal mountings are mica and silicon impregnated synthetic. If it comes with a shoulder washer you can use that to insulate the tab, and probably a #4 or metric 3mm steel screw and nut instead of the nylon screw. I used a nylon screw because it's one less part and the bigger screws are easier to handle.

  Posted by BAUHAUSLER on February 09, 2011 at 08:59:56 Pacific Time

• Zobel Network Necessary?
  You must be logged in to reply.
  
  
  The type of insulator is not critical, but you have to use a nylon screw so the chip tab doesn't short to the chassis through the screw. I used mica insulators, but you can get Bergquist or similar synthetic ones. Whatever's cheapest. Before screwing the chips down feel the mounting area with a finger tip and take off any burrs in the metal. They can puncture the insulator and cause a short.
  In the past when I've needed a large and odd sized insulator I've made one from sheet mica. Craft stores sell the stuff in bagged assortments for something I can't imagine. It's thick so I took a big chunk and working from the edge I peeled it down to about 100 microns. Maybe less. It was pretty easy. Then I put it on a glass surface and used a straight edge and a razor blade to trim it to size.
  This is a fairly small chip, so if you have larger insulators available you can cut one down. A few years ago I needed 32 TO3 transistor insulators and the best deal was on a box of 1,000. I hack those up for most of my insulator needs now, and so will my descendents for generations to come.

  Posted by BAUHAUSLER on February 09, 2011 at 08:54:15 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  Hey there, almost done with this project but I'm a little confused on the directions for adding a headphone jack. Can anyone give me more of a step by step process for how exactly to connect the 150 ohm resistors and how those connect to the headphone jack? To the author: thanks for this fun build as it has been my first 'big' project. I'll post some pictures of it when it's done.
  Thanks.
  

  Posted by jerkdrive on January 23, 2011 at 11:54:05 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  Thanks for your interest and support. If you're almost done, you've done a lot of work. The headphone jack hooks up like a set of speaker terminals except that the right and left signal leads of the headphone jack each connect to the board's speaker outputs through a 150 Ohm resistor rather than through a wire. You can make this connection at the speaker red terminal nuts inside the cabinet if that's more convenient than soldering the resistor to the board. 1/4 Watt resistors or larger should be fine. There are many types of headphone jacks so you'll have to figure out the terminals. The headphone plug has three connections. The body, or 'shield' acts as a ground return for both the left and right headphone. The 'tip' of the plug is the left signal connection. The 'ring' between the tip and the body is the right signal connection. Look at your plug and jack and see what terminals connect to these plug parts when the plug is inserted. As well as the left and right signal connections you have to connect the headphone jack ground. If you have a metal body jack directly mounted to a metal chassis and the chassis is grounded, the jack is grounded and you don't have to run a ground wire. If you have an insulated jack, a plastic mounting panel or the chassis is not grounded, then use a wire to connect the jack 'body' connection to the central ground point on the circuit board.
  I'm listening to one of the Squelette prototypes and JBL Control 1 Monitor speakers now. I think you'll be really pleased with the sound.

  Posted by BAUHAUSLER on January 23, 2011 at 13:31:28 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  cool! thanks for the help with the headphone set up. I think I got that setup right.
  However, I have run into some problems when I am testing the audio. The volume coming out of the amp is really really low. I have to put my ear to the speaker to hear it. And if I turn up the volume past halfway on either the pot or the audio source, the audio sounds bad (best way to explain - it sounds like when you have too much gain on a guitar amp and the audio quality gets real terrible.)
  also the pot rotates counter-clockwise to increase volume ( what little volume there is) which seems wrong to me. Could I have wired it wrong?
  Also, here's a picture
  http://i83.photobucket.com/albums/j298/breakneck_rust/DSC09488.jpg
  
  not the best fabrication job, hacking up a project box on the kitchen table of my apartment, but i'm proud of it none the less.
  thanks for your help and any suggestions you have to help me track down this problem.

  Posted by jerkdrive on January 25, 2011 at 20:22:38 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  Sorry for the late reply. That's a GREAT looking cabinet. If you have a Flickr account you can add your pix to the Make flickr pool. Did you fabricate all of that or is it a stock cabinet modded?
  Since the volume problem occurs on both channels it's less likely to be an accidental mis-wiring on the PCB.
  The volume problem you're describing sounds like the input jacks might not be grounded. If you used a 2P2T switch instead of the 3P2T switch in the article you have to wire all of the input jack grounds to the common ground on the circuit board. Several people have had this problem.
  If you have your input grounds connected properly then look at the orientation of the wiring on the volume pot.
  One other builder accidentally grounded the signal (red) speaker post to the chassis, shorting out the output. That causes a similar problem. I can't see the inside of the chassis, so check that as well.

  Posted by BAUHAUSLER on January 27, 2011 at 07:29:07 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  Shameless bump

  Posted by jerkdrive on February 11, 2011 at 19:47:04 Pacific Time

• Headphone Jack setup
  You must be logged in to reply.
  
  
  Yea, I bought a project box and cut it all to pieces. I couldn't find any cheap aluminum in small quantities, so I took this route.
  
  I wired up a 3PDT switch and I double checked the wiring - I looks like everything is running to the star ground properly. I did rewire the volume pot so that it works clockwise - however that did not resolve the audio issue.
  I don't see any problems with the connections to my speaker terminals either.
  One thing I did notice, is the On/Off switch that is connected to the transformer has a light in it. when I power on the amp the light in the switch comes on for an instant - then turns off. However the LED on the PCB stays lit normal - Is this a concern? On the topic of the power switch - in the magazine you never mention connecting a ground from the On/Off switch. I have one on this switch so I wired it to the star ground. Also i have the power cords live side wired to the switch tab marked Power, the cord neutral wired to a transformer lead and then the switch tab marked ACC wired to the other transformer lead. is that correct?
  image below:
  http://i83.photobucket.com/albums/j298/breakneck_rust/jpg-2.jpg
  
  Other than that, I don't know where or how to check for other errors.
  Any advice would be great, as I don't even know what I'm doing :P
  thanks
  

  Posted by jerkdrive on February 02, 2011 at 22:20:12 Pacific Time

• Saw
  You must be logged in to reply.
  
  
  What kind of saw did you use to cut the aluminium

  Posted by jcoster on January 23, 2011 at 04:53:03 Pacific Time

• Saw
  You must be logged in to reply.
  
  
  I cut the aluminum on a band saw with an all purpose blade. Other options: A hack saw or a Skil reciprocating saw. In the last two cases I would clamp a guide of some kind in place to keep the cut straight. The material is only 1/16 thick and cuts easily.

  Posted by BAUHAUSLER on January 23, 2011 at 07:25:47 Pacific Time

• Input jack
  You must be logged in to reply.
  
  
  I was reading ahead and I looked at step 4b, and I don't quite understand. Are the input jacks the 1/8" audio jacks or the RCA jacks? Thanks

  Posted by jcoster on January 17, 2011 at 17:20:24 Pacific Time

• Input jack
  You must be logged in to reply.
  
  
  Both the 1/8" headphone type jack and the RCA jacks are inputs. The double pole switch selects between them as sources.

  Posted by BAUHAUSLER on January 17, 2011 at 18:09:47 Pacific Time

• Capacitor issue
  You must be logged in to reply.
  
  
  When I reached step 1g. I powered up the board and the first cap read 15.5 volts, but the other one popped open, started hissing and let out a really strange liquid. I removed it from the board. Did I do something wrong in wiring, or was the cap deformed?
  And will anything happen if I replace the 1Ω resistor with a 10Ω resistor?

  Posted by jcoster on January 17, 2011 at 06:45:13 Pacific Time

• Capacitor issue
  You must be logged in to reply.
  
  
  Capacitors will overheat and pop when they're wired in backward. Note the orientation of the caps. One has its negative lead to the ground in the center and one has its positive lead there. The power supply is bipolar, with both negative and positive voltages. If yu wired the negative leads of both power supply caps to ground then the capacitor for the negative supply would be reverse biased and would fail. Replace the cap, noting correct orientation and test again. It's unlikely that anything else was damaged.
  The 1 Ohm resistor is there to load down the output at very high frequencies and prevent oscillation. 1 Ohm is about the right value but you could go as high as 2.5 Ohms if that's what you have on hand. 10 Ohms won't pull enough current at high frequencies to provide any damping. This part is not essential to functioning but it does make the circuit stable under unexpected operation conditions.

  Posted by BAUHAUSLER on January 17, 2011 at 08:00:24 Pacific Time

• Capacitor issue
  You must be logged in to reply.
  
  
  Well, I did use axial lead caps, so I most likely did something wrong there, but I wired the other 4700µF cap in the same way, but it didn't pop.

  Posted by jcoster on January 17, 2011 at 12:51:45 Pacific Time

• Capacitor issue
  You must be logged in to reply.
  
  
  You were right, I powered up the circuit again and the other cap popped. I thought that the arrow that points to the other lead would be the positive lead.

  Posted by jcoster on January 17, 2011 at 17:05:32 Pacific Time

• Capacitor issue
  You must be logged in to reply.
  
  
  Step 1D refers to electrolytic caps having the (-) terminal labeled. By convention, polarized components have the negative or cathode identified.
  
  In the pictures on pages 74, 76 you can see the markings on all of the capacitors.

  Posted by BAUHAUSLER on January 17, 2011 at 18:13:39 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  Squelette readers:
  I've been kicking around a bunch of ideas for new audio electronics projects. One that interests me is a tube preamp with bass and treble controls. It would be relatively simple to build and use, and could add functionality to a chip amp. Any interest in building this? Because of the parts count and the hassle of working with sockets I think a custom PC board would be needed to ease the build process. One drawback of the Squelette build is the difficulty of hand-wiring the blank PC board and I'd like to avoid that for the next project. I'm thinking cost would be in the neighborhood of $100. Any interest? I'm going to the TED Detroit conference today and hope to bend Dale Dougherty's ear about this.

  Posted by BAUHAUSLER on January 09, 2011 at 10:02:43 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  This project is interesting to me for several reasons. Tubes look cool in a retro way. They're simple to implement, pretty rugged and cheap. I like tone controls because I listen to a variety of music with widely differing production values. A little bass and treble tweaking normalizes the presentation for my room. Currently I do this with a fancy digital mastering processor but that's vast overkill when a simple resistor and capacitor network will accomplish the same thing. I like things to be as simple as possible to get the job done.

  Posted by BAUHAUSLER on January 11, 2011 at 07:43:38 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  I know this is late, but I would be all over that. I am currently doing a build on both this project and a tube style guitar amp from the AX84 project website. A tube pre-amp would be cool. AND the pre-amp tubes could be set up to run on voltages that aren't "to" scary.

  Posted by Sarek98 on June 01, 2011 at 20:44:18 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  An easy and cheap way to get high voltage B+ plate supply for a preamp is to do the classic 'backward' transformer trick. Use an ordinary 6.3V or 12.6V filament transformer for the heaters. Chose one with extra current capacity. Connect the secondary to the secondary of another filament transformer (6.3V winding to second transformer's 6.3V winding). The second transformer's 'primary' now has a handy 120VAC signal that can be rectified or voltage doubled for up to 300VDC at low current to run the plates of the tubes. It works, and filament transformers are cheap.

  Posted by BAUHAUSLER on June 02, 2011 at 19:38:05 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  I'm way too tied up with a new job to be working on a tube preamp. I did discover an excellent site that addresses tube line amps and tone controls in an easy to understand and practical way. This is Max Robinson at his best. You have to wade a few pages in but he analyzes practical examples of all of the classic tube tone circuits and describes their design and tradeoffs. It's outstanding retro engineering and I highly recommend it:
  http://www.angelfire.com/electronic/funwithtubes/Amp-Tone.html

  Posted by BAUHAUSLER on June 02, 2011 at 19:33:21 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  Do it with input controls and volume knob to make a full-featured preamp and it would certainly be a hit. In fact, pitch the idea of a semi-regular DIY audio column, including all types of economical hi-fi components from speakers, turntables, USB DACs, and CD transports. The Cmoy headphone amp has been wildly popular online; why wouldn't a few similarly easy to follow articles be successful?

  Posted by tommy_o on January 11, 2011 at 08:31:57 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  BTW, anyone with an interest in tube audio and the willingness to come at it with a little algebra should start here: http://www.vias.org/crowhurstba/index.html. This is a superb basic text on audio from 1959 by the great Norman Crowhurst. It is clear, comprehensive and easy to follow. It also covers loudspeakers and other audio components, but the sections on tube circuits are particularly useful.

  Posted by BAUHAUSLER on January 11, 2011 at 09:03:14 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  Minimum specs would include 2 inputs, volume control, bass and treble. That takes 3 dual triode tubes to accomplish. Add 3 more tubes and you can have a phono section to plug a turntable into. This wheel was invented before most of us were born so there are great circuits available that can easily be adapted to modern materials and parts.
  I like the idea of an audio corner at Make because it's my primary interest but audio electronics is covered pretty heavily already by hobby discussion sites.

  Posted by BAUHAUSLER on January 11, 2011 at 08:43:10 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  I'd certainly have a look. Sounds cool!

  Posted by aviatorbja on January 11, 2011 at 03:42:49 Pacific Time

• Tube preamp with tone controls
  You must be logged in to reply.
  
  
  I would be quite interested, although the tone controls do little for me. Point-to-point wiring a tube preamp or buffer would be quite fun indeed.

  Posted by tommy_o on January 10, 2011 at 18:24:33 Pacific Time

• C107 wrong size?
  You must be logged in to reply.
  
  
  I built myself one of these and am extremely happy. I used the PDF link above to get the schematic.
  
  I noticed that the schematic is identical to the LM1875 datasheet reference except for one thing: You have C107 (the feedback highpass) as .22uF while the datasheet has it as 22uf. When I finished building mine, it had no bass response until I replaced C107 with a larger cap.
  
  Also, I used a much larger +/- 18v transformer. This change (and upgraded caps, of course) increases power to 25W while staying well within the LM1875 specs. A large heatsink is needed with this, though.

  Posted by Kevmatic on January 08, 2011 at 18:28:52 Pacific Time

• C107 wrong size?
  You must be logged in to reply.
  
  
  Thanks for pointing out the error in the labeling of the feedback cap in the online schematic. The diagram has been corrected.

  Posted by BAUHAUSLER on January 10, 2011 at 17:36:39 Pacific Time

• C107 wrong size?
  You must be logged in to reply.
  
  
  Thanks for pointing out the feedback capacitor value error in the schematic. In the magazine the parts list, construction text and photos all have a 22uf cap but there must have been a mistake in redrawing my schematic.
  The chip can be used for higher power by increasing the power supply voltage and the heat sink efficiency. Note that this only works into higher impedances. The small die and case tab limit the maximum heat dissipation and therefore the current throughput. Many speakers have impedance minima of only 50% of their rated impedance. A 4 Ohm speaker therefore could present a 2 Ohm load at some frequencies and run into the chip's 4A current limit before voltage clipping occurs. I recommend that anyone contemplating alternative applications look over the chip datasheet and make sure that their intended use stays within the chip's specs:
  https://docs.google.com/viewer?url=http://www.national.com/ds/LM/LM1875.pdf
  For higher powers up to 50W the 1875's bigger brothers the LM3875 and LM3836 are ideal, and are just slightly more difficult to work with. I've used them all and they're great performers.
  That said, by all means dream up new uses for this great little circuit. It's an outstanding performer and is cheap and easy to experiment with.

  Posted by BAUHAUSLER on January 09, 2011 at 09:56:29 Pacific Time

• C107 wrong size?
  You must be logged in to reply.
  
  
  Above I meant the larger, higher power chips are LM3875 and LN3876. The LM3886 is similar.

  Posted by BAUHAUSLER on January 09, 2011 at 10:12:18 Pacific Time

• Amp Complete
  You must be logged in to reply.
  
  
  Finally finished the amp. Instructions were easy to follow. All parts purchased from Mouser except the transformer which was from Radio Shack. I had some binding posts laying around and some RCA jacks. I built my chassis from brass sheet and some cherry for the end pieces. The metalworking is poor. But hey, I'm a woodworker! Pictures of the amp can be seen here:
  http://brianandonian.com/LM1875Chipamp.aspx
  
  The amp sounds clean. No humm, noise, buzz, etc. The headphone jack is a must - this sounds really good!
  
  Ross, Thanks for the plans. This was a fun build!

  Posted by aviatorbja on January 08, 2011 at 14:04:58 Pacific Time

• Amp Complete
  You must be logged in to reply.
  
  
  Thanks for the pictures to drool over. It really looks outstanding and I love the wood/brass combo.
  As for the sound, don't be shy about throwing it at some high end speakers. More than one audiophile has been amazed by what these little chips can do.

  Posted by BAUHAUSLER on January 08, 2011 at 14:21:17 Pacific Time

• Amp Complete
  You must be logged in to reply.
  
  
  I'm using some homebuilt speakers with Viva drivers. They are very good, but lack low end. I have a pair of 1970's JBL Horizon's. I'll try those next.

  Posted by aviatorbja on January 09, 2011 at 09:14:42 Pacific Time

• Amp Complete
  You must be logged in to reply.
  
  
  If they're the JBL Horizon L-166 here:http://www.lansingheritage.org/html/jbl/specs/home-speakers/1976-l66.htm
  They're very good speakers ans should sound great with a quality amp. Those have the excellent 122A or 129H woofers and carefully designed midrange and tweeter drivers.

  Posted by BAUHAUSLER on January 09, 2011 at 09:37:53 Pacific Time

• Sparks
  You must be logged in to reply.
  
  
  So I have built the project up to step 1g. When I connect the leads and plug in the transformer, to test the power supply, nothing happens. No pops or sparks. But when I try to test the caps at P10 to P13 and P10 to P3 the leads I try to test spark. I am getting 0V on the meter. I checked the wiring and then rechecked it. It looks right. The caps are in the correct way; as are the diodes. I am not sure what I am doing wrong. Any suggests? Please and thank you.

  Posted by Hawkeye13 on January 07, 2011 at 15:30:56 Pacific Time

• Sparks
  You must be logged in to reply.
  
  
  I forgot to mention that along with sparks. The fuse blows. I tried giving it power without testing the leads. No sparks and the fuse is okay. when I do test it the fuse blows and there are sparks at the contact points of the volt meter.

  Posted by Hawkeye13 on January 07, 2011 at 15:36:24 Pacific Time

• Sparks
  You must be logged in to reply.
  
  
  It sounds like your meter is shorting it out and that should not happen. A voltmeter is very nearly an open circuit and draws a minuscule current. An ammeter is a short circuit with almost no resistance. If you have your meter set at amps or milliamps, or have the leads in the wrong holes in the meter then the meter will short out the power supply and give the effects that you describe.

  Posted by BAUHAUSLER on January 07, 2011 at 19:34:48 Pacific Time

• Capacitors, and General Questions
  You must be logged in to reply.
  
  
  I am trying to buy all the pieces for this very cool project, but I am rather new at electronics so most of this is over my head. I have a few questions:
  
  -what kind of film capacitors should I buy? I found some on Jameco.com which are Mylar 50v .1uf and Mylar 100v .22uf. Is 50v and 100v too much? Does it make a difference? Is Mylar the right kind of film cap?
  
  -Is it okay if the film and the electrolytic capacitors have a different voltage rating? I plan to use 50v and 35v.
  
  -For the insulated chip mounts, what is the point of them? Are we trying to keep the metal of the TO-220 chip from contacting the case? Or is it to transfer the heat more easily to the case?
  
  -What kind of heat should I expect this unit to generate? I am planning on using a plastic project box to house it until I can build a case.
  
  Thanks in advance!
  

  Posted by Hawkeye13 on January 02, 2011 at 21:59:06 Pacific Time

• Capacitors, and General Questions
  You must be logged in to reply.
  
  
  Good questions. All capacitors must be 25V rating or higher. The highest voltage in the amp is 17V. More than 25V rating does not accomplish anything but does not hurt. Very high rated caps like 630V would likely be larger and might not physically fit. Mylar film caps are fine, and that's what I used.
  The chip insulators are very important. The chip tab cannot be in direct contact with the heat sink. The heat sink is at zero volts (ground) and the chip tab has -17V on it. If the tab is connected to the heat sink electrically it will short out and blow the fuse. That's why the insulators and nylon screws for mounting the chips. The heat is very low. I built one prototype in a radio shack plastic project box with the metal bottom used as the heat sink. It plays for days and barely gets warm. I did provide it with air holes, though.
  Scroll to the bottom here for a pic of that:
  http://boingboing.net/2010/08/02/my-favorite-makers-a.html

  Posted by BAUHAUSLER on January 02, 2011 at 22:12:57 Pacific Time

• Capacitors, and General Questions
  You must be logged in to reply.
  
  
  Cool thanks. That helps a lot!! So I almost have everything I need except for the resistors. I am having problems finding a 1 watt, 1 ohm metal resistor. I checked radio shack as well as jameco.com and neither had it. Suggestions anyone? Also I was wondering about running resistors in parallel. I have never seen it done before and I don't understand it. I can't find a 20k but I have plenty of 22k and 220k resistors lying around. Am I running the them in line? Like the negative side of one to the positive side of the other? Or is it simply connecting one side of each to the same location and the other side of each together to a different location? I hope that made sense. Thanks again!

  Posted by Hawkeye13 on January 05, 2011 at 00:09:10 Pacific Time

• Capacitors, and General Questions
  You must be logged in to reply.
  
  
  Digikey part number RSF100JB-1R0. Digikey also (for US customers) will do economical USPS envelope shipping (for those last minute caps or resistors that were missed in the first parts order!).

  Posted by tommy_o on January 06, 2011 at 07:38:44 Pacific Time

• Capacitors, and General Questions
  You must be logged in to reply.
  
  
  Paralleling 22K and 220K resistors results in the 20K specified. paralleling 'piggybacks' the two resistors so the current divides to flow through both, resulting in a lower effective resistance than a single resistor. The alternative, series connection, attaches the two resistors in-line so there is only one current path, through both sequentially. That would result in a 242K resistance in this case.
  The 1 Ohm can be any type except wire wound. You can parallel two 2.2 Ohm or three 3.3 Ohm 1/2 watt resistors if that's easier to do. I routinely 'fake up' odd resistance and capacitance values from what I have on hand, and that's legitimate.

  Posted by BAUHAUSLER on January 05, 2011 at 06:55:46 Pacific Time

• Power Supply for LM3875?
  You must be logged in to reply.
  
  
  Would it be possible to use this power supply and transformer for a pair of mono blocks using LM3875 (as in, 1 transformer and power supply circuit per LM3875 circuit)? The afford-ability of this transformer and power supply setup are wonderful!

  Posted by tommy_o on January 02, 2011 at 17:49:10 Pacific Time

• Power Supply for LM3875?
  You must be logged in to reply.
  
  
  I've done that and it works fine. With 3875 you get a little more power into lower impedances because the transformer isn't running as close to its current limit.

  Posted by BAUHAUSLER on January 02, 2011 at 22:04:41 Pacific Time

• Power Supply for LM3875?
  You must be logged in to reply.
  
  
  I meant you get a little more power driving only one channel with one power supply because the current draw and resistive voltage drop are lower.
  This transformer has been showing up on a certain auction site lately. I bought 2 boxes of them for general use.

  Posted by BAUHAUSLER on January 02, 2011 at 22:23:31 Pacific Time

• Component Sourcing
  You must be logged in to reply.
  
  
  I was compling a parts list for most of the parts on digikey. This included the parts that did not have a specific part number associated with them (resistors & capacitors). As I was looking for parts some of them advertised they were for "general purpose" or "audio" should i steer towards any specific type? And is there any specs besides the obvious resistance and capacitance values that I should look for in order to make the best possible amp?

  Posted by minimaker on January 01, 2011 at 08:06:58 Pacific Time

• Component Sourcing
  You must be logged in to reply.
  
  
  Some audio folks prefer the Panasonic FC capacitors that Digikey carries.

  Posted by tommy_o on January 01, 2011 at 17:50:41 Pacific Time

• Component Sourcing
  You must be logged in to reply.
  
  
  Good questions. "Audio" is the default classification for standard grade parts because they do not have special characteristics for high power or frequency. The only significant characteristics are the voltage rating of the capacitors and that the Zobel resistor on the output has to be 1W as specified. The circuit is insensitive to the type or construction of the parts. Some notes on this page give recommended lead spacing dimensions for the best fit of the capacitors on the board, although other parts can just have the leads bent to fit with no penalty to function.
  I used 1/8W metal film resistors for most because they're small. The electrolytic caps can be ordinary types. For the signal capacitors and power supply bypasses I used metalized Mylar parts because they're smallest. I think that's what's specced on the parts list.
  Some people have asked about making an 'upgraded' version with audiophile parts. To my knowledge, changing part types has far less audible effect on this circuit than it does in, say a single ended tube amp. If you put in garden variety inexpensive parts you're not missing anything.

  Posted by BAUHAUSLER on January 01, 2011 at 08:23:10 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  I'm working on a parts list to order from Digikey for this project, but unfortunately there is no Radioshack near me. Is there a comparable transformer from Digikey that someone could direct me towards? Thanks for the help.

  Posted by tommy_o on December 31, 2010 at 14:01:50 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  I checked DigiKey and they list a Hammond transformer that meets the electrical spec. But it's 3 times as costly as the Radio Shack one and non-stock. Radio Shack will ship the transformer directly to you from a website order: http://www.radioshack.com/product/index.jsp?productId=2102703
  
  If you're using the aluminum angle chassis metalwork specced in the article the Radio Shack transformer will fit properly but I don't know about the dimensions of any alternative units.

  Posted by BAUHAUSLER on December 31, 2010 at 14:13:24 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  Is there a common diode offered by Radioshack that works for this project? Their website lacks some detailed info (at least that I can understand). Does a 1N5402 or 1N5404 diode work (item number 276-1143). Thank you.

  Posted by tommy_o on January 03, 2011 at 08:17:18 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  5402, 5404 OR 5408 will work. It needs to be at least 2A rating and 40V or higher.

  Posted by BAUHAUSLER on January 03, 2011 at 08:25:25 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  Thanks for the quick reply. I'll just make a note of the need and drop by Radioshack next time I'm in the next town over.

  Posted by tommy_o on December 31, 2010 at 14:38:48 Pacific Time

• Transformer Choice?
  You must be logged in to reply.
  
  
  Some of these once common items are no longer stocked in-store. If you call ahead they can tell you if they have it, or order it in. Sadly, electronic parts have been moved online to make way for more gadgets at RS.

  Posted by BAUHAUSLER on December 31, 2010 at 14:49:47 Pacific Time

• Using a Bridge Rectifier
  You must be logged in to reply.
  
  
  Hi,
  
  In your design you use 4 individual rectifier diodes, but would it be acceptable to use a rectifier package (like one of these: http://www.rapidonline.com/Electronic-Components/Discrete-Semiconductors/Bridge-Rectifier-Diodes/4A-SIL-Bridge-rectifiers/73805) in place of those diodes? Would anything need to be changed to accommodate for that?
  
  Also, (this is just out of interest rather than because I'd be brave enough to do it) is there any reason why you don't use the separate secondaries of a 12.6-0-12.6V supply to power the chips/channels?
  
  Thanks!
  

  Posted by JParkinson on November 21, 2010 at 04:04:37 Pacific Time

• Using a Bridge Rectifier
  You must be logged in to reply.
  
  
  Thanks for your questions. The 4 individual diodes are wired to form a bridge rectifier in this circuit. I used 4 diodes for space efficiency. A packaged bridge rectifier will function the same bit will have to be mounted differently. This circuit has a bipolar power supply with positive and negative voltages. A center tapped transformer winding and bridge rectifier, with the center tap grounded is the most parts-efficient way to achieve this. This is what the majority of solid state amplifiers do. There are other ways to configure bipolar supplies but they offer no advantages that I know of for this circuit.

  Posted by BAUHAUSLER on November 21, 2010 at 06:37:31 Pacific Time

• Using a Bridge Rectifier
  You must be logged in to reply.
  
  
  Thanks for the quick reply!
  
  Oops - I misread the power supply schematic, I can see now that it is indeed bipolar. For some reason 25.2V supplies don't seem easy to get in the UK so I'll probably be using a 24 or 28V supply instead.
  
  I'm not sure if I'll use the rectifier package or not, it depends what's in my parts bin...
  
  Thanks again for your time!

  Posted by JParkinson on November 21, 2010 at 11:35:28 Pacific Time

• schottky diode values
  You must be logged in to reply.
  
  
  the diodes are spec'ed at 4 amp or better in the parts list and 5 amp on the schematic, but the IN5822 on the list is 3 amp. We've got a bunch of these, but will they actually be ok in this circuit?

  Posted by Robin Biscuit on November 19, 2010 at 20:14:20 Pacific Time

• schottky diode values
  You must be logged in to reply.
  
  
  1n5822 is fine for this circuit. It's running near its max current rating only under extreme and unrealistic test conditions. I've torture tested the amp and never had a diode heat up or fail. I originally built this with fancy 10A Schottkys but they're difficult to source and work with. The 1N5822 was chosen because it's cheap, common and works just as well. You can also use SB340, SB360, MBR360 or any unit of at least 3A and 40V rating. The Schottkys are slightly quieter and have a little less voltage drop than ordinary diodes. I ran out of 5822s for one prototype and used 1n5408 instead (3A 1000V silicon) and those worked fine. They're about $0.04/ea in quantity.

  Posted by BAUHAUSLER on November 19, 2010 at 20:31:31 Pacific Time

• Ground Loop?
  You must be logged in to reply.
  
  
  So, I'm getting a bit of noise when the amp is at idle (music drowns it out, but it is still there, read more at http://www.gilberti-industries.com/2010/11/09/what-a-test/). I think my problem is a ground loop formed by the IC chassis being tied to each other electrically. Am I right thinking this? Will some nylon hardware fix my problem?

  Posted by scienceguy8 on November 08, 2010 at 23:42:41 Pacific Time

• Subwoofer?
  You must be logged in to reply.
  
  
  is there any way to mount the speakers inside the cabinet. maybe with a possible subwoofer.
  Thanks

  Posted by gmanx73x on November 13, 2010 at 12:26:38 Pacific Time

• Ground Loop?
  You must be logged in to reply.
  
  
  Nice looking build, and thanks for the mention on your website. Buzzing noises are usually caused by charging currents from the power supply getting into either the input or feedback circuits. This can be cause by combining grounds on the circuit board or by having a power supply lead close to one of the input pins or leads so inductance couples the signal to the input. This is why the circuit board in the article uses 'star' grounding of all commons to one point.
  Yes, the chips should be insulated from each other and from the heat sink. It looks like you have insulators under the chips. Nylon screws will work (you can use a metal nut). I buy nylon hardware at my local ACE hardware store, or you can order it cheaply in bulk off of eBay.

  Posted by BAUHAUSLER on November 09, 2010 at 06:02:10 Pacific Time

• Practice amp for bass guitar?
  You must be logged in to reply.
  
  
  Hi Ross,
  
  Thanks for the great article. This came to my door right around the time I was thinking of making a practice amp for my bass guitar. Other than the single source modification mentioned in the article, can you recommend any other modifications? Is there a way to combine the stereo outputs into a single speaker to double the power in mono?

  Posted by hbar on November 06, 2010 at 15:42:58 Pacific Time

• Practice amp for bass guitar?
  You must be logged in to reply.
  
  
  I'm glad you enjoyed the article. This is a very good circuit for a musical instrument amp. I wouldn't recommend trying to connect together the two separate channels into one to increase output. To do this you have to carefully match the gain from channel to channel, and use series current limiting resistors on each chip to force the 2 chips to equally share output. Even then you would get more volume only if you used a very low impedance speaker like 2 or 3 ohms.
  To use the full power of the amp from one mono source I would recommend building your practice amp with two identical drivers in one cabinet. Drive each driver from one channel of the circuit board. Connect your mono source to the input of each channel. This allows each channel to deliver full power into a speaker without changing the circuit.
  You could make a low-profile combo amp with a pair of inexpensive 8" or 6" drivers in a cabinet wider than it was high. I guarantee that with this circuit, two relatively sensitive drivers will play WAY over 'practice' levels of volume.
  For a wide range of moderately priced drivers, good specs and help on cabinet design I recommend the Parts Express website. They're a very large retailer of home, pro and musical instrument speakers and parts. I've used them often. Good luck, and check back with your results if you decide to go ahead with this.

  Posted by BAUHAUSLER on November 07, 2010 at 18:18:31 Pacific Time

• off-board components
  You must be logged in to reply.
  
  
  Ok so I am to the point of connecting the off-board components and have, what I hope to be, one final question. Step 4b says to connect the input jacks and switch terminals as follows, "input 1 left to switch 7, input 1 right to switch 9, input ground to switch 8". I understand the directions about left and right inputs, but am slightly confused about the ground. What exactly is the ground for the RCA jacks that hooks up to switch 8?

  Posted by connellydc on October 21, 2010 at 20:01:47 Pacific Time

• off-board components
  You must be logged in to reply.
  
  
  And BTW, congratulations for getting this far. That circuit board is pretty confounding to build. You're almost home free.

  Posted by BAUHAUSLER on October 21, 2010 at 22:14:12 Pacific Time

• off-board components
  You must be logged in to reply.
  
  
  The RCA input jacks have a signal terminal, which is the inner sleeve of the jack (this contacts the 'pin' at the center of the RCA plug). The metal body of the jack is the ground. For a pair of RCA jacks, connect the ground (body) connections together and wire to the indicated switch terminal.
  The headphone type jack indicated in the article has a 'body' ground connection that is for both the right and left channels on that jack. The right and left signal terminals on the headphone type jack are contacts that meet the 'tip' and 'ring' of the headphone type plug when it is plugged in. If you put a headphone jack into the plug and look inside you'll see how this works.
  Note that the instructions in the article assume the following: 1) you have the jacks mounted on an insulating plastic panel as described, so the grounds have to be connected through wires, and 2) that you are using a 3-pole 2-throw input selector switch where the 3 'poles' that it switches will be right signal, left signal and input ground.
  Several people have used 2-pole switches and have ended up with the grounds unconnected, causing problems. Search this page for the word 'hum' and you'll find a discussion of this.

  Posted by BAUHAUSLER on October 21, 2010 at 22:03:59 Pacific Time

• off-board components
  You must be logged in to reply.
  
  
  Thanks for the answer to my question and thanks for your ongoing support of this project. I would have never made it this far without it.

  Posted by connellydc on October 22, 2010 at 16:14:49 Pacific Time

• Mono or Stereo?
  You must be logged in to reply.
  
  
  This amp looks like it is configured as mono only. Is that correct or did I mis-interpret something?
  
  Thanks.

  Posted by FPV9090 on October 21, 2010 at 13:25:41 Pacific Time

• Mono or Stereo?
  You must be logged in to reply.
  
  
  Stereo. Two independent channels sharing one power supply and volume control.

  Posted by BAUHAUSLER on October 21, 2010 at 21:48:36 Pacific Time

• Mono or Stereo?
  You must be logged in to reply.
  
  
  In looking at the schematic, its very hard to see the two channels. The schematic only shows one speaker.

  Posted by FPV9090 on October 22, 2010 at 17:43:26 Pacific Time

• Mono or Stereo?
  You must be logged in to reply.
  
  
  The two channels are identical, sharing the power supply, volume control and selector switch. The schematic is therefore representative of either channel.

  Posted by BAUHAUSLER on October 22, 2010 at 21:59:43 Pacific Time

• signal generator
  You must be logged in to reply.
  
  
  I'm to the testing stage of my amp, but I don't have a dedicated signal generator and cannot really afford to buy one. Can anyone recommend a good free one I can get online?
  
  thanks

  Posted by connellydc on October 21, 2010 at 11:15:02 Pacific Time

• signal generator
  You must be logged in to reply.
  
  
  I'm not sure it's 'the best' but I've used SIGJENNY in the past and it operated as advertised. Some time ago I generated a set of sine wave files, 20 to 20K frequencies, 1/3 octave apart, 2 minutes long each. I burned them to a CD and I use that with a battery powered portable CD player for mobile testing.

  Posted by BAUHAUSLER on October 21, 2010 at 11:29:08 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  Hi Ross, thanks for writing this article. I used your core circuit but used a center tapped 12.5 x 12.5 torroid transformer and installed in a hammond chassis with headphone jack and it is working out great. It inspired me to build a speaker pair from plans (Microbe RS desktop speakers). However, the speakers are 4 ohms, and have custom built crossovers. When I play audio through my speakers, it sounds great until past 50% volume, then it starts to clip and the led begins to dim with each beat. I installed heat sinks on the outside of the chassis and they did get quite hot. I did the power calculations and found that 11 watts output into 4 ohms results in an impedance value of 7.56 amps and that RMS is 30.25 volts. These values exceed the ratings of the transformer. Could I substitute a higher rated transformer (ie Avel 160v 18vX18v) and would this solve my problem? Thanks again. M

  Posted by DJStugots on October 15, 2010 at 09:55:52 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  I'm glad you found a good use for this circuit.
  The current for one channel, 11W into 4 Ohms is sqrt(11/4) = 1.66A. The rms voltage at that current is 1.66 X 4 = 6.64v. the peak voltage is 6.64 X 1.4 = 9.3v.
  If the LED is blinking, then the power supply is being exhausted and the supply voltage is dropping. This is typically due to the transformer saturating from excess current being drawn. To produce 11W into 2 4 Ohm loads at the same time the transformer would have to be able to maintain full output voltage while delivering 3.3A.
  I specified a 2A rated transformer for this project because in practice the circuit would voltage clip on peaks before a full long term overload of the transformer's current capacity was reached.
  Try this: Put a DC voltmeter from ground to either power supply node. You should get about 17 to 18V with no signal. If you increase the volume to the point of distortion and watch the power supply voltage you'll probably see the voltage drop way down on musical peaks. This would indicate a transformer with insufficient current capacity.
  The chips will source up to 2A of current, so that's not the weak point. I drove this circuit to full output with sine waves at clipping several times (an abuse condition) to see what would happen. When the chips get too hot their internal protection circuitry just shuts them off until they cool off.

  Posted by BAUHAUSLER on October 15, 2010 at 10:10:14 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  Thanks, I miscalculated the values by squaring the quotient of power and impedance instead of taking the square root. The transformer I am using is a Triad 24V CT at 2.08 amps and is rated at 50VA. I have an Avel torroid 160 VA transformer that is 18Vx18V CT @ 3.64 amps that I could use. Would this exceed the power specs for the circuit? The rail voltage would be somewhere around 23 volts.

  Posted by DJStugots on October 15, 2010 at 11:18:28 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  Hi Ross, I decided to swap out the 12x12 volt 2 amp torroid for an 18x18 3.4amp one. I powered up and sounded great, but amp started smoking. I cut the power and found that both caps on the rectifier circuit popped. I replaced them with same caps and reverted back to 12x12v power supply and it is working fine. How would I calculate new capacitor values to handle the larger 18x18 power supply? Thanks.

  Posted by DJStugots on October 22, 2010 at 07:40:03 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  In the equations above is a formula for calculating peak voltage from RMS. All of your power supply capacitors experience peak voltage and must be rated at no less than 120% of that for safety.
  Have you checked the chip's data sheet to see if you will be exceeding the chip's allowable DC supply voltage with your substitute transformer?

  Posted by BAUHAUSLER on October 22, 2010 at 08:13:03 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  I checked the NatSemi spec sheet for LM1875 and it says the chip can handle up to +/- 30volts. With the new transformer I calculated that 18 volts per channel with an 8 ohm load would produce 40.5 watts at a current of 2.25 amps. The RMS voltage would be 17.92 volts, Peak of 25 volts so the cap would need to be rated at at least 30 V? The caps in your design were rated to 10 volts I believe.

  Posted by DJStugots on October 22, 2010 at 11:07:27 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  You're confusing transformer RMS voltage with amplifier output RMS voltage. The chip is limited to 2A of output current or 20W of output power, whichever comes first. You won't get more power out of it into an 8 Ohm load just by increasing the transformer size. Higher transformer voltages are useful for driving high impedance loads like a 16 Ohm speaker, where the limiting factor is not current swing but voltage swing and more voltage headroom is needed.
  No capacitors in the power supply are rated at 10V.

  Posted by BAUHAUSLER on October 22, 2010 at 11:21:56 Pacific Time

• Power Supply
  You must be logged in to reply.
  
  
  There's one other possible cause for power supply voltage to drop at high output. If the power supply capacitors are too small they won't store enough charge to maintain the voltage under heavy draw. I've built chip amps with as little as 1000uf/rail/channel and they work fine but the rails don't regulate (maintain voltage) very well. The power supply loses 'headroom' on peaks. That's why I specced 4700uf/rail for 2 channels, which seems to be adequate.
  An undersized power transformer or capacitor will lead to increased power supply impedance at high currents. The chip itself has a very high capability to reject power supply variations from its output (high CMRR) but it does need the voltage headroom to swing the output voltage adequately.
  Another issue to look at is the volume level and the efficiency of the speakers. If you're going for fairly high volumes and the speakers' efficiency is below average 11W might be marginal.

  Posted by BAUHAUSLER on October 15, 2010 at 11:01:00 Pacific Time

• You also build turntables?
  You must be logged in to reply.
  
  
  This is a really nice project! If i could find the time, i'd love to try to build one !
  
  You mention that you also build turntable in the article. Are there any photos/specs of a turntable you've worked on?
  
  I'm currently working on a special turntable project over at www.makingaturntable.com and I'd be curious to hear your feedback on what we're doing.
  
  Cheers!

  Posted by logikdev on October 12, 2010 at 08:23:02 Pacific Time

• You also build turntables?
  You must be logged in to reply.
  
  
  I repair and modify turntables but I haven't scratch-built one. I'm in a room right now with about 35 turntables, which is way too many. My most elaborate mod to date was adding full-function RF remote control to a B&O 4004. This allows me to control its functions - start, stop, pause, back, forward - from anywhere in the house.
  Here's a link into the photostream with some pix. http://www.flickr.com/photos/ross-hershberger/3327336572/

  Posted by BAUHAUSLER on October 12, 2010 at 08:48:36 Pacific Time

• You also build turntables?
  You must be logged in to reply.
  
  
  Nice work on the B&O! That's a pretty cool mod!
  
  What kind of TTs do you have lying around? Any projects in mind for any of those?

  Posted by logikdev on October 12, 2010 at 09:54:34 Pacific Time

• power supply
  You must be logged in to reply.
  
  
  Ok, so I finally got all the pieces I needed together to build this amp and went about constructing the power supply. When it came time to test it, I turned it on and there were no pops and the fuse did not blow, so that was a good sign, but I came across a problem when measuring the voltage from P10 to P13. The instructions say the meter should read about -18V, but when I put the positive lead of my multimeter to P10(which is the positive of the capacitor) and the negative lead of the multimeter to P13, the meter read a positive 18V instead of negative. I reversed the leads leads of the meter and then got a reading of -18V. Now I won't pretend to know much about circuits, but that doesn't seem right. I went back over the construction of the power supply step by step to see if maybe I had installed something wrong, but I didn't. Any advice on what might be wrong?

  Posted by connellydc on October 08, 2010 at 22:02:20 Pacific Time

• power supply
  You must be logged in to reply.
  
  
  Your circuit and measurements are correct. P10 is the ground of the circuit and the zero volts reference. P13 is the -18V point. If you connect the Common (black) of your meter to P10 (ground) and the red lead of the meter to P13 you should read -18V, as that node is 18 volts below ground. It's a bipolar supply, which many builders will be unfamiliar with. The chips require both a positive and a negative voltage to run. This arrangement allows both the input and the output of the chips to be referenced to ground, halfway between +18 and -18v. The audio signal then 'swings' negative and positive alternately, measured from ground.

  Posted by BAUHAUSLER on October 08, 2010 at 23:11:16 Pacific Time

• power supply
  You must be logged in to reply.
  
  
  Thanks so much for your help Mr. Hershberger. It is a relief to know that I won't have to take it all apart. Your explanation was quite useful and helped me gain a clearer understanding. Thanks again!

  Posted by connellydc on October 09, 2010 at 08:17:11 Pacific Time

• Advice for 220-240v?
  You must be logged in to reply.
  
  
  Can anyone recommend a transformer to use when building this circuit to be used with Australian (220-240) voltage? And am I right that the rest of the circuit needs no modification once the input voltage and current are sorted? Thanks!

  Posted by mark in pdx on September 24, 2010 at 21:28:20 Pacific Time

• Advice for 220-240v?
  You must be logged in to reply.
  
  
  25.6V is a standard voltage because it's a multiple of 6.3V. You should be able to find 240V/25.6V transformers locally. The current rating should be at least 2 amps. The circuit will work without modification with transformers from 24VCT to 30VCT if you can't find 25.6VCT. For 240V use, change the line fuse to 1/2 the value given. I think I specced 500ma, so use a 250ma fuse for 240V and make sure your power switch is rated for 240V. A lamp or appliance power switch almost certainly will be.

  Posted by BAUHAUSLER on September 24, 2010 at 21:55:11 Pacific Time

• Advice for 220-240v?
  You must be logged in to reply.
  
  
  Thanks for the quick & clear reply, Ross!

  Posted by mark in pdx on September 24, 2010 at 22:40:26 Pacific Time

• 830ohm, 1/2w resistor
  You must be logged in to reply.
  
  
  Anyone know where I can find the 830ohm, 1/2w resistor needed for the project? I've done some searching online and the closest I've been able to come up with is a 820ohm, 1/2w resistor from parts-express.com. Thanks in advance.

  Posted by connellydc on September 22, 2010 at 14:26:20 Pacific Time

• 830ohm, 1/2w resistor
  You must be logged in to reply.
  
  
  The author here. That must be for the LED circuit, right? There's no 830 in the audio circuit. Anything between 680 and 1K will work fine. Electronics is pretty much a +/- 10% business anyway. 1K will give you 17ma and 680 will give you 20ma. 1K should be fine. The power dissipation is 300mw or more, so use a 1/2w part, or if you're not proud, use 2 1/4W 2200 or 2K resistors in parallel, which might be easier to source.
  Good luck, and check in here if you have more questions or issues.
  Note: the LED has a narrow beam, so leave the leads longish and point it slightly upward for best visibility.

  Posted by BAUHAUSLER on September 22, 2010 at 21:52:21 Pacific Time

• Mini Chip Amp
  You must be logged in to reply.
  
  
  For the capacitors what temperature rating should i use. How can i determine how many amps the amp puts out per channel, also if the LM1875 is rated to put out 25 watts how come this amp only puts out 11 watts per channel?

  Posted by RHINO92 on September 18, 2010 at 19:00:21 Pacific Time

• Mini Chip Amp
  You must be logged in to reply.
  
  
  BTW, the math for current in an amp output is A = sqrt(P/R) where A = current in amps, P = power in watts and R = load impedance. So for 11 watts into 8 Ohms, A = sqrt(11/8) = 1.17A. The rms voltage is A X R = 1.17 X 8 = 9.36. Peak voltage = rms X 1.4 = 1.17 X 1.4 = 13.

  Posted by BAUHAUSLER on September 18, 2010 at 19:44:18 Pacific Time

• Mini Chip Amp
  You must be logged in to reply.
  
  
  Oops. I meant peak voltage = rms X 1.4 = 9.36 X 1.4 = 13.

  Posted by BAUHAUSLER on September 18, 2010 at 19:45:34 Pacific Time

• Mini Chip Amp
  You must be logged in to reply.
  
  
  Good questions. This amplifier does not stress its capacitors with heat, so any temp rating part will be adequate.
  Regarding power and current, the amplifier puts out about 2.5A total, 1.25A/channel. The chip is rated to put out more power than that. This amp uses a power transformer that is intentionally of lower voltage and current. This limits the power output of the amplifier. If it is built for full power output the transformer and heat sinking have to be much more substantial, expensive and difficult to build. A full-power version would gain maybe 3 or 4 dbs of volume, due to the logarithmic nature of the power/volume relationship. You would hardly notice the difference. We chose a compromise between power output spec and size/cost/complexity.

  Posted by BAUHAUSLER on September 18, 2010 at 19:20:11 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  I have finally completed the build on this amp, with everything going relatively well. But after powering up my amp I get a constant feedback over the speakers even when the volume pot is turned all the way down. I have checked the feedback resistors and they read 20K as they should. Does anyone have an idea where to start looking to fix this problem?
  
  

  Posted by gruntled on September 16, 2010 at 09:22:09 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  Please describe the feedback sound you get. How loud? Is it in both channels? About what frequency? Does the music signal come through with the feedback?

  Posted by BAUHAUSLER on September 16, 2010 at 09:28:49 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  The feedback is coming through both channels as a constant midrange humming Also has an occasional crackling sound. As the volume is increased the feedback seems to stay at a constant volume, as if it were independent from the input source. I did notice that when my finger brushed across the heatsink on the LM chips the feedback got much lower.

  Posted by gruntled on September 16, 2010 at 09:47:23 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  That sounds like a grounding issue. The circuit is picking up radiated 60hz signal out of the air and amplifying it. Check all of the grounds, starting at the inputs. Note that the build for the amp uses a 3-pole input switch that switches the grounds along with the L & R signals and connects them to the circuit ground. If you use a 2-pole switch instead you have to connect the input jacks' grounds to the circuit central ground. My money's on this as the solution.
  I'll be offline until around midnight EST, but I'll check in then.

  Posted by BAUHAUSLER on September 16, 2010 at 09:57:54 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  After soldering the main grounding wire that bolts to the transformer the humming was pretty much eliminated. but now the crackling is much. It is very loud even with the pot turned all the way down. i will check the rest of my grounds to see if anything else may be a bit loose.

  Posted by gruntled on September 16, 2010 at 16:39:38 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  Intermittent noises can often be located by tapping. Use a chop stick or other nonconductive poker and gently tap on parts and wires while listening to the crackling. Often you'll poke something and the crackling changes abruptly. This narrows down your search for the problem. It's also useful for locating cracked solder joints, which can be invisible.

  Posted by BAUHAUSLER on September 16, 2010 at 21:42:44 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  It's been a while since I have worked on this project. But I'm giving it another go. The main issue I have now is that when I cut the power to the amp, the fuse often pops. I have tried this with several power bars but with the same results. As soon as i turn off the switch there is a fairly bright flash in the fuse and that is the end of it. (It doesn't happen every time, but it is fairly often) I need to figure out what may be causing this before I can move on. Any ideas on where to look.
  I have removed and re-soldered all of the grounding lines to ensure that they are secure.

  Posted by gruntled on November 11, 2010 at 06:17:15 Pacific Time

• Speaker Feedback
  You must be logged in to reply.
  
  
  I'm not sure where you are with this project so I'm going to assume the following: 1) power supply is constructed. 2) two audio circuits are constructed and connected to the power supply.
  3) the fuse blows when you turn the unit ON. ('cut' usually means turn off).
  The first thing to look at is the insulation of the chips from the chassis/heatsink. The metal chip tabs are internally connected to the - power supply. If the chips connect through their tabs or through metal screws to the chassis, there will be a short from the power supply to ground that will cause the fuse to blow. Check that first & report back...

  Posted by BAUHAUSLER on November 11, 2010 at 07:13:51 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  Any ideas for integrating a pre-amp? I'm basically a novice but am playing with the idea of expanding this for vinyl. I'm curious if anybody has a quality schematic suggestion...

  Posted by jdbco2001 on September 14, 2010 at 15:56:04 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  This circuit has 20 dbs of gain, so it can be driven to full output by less than 1/10 volt of input. Any source component will drive it without additional amplification, and it has 2 inputs and a volume control built in. What are you trying to achieve with a preamp? More inputs? phono capabilities? Tone controls? Channel balance?

  Posted by BAUHAUSLER on September 14, 2010 at 20:51:36 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  I’m not yet familiar with typical output values from a turntable or the dbs gain required for reasonable vinyl listening volume, but that’s the idea. It would be nice to add phono capabilities (ideally as an additional input). Tone controls and channel balance would be great; however as a novice, I figured to hold off on this first time around. Any assistance for moving in this direction are certainly welcome. Although, the project and expansion alone are enough to begin with! Thanks!

  Posted by jdbco2001 on September 15, 2010 at 07:36:36 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  Phono is tricky. The gain required to step up a cartridge output to line level is between 30 and 40 dbs. Plus there's the RIAA equalization curve involved, so it's not straight gain. Noise is a problem. For turntable use I would recommend purchasing one of the small stand-alone phono preamp "adapters" that will take the turntable output and produce a line level signal. This signal could then drive the Squelette.
  Tone controls are another set of filters. For that you can buy assembled circuit boards with working, tested circuitry. You add a power supply and input/output connections. They're inexpensive and almost always based on opamp chips. I look on eBay for that kind of thing. Start with the search terms "diy tone control' and see what comes up. If you decided to incorporate the tone controls into a Squelette build you could derive the power supply for the tone board from the Squelette power supply with a couple of 12V/-12V fixed voltage regulators.

  Posted by BAUHAUSLER on September 15, 2010 at 07:57:54 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  Any possibility for a future MAKE project? I seem to know a lot of audiophiles with a nice vinyl collection. Just like amplifiers there is debate about what is a quality phono preamp. I can settle with incorporating an adaptor, but how great would it be to make one!
  
  Thanks for sharing this and your lightening fast advice!

  Posted by jdbco2001 on September 15, 2010 at 10:09:04 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  Regarding making a phono preamp, it's one of the more challenging things in audio to tackle. With signals on the millivolt scale being amplified 4 orders of magnitude it's critical to control noise, so phono circuits come with a fairly demanding set of requirements. If I developed one it would probably be based on the old RCA vacuum tube phono preamp from the '50s. Many people still use that simple circuit and tubes are awesome.

  Posted by BAUHAUSLER on September 15, 2010 at 10:28:10 Pacific Time

• pre-amp
  You must be logged in to reply.
  
  
  Thanks for asking. It's possible, and I've researched a number of vacuum tube based phono and tone control circuits for possible future projects. They're not especially complex, but the build description is lengthy because of the number of parts and the need to hand-wire it all. A long build means a long article, and the project had better have wide interest and applicability if it's going to occupy a lot of pages. I certainly think it's cool, but for Make readers it might be more niche than something like Squelette.

  Posted by BAUHAUSLER on September 15, 2010 at 10:22:58 Pacific Time

• Working on my own.
  You must be logged in to reply.
  
  
  Thank you for publishing this article. It could not have come at a better time. I decided to make a few modifications, however. You can read about my progress here.
  
  http://gilberti-industries.com/?p=4

  Posted by scienceguy8 on August 29, 2010 at 15:09:56 Pacific Time

• Working on my own.
  You must be logged in to reply.
  
  
  Thanks for your interest in this article and for the work that you put into adapting it to your needs. I don't have the reader to look at your PCB layouts. I want to warn about one thing. NatSemi is very specific about the handling of the grounds in the circuit. They specify star grounding to a central point for the power supply and audio circuit grounds. The input ground reference, power supply bypass caps, feedback, etc. should all have their currents carried to the 0V star ground on separate leads. The manufacturer warns that mixing currents could lead to noise, distortion and instability. Typical PCB ground plane practice allows currents to interact in unpredictable ways. The chip has an open loop gain of 90dbs and any spurious signals on the input or feedback circuits could be amplified.

  Posted by BAUHAUSLER on August 29, 2010 at 15:24:01 Pacific Time

• Working on my own.
  You must be logged in to reply.
  
  
  Thanks for the quick heads up. Good thing I didn't send off the files for fabrication yet. So all I have to do is tie each grounded device back to the ground between the two 4700uF caps with their own dedicated lines? Sounds simple enough.

  Posted by scienceguy8 on August 29, 2010 at 16:16:10 Pacific Time

• Working on my own.
  You must be logged in to reply.
  
  
  That's correct. To make sure, you should look at the chip datasheet's section on grounding and follow their recommendations.
  I think you'll really like the sound of this amp. It's ridiculous the quality of reproduction considering the small parts count and low cost. Please post impressions and pix as your build goes along. Good luck!

  Posted by BAUHAUSLER on August 29, 2010 at 16:25:48 Pacific Time

• Transformer.
  You must be logged in to reply.
  
  
  The specified transformer is 25.2VCT (12.6/0/12.6). A 28VCT transformer (14V/0/14V) could be used. Expect the DC voltages measured in circuit to be 1 to 2 volts higher. Make sure all of the capacitors that you use have a voltage rating of at least 35V to give a safety margin. The power output will be slightly higher. Use the same value line cord fuse for safety.

  Posted by BAUHAUSLER on August 14, 2010 at 21:16:10 Pacific Time

• Transformer Substitution
  You must be logged in to reply.
  
  
  Any chance I can use a 28V transformer instead, or is this too much juice?

  Posted by Racharg on August 14, 2010 at 14:43:03 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  I'm pretty much a novice at this but I've read through this article (twice) and think I can build this. Are there any suggestions on how would I modify the project to function as a dock and charger for an Ipod Touch (or iphone). Ideally I'd like to use the chipamp as my volume control.

  Posted by wrlinville on August 14, 2010 at 08:42:05 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  The simplest way to incorporate a dock/charger for an Apple product is to purchase an aftermarket dock of appropriate size/shape, take it apart and build the connectors into the amp cabinet. I'm not sure what the power source for the charger is, but if it's 5VDC you could 'steal' some current from the positive voltage rail of the amp and run it through a LM7805 voltage regulator chip to step it down to 5v for the dock. I'm not certain that's what's needed, but that's where I'd start.
  As for the audio outs, you'd have to trace the wiring in the dock that goes to the right and left audio outputs and connect those wires to the volume pot of the amp circuit. That's actually simpler than the 2 inputs and selector switch described in the article.
  If you're new to this kind of construction take extra care building the circuit board. There are a lot of solder joints. It's a good idea to re-trace your last work and check the steps every time you pick up the project again to make sure no errors creep in.

  Posted by BAUHAUSLER on August 14, 2010 at 21:24:35 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  So, if I borrow the 5 volts would it lessen the voltage available to the speakers and if so, would that cause fluctuations between when the ipod was charging and when it was not charging? You can probably tell by the question that I'm shooting in the dark here, but I'm reading more and trying to learn.

  Posted by wrlinville on August 20, 2010 at 10:16:19 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  The voltage regulator will take the +17V of the amp power supply and chop it down to (we think) the +5V required for the Ipod. Better check what the charging needs of the iPod are. The power for the amp would only be affected if the iPod was drawing a lot of current (more than 750ma) while the amp was trying to put out full volume. Below full power amp output the power supply of the amp has plenty of current to run both. The iPod has a very small battery and its charging current is likely to be a trickle that will never be missed by the amp power supply.
  I recommend Googling around to see if someone has posted the specs for charging your iPod. Or look up the spec from Apple. They might give it in terms of voltage and current, which would be ideal.

  Posted by BAUHAUSLER on August 20, 2010 at 10:29:22 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  I'm trying to do this as well.
  
  I was looking at doing the LM7805 but my teacher said it would get too hot and that I should use a switching regulator. They're much more efficient but are also a lot more expensive.
  
  I'll let you know what I come up with.
  
  I want to build some of these amps and put them inside vintage radios. Then sell them to my hipster friends at inflated prices.

  Posted by mukelarvin on January 28, 2011 at 08:08:53 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  I like the idea of chip amps in vintage radio cases! I have a heap of old radios here that would make nice desktop ipod docks.
  As for the efficiency issue, take a look at the specs for the LM7805 relative to the current draw of the device you're powering. Power dissipation is strongly dependent on the current drawn through the regulator. I don't think an iPod actually draws that much juice.
  If you're putting the chip amp in a larger case, you can make a dedicated 5V power supply simply with a filament transformer, a bridge rectifier, a single electrolytic cap and a regulator. That would be more power efficient since much less voltage is being dropped in the regulator.
  Or use the wall wart power supply that comes with many iPod docks and just wire it into the 120V power cord.

  Posted by BAUHAUSLER on January 28, 2011 at 08:29:19 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  The LM7805 I have here says it can handle 30V, 1A.
  
  Would I be able to branch off of the VCC+ like this?
  Poor Drawing #1
  
  I read how the mintyboost powers the USB port and thought I could just branch off of the 5V like so?
  
  < a href="http://mukelarvin.com/?attachment_id=692">Poor Drawing #2

  Posted by mukelarvin on January 28, 2011 at 23:02:39 Pacific Time

• Squelette with Ipod
  You must be logged in to reply.
  
  
  Yes, the 7805 should be able to handle the Squelette's +17V input and pass enough current for what you're planning. If you use this, monitor the temp of the 7805 and add a small heat sink if it gets uncomfortably warm.
  The USB powering scheme that you show is one that I've seen before. I haven't done that, but duplicating the Mintyboost setup for 5V power is probably pretty safe.

  Posted by BAUHAUSLER on January 29, 2011 at 14:44:51 Pacific Time

• Schematics don't match parts list
  You must be logged in to reply.
  
  
  The online schematics show a 100Ohm pot. The parts list in the magazine shows a 100KOhm pot. I'm guessing the latter is correct here, but you should probably clear up the confusion.

  Posted by Jack D. on August 11, 2010 at 23:07:25 Pacific Time

• About the sound
  You must be logged in to reply.
  
  
  And another thing... I belong to an audiophile club/beer drinking society with a lot of DIY-ers. I took a Squelette to a recent meet to show it around. Some prankster thought it would be amusing to substitute this lowly chip amp into the high-end system being demonstrated, in place of an exotic single ended triode tube amplifier. Everyone expected humiliation for the lowly solid state amp compared to the expensive tube job. The real surprise was how beautiful this circuit sounds connected to quality speakers and sources. Nobody could find a single thing to fault about the performance, and these are some pretty critical listeners. At the next club meet I was asked to speak about chip amps and handle questions. I think a few projects were hatched or restarted based on the sonic qualities of this circuit. So in addition to being a pretty easy audio amp to build, having a useful amount of power and looking nice it holds up extremely well sonically.
  The circuit is taken directly from the national Semiconductor datasheet for the chip. The build that I developed incorporates all of NatSemi's notes for best performance, stability and durability. I believe this is the same circuit and chip that 47 Labs uses in their multi-thousand-dollar Gaincard chip amplifier, which set the audiophile world on its ear when it was introduced.

  Posted by BAUHAUSLER on August 05, 2010 at 17:30:10 Pacific Time

• On enclosures
  You must be logged in to reply.
  
  
  The author here again. The aluminum and Plexiglas enclosure shown in the article is one option, but I hope that readers will use this circuit in their own unique variations on the amplifier. Only a relatively small 6 square inch piece of aluminum, with insulators, is needed to heat sink the chips. Apart from that the chassis mounted component locations are not critical. Just add longer wires and you're fine. The circuit will work perfectly well in a lunchbox, cigar box, a suitcase with built in speakers, a hollowed out dictionary hidden on a shelf, old radio cabinet or whatever you can imagine. Think Steampunk or futuristic. A Boing Boing post shows me with an early prototype built in a painted radio Shack plastic project box with aluminum bottom. The transformer is in a separate cylinder made of plumbing pipe with a trailer wiring harness for a power cable.
  Uh... link here:
  http://www.boingboing.net/2010/08/02/my-favorite-makers-a.html

  Posted by BAUHAUSLER on August 05, 2010 at 17:00:34 Pacific Time

• Cost of parts
  You must be logged in to reply.
  
  
  We've had a comment from one reader that pricing out the components from Radio Shack exceeds the $50 cost estimate in the article. I don't doubt this is true. The cost estimate assumed a builder with some common parts on hand. Buying these off the rack individually is an expensive way to do it. Here are some tips for reducing the cost of the build.
  I love Radio Shack because they're right around the corner when I need something in a hurry. I shop there because I want them available when I need them. They are about the most expensive way to buy small parts, however. I buy a lot of stuff online. For perplexing reasons of global commerce the gas to drive across town to buy a transistor from the commercial electronics supplier costs me more than having 10 or 20 transistors delivered to my door from China. I also buy 'kits' of resistors and caps online. 1,000 or 2,000 resistors in a range of values for $10. If it's way more than I'll ever use I'll split an order with a friend & we'll both have a lifetime supply. I get a lot of parts on eBay. I maintain a 'want list' of stuff I'm looking for or that friends want, and when the list gets long enough I'll go looking for one or two sellers with combined shipping and get a ton of stuff all at once for low shipping. Sometimes when it arrives I can still remember why I wanted high voltage depletion mode MOSFETS or whatever. Buying bulk and sharing can drive the cost of a $0.50 resistor down to $0.005 or less.
  As far as the metal and plastic go I'm a scrounger since I worked in a machine shop. Commercial materials suppliers often have offcuts - short ends and leftovers from big jobs - that are great for hobby work but too small for their regular clients. They will often sell you these for a nominal charge. Go in when they're not busy, respect their time, be cool and they'll often set you up with what you need cheaply. If someone takes care of you, tell your friends and promote their business.
  Techies tend to be packrats. Sometimes an email or two will turn up the part that you need from someone's stash in less time than it takes to order it. Keep in touch with your fellow DIY-ers and share to help each other out.
  There's more information on parts dimensions that will help builders construct the board according to the instructions. I'll post that in a day or so.

  Posted by BAUHAUSLER on August 05, 2010 at 16:37:18 Pacific Time

• Cost of parts
  You must be logged in to reply.
  
  
  "There's more information on parts dimensions that will help builders construct the board according to the instructions. I'll post that in a day or so." It's been almost two months now.

  Posted by tullyf on October 01, 2010 at 08:36:24 Pacific Time

• Cost of parts
  You must be logged in to reply.
  
  
  Right there at the top. ERRATA. That's the promised info on parts dimensions and it's been there for quite a while.

  Posted by BAUHAUSLER on October 01, 2010 at 09:01:23 Pacific Time

• Author of Mechanical drawings
  You must be logged in to reply.
  
  
  Ross Hershberger here. The mechanical drawings should be credited to Regina Tuskey. She's my wife and a fantastic CAD machine designer. This task was child's play for her and took less time and effort than making a cup of tea.

  Posted by BAUHAUSLER on July 13, 2010 at 18:17:35 Pacific Time