MAKE: Noise — Discuss this article

Showing messages 1 through 23 of 23.

• Did more testing
  You must be logged in to reply.
  
  
  Which leads to the need for more testing
  
  I put the 92 watt cooler on the rig, and it did great.
  
  it was able to output just over 8 volts no load, and 2.5 amps at .1 v.
  
  Again at over 400 degrees f, same as the other tests.
  
  now I need to test at what minimum temps the 92, and the 46 make useable voltage. I have the data for the 92 I just tested, and hopefully within the week will gather the same for the 46.
  

  Posted by mis4tun81 on December 02, 2008 at 14:08:30 Pacific Time

• different watts
  You must be logged in to reply.
  
  
  how does different wattage on the peltier cells affect the output?
  Will a 80w peltier over a candle produce more or less volts than a 40w?
  And how about the current, will it be higher or lower than the 40w?
  

  Posted by maneuver on November 24, 2008 at 02:12:01 Pacific Time

• different watt RESULTS!!!!
  You must be logged in to reply.
  
  
  I spent a lot of time trying to answer the question of which peltier to pick.
  
  I bought some 12 volt 46 watt and some 12 volt 136 watt coolers since they were the same physical size and built a test rig to clamp them one at a time between an aluminum block on the hot side, and a zalman 9500 heat sink (with the fan run off an external power supply).
  
  I supported the test rig over a heat gun and was able to get the hot side up to over 400 degrees f. ( I agree this is way too hot), however I wanted to see what the absolute highest output voltage I could get from them.
  
  Once I found that maximum output voltage, I used a programmable constant voltage load to see the maximum current output at fixed voltages. And the results are very predictable and can be used to figure out which cooler to use in your application.
  
  At some point I may do additional testing with a candle as a heat source, however up till now I was just looking at the capabilities of the different coolers. I believe the candle is the weak link in getting consistent power from the cooler and I am still looking for a more reliable and consistent source of heat (maybe an oil lamp) to use.
  
  The results can be categorized by the following characteristics:
  
  The lower the Peltier wattage the higher the maximum (no load) voltage output.
  
  The 46 watt cooler delivered 8 volts at 400f the 136 delivered just over 6 volts.
  
  The higher the Peltier wattage the higher the short circuit current output (0.1v).
  
  The 46 watt cooler put out a maximum of 1.3 amps at 400f, and the 136 watt reached just under 3 amps.
  
  As you know one is useless without the other, so I tested a range of voltages between these extremes to graph the output capabilities of both the coolers, and the graph is a straight line between those points.
  
  The 46 watt cooler was most efficient at delivering 4-5 volts, and the 136 watt cooler most efficient at 3-4 volts, if you can get them hot enough (which I believe is the problem anyone who is suffering from poor output is having)
  
  As I said I tested at temperatures too high for regular use, but it clearly shows if you ever want the possibility of 5 volts out of a single cooler, get a lower watt cooler, and if you are thinking about running coolers in series, get higher watt coolers.
  
  I think the 136 watt coolers I used will be useless at candle energy levels, so by higher watt I think maybe 60 to 90 watts should be ok. I have a 90 I haven’t tested yet maybe I will test that at some point but I assume its results will fall between the other 2.
  
  Finally, the fan on the zalman 9500 cooler I am using draws more power than either cooler will (almost) ever put out, I picked it because at some point I believe it is good enough to use it without a fan, and I can use all the power generated for whatever my intended output.
  
  L8r
  

  Posted by mis4tun81 on December 01, 2008 at 09:49:52 Pacific Time

• LM117?
  You must be logged in to reply.
  
  
  Where can I pick up 117 IC?

  Posted by LarryBrennan on October 24, 2008 at 19:40:56 Pacific Time

• Little Voltage
  You must be logged in to reply.
  
  
  I have followed the design with only reasonable differences, and have yet to produce anything greater than 1v. I have even tried two TEC's in series. When I shopped Ebay for the TEC's I noticed many varying voltages, and chose 12v. Was this a mistake or does it make little difference since the voltages produced fall well below this threshold.

  Posted by cable489 on October 05, 2008 at 22:24:19 Pacific Time

• Re: Little Voltage
  You must be logged in to reply.
  
  
  I had similar problems when I first put my kit together. I got up to a 4V, 200mA output by hacking the candle (mine is a 12V unit as well).
  
  The wick of my tea lights, when new, weren't embedded in the wax, so I made a replacement wick with some heavier string dipped in melted wax (from a lit candle). That got me a much bigger flame.
  
  Another candle mod I tried involved a bent piece of wire pushed into the wax and a piece of the same waxed heavy string draped over it. the result was 2 extra, heavy wicks that stood half an inch or so out of the wax. The candle had a pretty big flame, but within a few minutes, it melted all of the wax in the tealight which then drained out a hole in the bottom of the metal holder.
  
  Getting back to the possible power output. I figure that to be most useful, I'll need to use 4 Peltiers - 2 pairs wired in parallel, which are then wired in series. From earlier experiments, I could probably get 8V and 400mA. With a little voltage regulation, it should easily be enough to power a device from a USB plug. Plus, with the 8V maximum, the fan will get up to its maximum speed fairly quickly.

  Posted by Invader_Grug on October 07, 2008 at 15:14:18 Pacific Time

• Re: Little Voltage
  You must be logged in to reply.
  
  
  Excellent tips here - I really like the candle hacking approach. I found that you can just about get 5v from one cell, but it's really running it at the edge of capability.

  Posted by Andrew Lewis on October 07, 2008 at 15:58:32 Pacific Time

• Using a solar death ray?
  You must be logged in to reply.
  
  
  How about powering a group of these on a flat panel with heat from a solar death ray? (Parabolic mirror dish?)
  
  Possible?

  Posted by sherab on October 01, 2008 at 20:25:01 Pacific Time

• Thermopile?
  You must be logged in to reply.
  
  
  might this device also be called a thermopile?
  
  http://www.dself.dsl.pipex.com/MUSEUM/POWER/thermoelectric/thermoelectric.htm

  Posted by diyed on September 30, 2008 at 06:58:40 Pacific Time

• Thermopile?
  You must be logged in to reply.
  
  
  Yes, that is a valid name for this technology. :)

  Posted by Andrew Lewis on September 30, 2008 at 07:16:21 Pacific Time

• Seebeck improvements.
  You must be logged in to reply.
  
  
  In the article you say "use plenty heatsink grease" to increase the output - that will have the opposite effect ! You need to improve the connection between the cell and the sinks sure, the only way to do that is to make the surfaces FLATTER and use LESS compound. Lap the surface of your heatsinks on wet and dry paper tacked on a glass plate, or, better still, use some grinding grit on the glass directly. Correctly flattened, all you should need is a translucently thin film of grease.
  
  Steve

  Posted by steve@thetaylorfamily.org.uk on September 29, 2008 at 09:02:33 Pacific Time

• Seebeck improvements.
  You must be logged in to reply.
  
  
  I agree that the contact surfaces need to be as flat as possible, and polishing the surface is a good idea.
  
  The noun "plenty" in this context is intended to convey the term "A full or completely adequate amount or supply", as defined in http://www.thefreedictionary.com/plenty.
  
  I advocated the use of thermal transfer compound because the steel surface of of a tin can is typically too thin to hold it's shape without warping under heat stress. Polishing the end of a can to get a mirror-flat surface is sometimes difficult because there is usually a ridge around the end of the can, and the tendancy towards flexation under pressure can result in an uneven surface.
  
  A possible solution would be to replace the end of the can with a copper disc thick enough to resist flexation. The thicker copper disc could be polished more effectively, and would conduct the heat more effectively to the cell.

  Posted by Andrew Lewis on September 30, 2008 at 07:14:01 Pacific Time

• Seebeck improvements.
  You must be logged in to reply.
  
  
  this is good stuff.
  
  Also, What about water cooling? put a reservoir cup above and solder together a tube that runs to a lower "tank" that is flush to the surface of the peltier module?

  Posted by diyed on September 30, 2008 at 06:32:40 Pacific Time

• Seebeck Generator Efficiency
  You must be logged in to reply.
  
  
  Out of curiousity, did you go through any revisions in creating this design? At first I thought that a fan would be unneccessary, same for the gasket. As I thought about it, I realized that both would create a larger temperature differential, by helping the cool side pull more heat from the hot side, making the peltier module come closer to optimal output. I am inferring that the higher your temperature differential, the closer you can come to max output. I also read that the peltier modules are only about 10% efficient, so I am guessing that if you need to generate close to theoretical maximum output on the module you need to really try to optimize.
  Lastly, I am not an EE or even an electrical hobbyist at this point, but I am intrigued by the possibility of stepping up the voltage to power a rugged notebook. I have seen a wide variety of Peltier modules on ebay but what I am not understanding yet is how far I can boost output on a smaller unit, or when I need to go to a larger units. I saw some that came close to the 90 watts I'd need, but I figure there's more to it than that. Need to read up!

  Posted by diyed on September 29, 2008 at 06:31:50 Pacific Time

• Seebeck Generator Efficiency
  You must be logged in to reply.
  
  
  In trying to educate myself, found this good source for formulas aiding generator design decisions:
  http://www.ferrotec.com/technology/thermoelectric/thermalRef13.php

  Posted by diyed on September 30, 2008 at 06:40:38 Pacific Time

• A design that could withstand heat from a cooking fire?
  You must be logged in to reply.
  
  
  I have done a little research and it appears that the TEC modules have an upper end of about 180 degrees C. However, I though that if you used some Thermo-trex 2800 wire (resists up to 600 deg. C continuous heat) and a good enough heat sink, you could have a power generation module that you could suspend over a cooking fire or mount in a chimney type arrangment that could be used to power laptop computers by burning biomass...
  
  interesting idea, but I think the TEC modules are the weak link. any thoughts?

  Posted by diyed on September 28, 2008 at 14:56:51 Pacific Time

• Circuits and Diagrams
  You must be logged in to reply.
  
  
  
  The link appears to be broken.

  Posted by 3LAN on September 25, 2008 at 20:10:27 Pacific Time

• Seebeck generator - What is the gasket for?
  You must be logged in to reply.
  
  
  
  Cheers,
  
  just wondering what's the exact function
  of the gasket cut from the PCB-board?
  Is the PCB-board plated with copper from
  one side or two? Somehow I guess you shouldn't allow the heat to escape via gasket, past the Peltier-cell.
  (So the plastic layer inside PCB-board
  is important, and one should not use a pure copper plate instead?)
  
  Yours,
  
  Kartturi
  
  PS. Same problems with Circuit-page as the above poster had.
  
  

  Posted by Kartturi on September 18, 2008 at 13:49:31 Pacific Time

• Seebeck generator - What is the gasket for?
  You must be logged in to reply.
  
  
  absolutely correct - the gasket is there to help stop heat transfer - I just used a piece of PCB because I had some handy - anything that produces a thermal break would be fine. Copper sheet would be a very bad idea :)
  
  Andrew

  Posted by Andrew Lewis on September 18, 2008 at 14:10:55 Pacific Time

• Seebeck generator - How much power?
  You must be logged in to reply.
  
  
  Thanks for the reply.
  Another question: How many watts (well, approximately, with some "standard candle", huh) you can take from
  this model, after you subtract the
  power needed by the fan?
  That is, how much juice you have for
  some external task, like recharging
  a mobile phone?
  
  I guess one needs to measure this only
  after the heat sink has set to some
  constant temperature, where it is kept
  by the action of fan. While immediately after starting the machine, the temperature difference between the can top (licked by flames) and heat sink (still at room/outdoor temp.) is larger,
  thus the amount of electricity generated by Peltier cell is also larger?
  (Please correct me if I'm wrong.)
  

  Posted by Kartturi on September 21, 2008 at 12:05:17 Pacific Time

• Diode part numbers
  You must be logged in to reply.
  
  
  I've just spotted a typo in the part numbers for the diodes mentioned on page 129 of Make magazine. They should be 1N4001 or 1N4002, not "LN".
  

  Posted by c0redump on September 15, 2008 at 04:20:30 Pacific Time

• Diode part numbers
  You must be logged in to reply.
  
  
  well spotted ;)

  Posted by Andrew Lewis on September 15, 2008 at 05:06:13 Pacific Time

• Circuit diagram
  You must be logged in to reply.
  
  
  The link to the diagram does not seem to work - I get a 404 error.
  
  I'm using explorer 6 and the latest firefox - neither worked.

  Posted by AlexanderDumas on September 13, 2008 at 14:31:08 Pacific Time