Workshop
Espresso Machine Brain Repair


I returned home from a recent trip to a tragically non-functioning espresso machine. When I powered it on, I could hear a relay clicking, probably trying and failing to turn the pump motor on. On a commercial-grade machine like mine, a Fiorenzato Bricoletta E-61 design, this often means your microcontroller box, or brain, is kaput. This being an Italian-made part with no benefit of consumer-level economies of scale, the replacement is about $200. Yikes.

I put my multimeter on the pump motor wires and could see it spiking at about 28V, but never supplying sustained current. This helped confirm that something was ill in the machine’s brain, but I hoped it might be a single component, rather than the whole unit.

I put this video up on my favorite coffee forum, Home-Barista, to see if anyone could help me diagnose it:

This lead to a correspondence with James “Pat” Boyt, who said it looked like a “dried up” power capacitor to him. I pulled the brain, tested the big brown capacitor in the above photo, and sure enough, it was dead. I desoldered it, replaced it with a new 330uF 35V capacitor rated to 106 C, and reconnected the brain. I plugged in the machine, flipped the power switch, and it powered right up. Hurray! That’s a 19 cent part, instead of $200.


There’s nothing quite like the internet for getting expert advice on a problem. I’m grateful for Pat’s guidance in bringing my machine back to life.

30 thoughts on “Espresso Machine Brain Repair

    1. Congratulations!

      Based on my experience, electrolytic caps are usually the culprit. Especially on a solid-state board that lives in a warm environment. It’s where I always look first.

  1. It’s truly amazing what a bad cap can do, glad to see you solved the problem, rather than creating a landfill problem. Seriously, big kudos for that and the example you’ve set.

    This reminds me of the famous bad capacitor plague of the mid 2000’s, thousands of really expensive (and toxic) machines ended up as landfill because of shoddy corporate espionage. The chemical formula was improperly copied, leading to bad capacitors that failed in about a year.

    Modern supply chain quality control has become so very critical in general, but you don’t hear much about it.

  2. It’s truly amazing what a bad cap can do, glad to see you solved the problem, rather than creating a landfill problem. Seriously, big kudos for that and the example you’ve set.

    This reminds me of the famous bad capacitor plague of the mid 2000’s, thousands of really expensive (and toxic) machines ended up as landfill because of shoddy corporate espionage. The chemical formula was improperly copied, leading to bad capacitors that failed in about a year.

    Modern supply chain quality control has become so very critical in general, but you don’t hear much about it.

  3. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

    1. Great question. I think I spent an hour or two trying to isolate the problem, 1/2 hour shoot a video and posting on home-barista, corresponding, finally about 15 minutes desoldering the bad cap and soldering in the new one, and reconnecting things. If I’d relented and bought the $200 Gicar unit, it would have been after the first two stages (poking around, video), so the post-capacitor-revelation labor cost would have been fro 15 minutes.

  4. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

  5. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

  6. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

  7. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

  8. Way to go on the diagnosis! Not to mention avoiding that $200 price tag. Out of curiosity, how many people-hours did it take from “something’s wrong” to finding the dead capacitor? I’m wondering what the “labor cost” was on this.

  9. Good job! Sometimes, the most satisfying repairs are the ones you do on the things you rely on the most. Glad to know that you could isolate and replace the broken part yourself. I hate it when products are difficult to repair and you have to throw it out.

    1. I have to credit Pat with that — I think he’s learned it by experience. The motor was trying to turn, in tiny little pulses, and when I put a meter on it I could see the brief spikes in current. Since the motor turned freely when I put a screwdriver in the rotor slot and turned, it wasn’t (thankfully) a damaged motor or pump. I’m conjecturing here, but I suspect all this lead Pat to think that the microcontroller was telling the system to turn the pump, but some discreet component was failing its job.There are few enough of them on the boards that you could test them one by one, but Pat’s guess pointed me toward the failed capacitor on the first try. I’ll ask him for a more scientific answer, if he has one.

    2. I have to credit Pat with that — I think he’s learned it by experience. The motor was trying to turn, in tiny little pulses, and when I put a meter on it I could see the brief spikes in current. Since the motor turned freely when I put a screwdriver in the rotor slot and turned, it wasn’t (thankfully) a damaged motor or pump. I’m conjecturing here, but I suspect all this lead Pat to think that the microcontroller was telling the system to turn the pump, but some discreet component was failing its job.There are few enough of them on the boards that you could test them one by one, but Pat’s guess pointed me toward the failed capacitor on the first try. I’ll ask him for a more scientific answer, if he has one.

  10. Your post is awesome!
    I turned on my ECM Giotto today and heard the exact same clicking noise you mentioned. I did a couple of google searches and came across your posting. I knew immediately that what you described was my problem, then I heard it in your video.
    Based on your diagnosis of a dead capacitor we got the machine up and running in no time, and it cost us only 59cents! Thank you, thank you, thank you!

  11. Super details… Well done. I found your story as I was trying to diagnose my Giotto’s new habit of Tick Tick Tick Ticking for a long (and ever longer time) before running. Was given courage to open the controller box, Relieved to find only one capacitor, removed it, tested (way out of spec) and Ebayed up a fresh one (ok 4 was the smallest number) of that same condenser. 330uf 35V 105c (I think 105) and look forward to getting back in business $199 to the good.

Comments are closed.

Tagged

John Edgar Park likes to make things and tell people about it. He builds project for Adafruit Industries. You can find him at jpixl.net and twitter/IG @johnedgarpark

View more articles by John Edgar Park