Hi All.

I messed something up here. I have an '80s era S27 that has been working beautifully. This is a rotary motor machine which produces full pressure almost instantly, and the espresso always tasted somewhat bitter to me. In the quest for improvement, I installed a delay-on-make relay before the motor a few weeks ago. I set it to a 6 second delay, and It definitely improved the flavor. The coffee seemed better balanced. I was pretty pleased with myself (always a predictor of trouble).

Things were going fine, except that it was now a pain to clean the machine, since it took 6 seconds before the pump would engage when I'm trying to clean soap out of the 3 way and portafilter head. I decided to install a dpdt switch, so I could bypass the relay for cleaning and rinsing. I found a nice looking toggle switch, wired everything in, and found that the pump immediately came on when the toggle was thrown, instead of being triggered by one of the dosing buttons. I assumed I'd wired something wrong, so I took out the dpdt switch and went back to just the relay. Same situation, pump goes on six seconds after the machine is turned on. I then completely took out the relay, returned everything to original wiring and got the same result--motor comes on when the machine is turned on. To be clear, the machine isn't dispensing any water but the pump is definitely engaged and running.

This machine has a funky old circuit board that visually looks OK--nothing seems burned. The fuses look good. Can anybody give me advice on where to look next? I now have the pump wired to a switch, so I can manually turn it on after I push one of the dosing buttons. This is better than a blinding withdrawl headache, but will get old in a hurry.

OK, I think I'm narrowing this down. It sounds like I shorted a TRIAC relay on the logic board. I found a few references to the pump running continuously when the TRIAC gets fried. It dies in the open position, allowing continuous current to the motor when the machine is turned on, instead of being triggered to open when it gets a low current signal from the dosing button. I hope this works, because a TRIAC is $5 and a new logic board is $850.

Here's a follow up for anybody who has this problem in the future. I sourced a new triac from ebay (couple of bucks) and a $20 electronics soldering kit from Amazon and went to town on the circuit board. I got excellent advice from my EE brother to find a similar, discarded circuit board to practice on. That helped, as well as advice to snip the triac off and remove the little legs on at a time.

The most difficult part is determining if a triac has failed. There are 5 in a row on this board, all of which I tested with an ammeter. I wasn't getting the readings that I expected, but one triac was acting differently than the rest, so I replaced that one.

The whole operation took around an hour, and my soldering job isn't very pretty to look at, but it worked! The machine is back to operating normally and now I have a (properly wired) toggle switch which allows me to switch between adjustable preinfusion or instant pump engagement. All is right with the world.

Here is a link posted by someone who came before me with the same problem and included helpful photos. http://www.phoebeandcharles.com/2015/02 ... cilio-s27/