The Oscar has been PID’d

It took much too much time. That’s inexperience for you. Had to learn a little about plumbing, wiring, power, and now I’m at the last hurdle of learning the machines temperamental nature.



Working through calibration has been just about as interesting as the rest of the project. I’ve been talking about this on Home-Barista a bit lately and I’m sure I’ll run the gammut of the alt.coffee group. For now though it’s been a self-study using Ken Fox’s article on “Should I PID my HX”, which he goes through similar motions to find out where the sweet spots of an HX/PID would be. He set the bar pretty high with the number of series he ran and the reporting he gave.

Below was my first thoughts. The graphs were accumulated under the following rules:
* 30m sitting
* 50ml flush
* 2m wait
* Pull for 22s
* Wait 2.7m
* Rinse Repeat


As you can see the machine doesn’t really harmonize that well under these settings. A 2 minute down time really wasn’t allowing the machine enough room to become stable. At this rate there could be up to a 7’f shift in temperature from one shot to the next. I then decided to try another series that went for the longer view.
* No flush
* 10m standing minimum
* 50m standing maximum


This showed a remarkable improvement. The data suggests that somewhere around the 10m mark the thermal zone is achieved (looking for 202’f +/-ish). I should have documented the times, but the longer towards 50m the higher the initial spike in temperature. A little flush for anything over 50m would probably help keeping it out of any burning temps.

My last series I ran was a simulated night with company that all liked milk.
* Grind Time
* PF Unlock
* Grind/Tamp/Clean
* PF Lock
* Extraction 22-24s
* Steam/Froth Milk
* Clean & Reset
* Repeat


Based on earlier data we know shot performance is weak on a 2m delay. It is no doubt that we would see a degrade in performance. The first person gets the best shot, everyone else is off markedly. (Yes we could get into the milk issue and need for best extraction but thats a digression from the general test) You do however start to see a trend that the machine is stableizing at a lower temperature. One thought I have kicking around is that I can raise the temp up quite a bit flush two shots (60-80ml) and perhaps it would even out at a desired temperature. However this may be faulty logic. I’ll test and see sometime.

I want to do another series at 4m and 7m to see what it looks like. I’m thinking somewhere between those times I’ll achieve a repeatable shot zone that the machine can handle for general use/walkup performance. Once that zone is understood finding as Ken would say a high and a low zone for other coffees temps would be good to know. Then trying to solve high performance out of consumer grade vehicle would be on the agenda. Boy oh boy is this a time consuming hobby!

A big big thank you to Jack Denver & Greg Lepore for holding my hand through so much of this process. To the outside world reading through a conversion sounds very simple, but in practice it has a steep learning curve depending on your natural areas of expertise. Without good people like these two and a great body of communities surrounding me this project would have been near impossible.

The culmination of this ditty was to show a picture of my first PID extraction but of course now my Canon has decided to malfunction after three years of perfect service. As soon as it is repaired we’ll do more show and tell. Least to say after a few misdials and not enough Tuscano, my last shot was at perfect time and had such great taste I know my efforts have been rewarded.

To the perfect cup! ChinChin!
-a