Comments about PID settings and temperature on the Pro 700 can be found in several threads but I wanted to regroup my recent findings here.

***

A recap of previous episodes...

After struggling with over-extraction for weeks with the Pro 700 I realized the water temperature with the default PID settings was way too hot in my case. I first had doubt because I was hearing what sounded like (and probably was) flash-boiling when activating the brew lever, similar to a cooling shot on an HX.

I purchased a cheap digital meat thermometer to try to measure water temperature right out of the group but the thermometer is too slow to get accurate measurements. Still, with the default PID settings I could get readings of 205F out of the group after several seconds of running water --- during a shot temperatures were likely even higher.

I had to decrease the offset parameter (E1) by at least 10F to get a reading of 200F out of the group, but even then the reading was after seconds of running water and likely under-estimated brew temperature by a large margin.

In the meantime I have identified a discrepancy in the PID settings. When set to degrees Celsius, the default offset (parameter E1) is 14C, but when set to degrees Fahrenheit, the default offset is 34F. These do not agree -- a temperature difference of 14C is 25F while a difference of 34F is about 19C. So just in the default PID setting the offset is 5C higher when set to Fahrenheit. I have noticed Profitec of this discrepancy but I have not yet got a reply.

***

Following the advice of many HB members, I decided to set the offset to 0F so as to directly assess the brew boiler temperature. I then reduced the temperature until I could taste a pleasant acidity in the shot.

I used fresh beans (about 6-7 days post roast) of Pilot's Big Bro blend (I think this qualifies as a light roast --- Pilot mentions an Agtron reading of 70, although I've seen lighter roasts). Pilot recommends a brew temperature of 199F for this blend. Right now I set my brew boiler temperature (with zero offset) to 206-208F to get a great balance between acidity and bitterness. Crema is more uniform and not too dark now, latte art is easier because crema is less stiff. Shots are *much* more enjoyable.

I cannot take a reliable reading of the group temperature but I will try to use an IR thermometer as soon as I can.

Using the default PID settings in Fahrenheit, the brew boiler was at (t1+E1) = 200 + 34 = 234F, so the temperature difference between the default brew boiler setting and what I find produce palatable shots is pretty large (28F/15-16C).

It would be interesting if other owners of the Pro 700 would comment about their temperature settings. I wonder if it's just my unit (temperature probe problem?) because different reviews mentioned awesome shots out of the box. I know Dan's review used the machine with the PID set to degrees Celsius, which likely alleviated part of the problem. And clearly the machine could stand its own compared to others (e.g. Vesuvius shoot-out) so experimented tasters didn't seem to notice any major problems with the Pro 700.

I also struggled with temperature adjustment on the Profitec Pro 700, which seemed to be running too hot. I was measuring 206°F from the shower screen. After a month, I purchased the Scace I with a Fluke 52-2 thermometer. To my surprise, the grouphead temperature matches the PID reading of 200°F with the offset at the default setting of 34. That said, the shots are not very good. My 10-year-old HX machine produces much more enjoyable espresso.

Did you take the scace readings with the fluke set on T or K couple?

The Fluke is set for "T"

T is correct then.

GC1 wrote:I also struggled with temperature adjustment on the Profitec Pro 700, which seemed to be running too hot. I was measuring 206°F from the shower screen. After a month, I purchased the Scace I with a Fluke 52-2 thermometer. To my surprise, the grouphead temperature matches the PID reading of 200°F with the offset at the default setting of 34.

So if you measured 200F with the Scace+Fluke, how did you measure 206F at the shower screen before? Digital thermometer as I did?

Also you're speaking of the Pro 700 in the past tense. Did you get rid of it? Or aren't you using it anymore?

That said, the shots are not very good. My 10-year-old HX machine produces much more enjoyable espresso.

I don't have much experience with espresso so take everything I say with a grain of salt. But lowering the temperature on my unit drastically improved the taste. I have no idea how far this is from the optimal taste that could be extracted from the coffee I'm used (and I understand this is also a question of personal taste) but correct 200F or not, you ought to try it.

I'm not saying your Scace measurements are wrong (I have no idea how to even make those) but as far as I can tell, either the settings in degrees F or those in degrees C (or none of them) are right. They can't be both right as they set different temperatures on the brew boiler side.

Furthermore I am pretty sure that with the default settings (34F offset, 200F target) I can hear flash-boiling in the group; and that digital thermometer readings are almost certainly low-ball for the reasons I mention above.

I have yet to hear from Profitec, and for now I'm pretty happy with the lower temperatures. My next move will be simply setting everything to Celsius. On one hand we're losing precision (1F = 5/9 C) on the display but on the other hand the PID controller was programmed in Celsius. The recommended settings for the proportional, integral and derivative gains are the same in both units but I wonder if they should be. After all they apply to different readings unless the units are only there for display purposes and all computations are performed in C. In any case, early reviewers didn't note anything wrong with the taste (on the contrary) and the machine was set to C.

You live in Canada, you should be using Celsius anyways .

Indeed but it took me some time to even acknowledge the existence of the PID controls as I wanted to concentrate on my technique first. Turns out it would have helped me tremendously right from the start.

Anyway, a word of caution if you are changing the F.01 setting (temperature units). The PID controller performs very suspicious operations and conversions trying to retain the settings across unit changes. I just set my machine to Celsius; I had 255F as for my T2 setting, and 0F as for my E2 setting (the reading offset for the steam boiler). After changing to Celsius T2 got changed to 124C (correct) but somehow E2 got changed to 45C.

This is in fact pretty dangerous as now the target temperature in the steam boiler was now 124 + 45 = 169C (334F). I noticed the boiler being active and saw the steam pressure go up and up; I stopped the machine before it reached 2 bars. The lesson is: unit conversion is screwy in this PID's firmware (v. 1.04 btw in case further revisions occur), and always review all settings before committing.

A couple of things to add in this thread...

First, I got a reply from Profitec and the issues I raised were supposedly sent to the development team.

Second, I wanted to say that using E1 = 0 and E2 = 0 (no offset), using a 95 C brew boiler temperature (t1) currently does a good job (as far as I can tell) with medium roasts. This is significantly lower than the default setting even in degrees Celsius.

Third, I just revised my system control theory basics and it looks to me like if you want a certain behavior out of a control loop, and you set the P, I, D parameters to get this behavior, then if you change the units of measurements of the output variable of the process and of the command, then you must also change the P, I, D parameters of the controller to if the same behavior is to be expected. As it stands, Profitec recommends and use by default the same parameters in degrees Celsius and Fahrenheit. This likely results in different control loop behaviors depending on the chosen units.

While I'm no expert, it appears to me that the PID is set pretty aggressively, with a rather large overshoot (aiming for 95 C, the controller typically overshoots up to 97 C before going down to 94 C, then 96 C and stabilizing to 95 C). While I have not explored different PID settings (yet) there should be different values of the P, I, D parameters which provide a smoother ramp up in temperature at a modest recovery time cost.