PID Offset Problem

[Marshall]

I should know better than to touch anything technical on my machine, but fools do rush in. A friend was kind enough lend me a Scace Thermofilter and Extech multimeter to calibrate my Zaffiro. Interestingly, it mostly confirmed the results from my probe-over-the-lip and cheapie Chinese multimeter. After an 8-9 second rise, brew temperatures were flat across the board for all boiler temps and consistent with other shots at the same boiler temperature.

With my PID probe mounted inside the boiler, the heat loss to brew temperature was about 22F. So, I happily packed up the borrowed equipment and shipped it back to my friend.

Then I started to think about simplifying things by changing my PID readout from C to F and setting a calibration offset (of -22F, of course). This is where things got sticky. The resulting brew was thin, pale and tepid. I thought, perhaps, I should have set a positive 22F offset. That produced highly entertaining waterworks, but a rather burnt cup of coffee.

I found Jim Gallt's very clear posting from last summer and confirmed that I was correct in setting a negative offset, but it really didn't solve my lukewarm espresso problem. Jim mentions a built-in offset that has to be added to the desired offset. But my Eurotherm PID instructions only refer to choosing a "factory calibration" or "user calibration," without saying how much the "factory calibration" is.

Any suggestions? If all else fails, I'll remove the offset and resume doing the subtractions in my head. But, would be kind of cool to see the actual brew temperature on the PID readout.

FOLLOW UP AN HOUR LATER: I set the Offset parameter back to "factory calibration," and the readout immediately showed that my boiler temp was 179F (22F BELOW the set point, instead of above it), which explained the tepid espresso. I just set it all back to "factory calibration," with the indicated set point at 223, and all is right with the world again. Obviously, there is more to learn about this PID offset business.

[jesawdy]

Marshall, IIRC you have the Eurotherm 2132? I have some manuals floating around if you do not have them.

[Marshall]

Marshall, IIRC you have the Eurotherm 2132? I have some manuals floating around if you do not have them.

Yes, Jeff, I have the 2132 and the manual. Thanks.

[HAL9000]

Before trying anything else I would try reversing whatever you did to change the readout from C to F and then try the offset again in C. I have a controller that requires me to change the temp unit in a number of different places in the configuration for it to work, and wonder if yours does too. If nothing else this will remove that as a potential cause of the problem.

Does your controller have the ability to control two separate processes?

[Marshall]

Before trying anything else I would try reversing whatever you did to change the readout from C to F and then try the offset again in C. I have a controller that requires me to change the temp unit in a number of different places in the configuration for it to work, and wonder if yours does too. If nothing else this will remove that as a potential cause of the problem.

Does your controller have the ability to control two separate processes?

The Eurotherm has about 60 parameters that are adjusted through a couple dozen submenus nested in 9 top-level menus, all controlled by 4 buttons on a 4-digit readout. I rate my chances of doing more good than harm as slim-to-none, even with the manual firmly in hand.

[HAL9000]

The Eurotherm has about 60 parameters that are adjusted through a couple dozen submenus nested in 9 top-level menus, all controlled by 4 buttons on a 4-digit readout. I rate my chances of doing more good than harm as slim-to-none, even with the manual firmly in hand.

Being a lawyer myself I know the feeling well. Though I am fortunate to have an EE for a brother and he has helped me through a lot of this stuff.

Just another thought:

I'm not sure the built-in offset or factory calibration is relevant for you. I am no expert (as I mentioned I am something else entirely), but I think you are looking for an offset from the displayed temp, which already takes that into account. I would be surprised if the offset programming refers to anything other than the factory calibration - which presumably could only be some unknown absolute measurement that is the basis for the undocumented factory calibration.

That said, I found your instruction manual and notice that there are two different offset adjustments that can be made - one is to apply the same offset to every reading along the entire range, and the other is to pick two boundaries of a range (e.g., freezing and boiling), and then set them separately so that, presumably, the controller will calculate the appropriate offset for any temperature within that range. Which type of offset were you trying to program?

Best,
Paul

[Marshall]

That said, I found your instruction manual and notice that there are two different offset adjustments that can be made - one is to apply the same offset to every reading along the entire range, and the other is to pick two boundaries of a range (e.g., freezing and boiling), and then set them separately so that, presumably, the controller will calculate the appropriate offset for any temperature within that range. Which type of offset were you trying to program?

Best,
Paul

Yes, I noticed that, too, Paul. Apparently I'm on the "PID for Dummies," setting, which only displays one offset parameter. I can see that there are ways to "promote" hidden parameters to visibility, but only wanted the single offset, anyway, at least until I find a way to use the Zaffiro to make gelatos.

[Marshall]

Before trying anything else I would try reversing whatever you did to change the readout from C to F and then try the offset again in C. I have a controller that requires me to change the temp unit in a number of different places in the configuration for it to work, and wonder if yours does too. If nothing else this will remove that as a potential cause of the problem.

But, as I think more about it, I think you're on to something. My experience only makes sense if the PID was applying a 22C offset in both directions. But, the chart on Page 5 of the Manual says the offset will adjust in X "display units," and my display is set for F. So, I'm pretty much at a loss at this point.

[JB130]

I'm sure someone who knows this controller will chime in eventually, but until then I will add my 2 cents ... Here are what seem to be the relevant excerpts from the manual:



It sounds like this is what you want, to set the OFS parameter to -22F. And it sounds like you tried to do exactly that.

I don't think you need to mess with the 2-point calibration discussion that follows:



At first I thought it might be that the offset is in units of C, even when the process value is displayed in F, but as you say, it seems pretty clear from the manual that the offset units are the same as the display units:



Another possibility is that you somehow got yourself into the 2-point calibration menu and set OFS.L = -22F and OFS.H = 0F, leaving OFS = 0. This combination would lead to your symptoms. But based on your "PID for dummies" comment it sounds like you were not in the 2-point menu. So I'm not sure what else might be wrong.

John

[gabriel]

Marshall,

I was testing my Silvia with LOVE PID using a Scace Thermofilter and found that at least with my config the offset was not static

when I set 210F in the boiler I get ~195F brew temp, but when I wanted to get 200F brew temp I had to set the PID to 218F and not 215F
All testing were done after the machine was on for a full hour

Not sure if any of this apply to the Zaffiro which IIRC has a dynamic brew head heating like the Alexia and not the static heating you find in the Silvia

/gabi

[Marshall]

John,
Thanks for all the layout work. Yes, that is the same manual I have been looking at and, yes, you summarized why I am so perplexed.

Gabi,
I also thought it might have been the ability to set two different low/high offsets, except that my PID is currently configured to only make the "low" temp offset visible. I suppose I should look into what is going on with the "high" offset, but am concerned about what I might screw up by accessing it.

[JB130]

I also thought it might have been the ability to set two different low/high offsets, except that my PID is currently configured to only make the "low" temp offset visible. I suppose I should look into what is going on with the "high" offset, but am concerned about what I might screw up by accessing it.

Marshall

Now I'm confused; when you say the "low" offset, do you mean OFS.L?

John

[Marshall]

Marshall

Now I'm confused; when you say the "low" offset, do you mean OFS.L?

John

Yes.

[JB130]

Yes.

Then maybe the problem is as I described yesterday?

Another possibility is that you somehow got yourself into the 2-point calibration menu and set OFS.L = -22F and OFS.H = 0F, leaving OFS = 0. This combination would lead to your symptoms.

In other words: If you only set the "low" offset to -22F, and leave the "high" one at zero, then you will only be offsetting boiler temperatures near the temperature specified by the Pnt.L parameter (defaults to 0), but you would not be offsetting boiler temps near the temp specified by the Pnt.H parameter (default is 100). Since your boiler is nowhere near freezing, you are not getting the offset you expect. This combination of parameters would lead exactly to the symptoms you describe.

One fix would be to set both OFS.L and OFS.H to the same value (namely -22).

But the easier fix would be to set both OFS.L and OFS.H to zero (their default values), and to use the OFS parameter instead. According to the manual, you shouldn't have to use the 2-point calibration if you want the same offset for all temperatures, the single OFS setting will do the trick.

John

[cannonfodder]

And this is why I still use a commercial pressurestat.

[cafeIKE]

And this is why I still use a commercial pressurestat.

clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK

[HAL9000]

clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK

LOL, literally. I described that once somewhere as the sound of a blacksmith shoeing horses in the next room, slowly and deliberately. IMVHO the PID is worth the trouble and expense just to get rid of that racket. At least for home dinner party use.

[cannonfodder]

That is just the sound of love on standby.

It never bothered me but I have heat exchanger machine, Marshall has a single boiler Zaffiro which would benefit greatly from a PID control. Sorry, I just don't see the need for one on a HX box. Single and double boiler I do.

[HB]

clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK
clickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACKclickCLACK

I've never given it much notice over the "normal" household buzz. Maybe in another 20 years I'll have the luxury of worrying about sounds less than 50 dB.

[cafeIKE]

Sorry, I just don't see the need for one on a HX box.

Not much of what we have, do we "need."
Consider for a moment the energy required to heat AND the "gallons" used flushing a HX that is far too hot.
Consider that a PID will likely out last a half-dozen pressurestats that otherwise end up in a landfill.
We must all do our part.

EDIT : Just this morning I started a new blend. Underwhelmed on the premier, a 2°F drop in boiler temp rectified the defect.
Much simpler than fannying about flushing an extra ?.?.?ml and remembering to do it as long as the blend persists. If I return to the previous blend, and assuming the seasons haven't changed, +2°F and I'm exactly where I left off. What could be simpler?

I've never given it much notice over the "normal" household buzz. Maybe in another 20 years I'll have the luxury of worrying about sounds less than 50 dB.

Some machines have quiet tiny single contact pressurestats, some have full size triple contact commercial models. The large 2" diaphragm in a commercial model makes a snap with a peak level considerably louder than 50db. When guests at a dinner party in the next room ask "What's that clicking?", it's too loud.