Has anyone asked WLL to check into it if Profitec isn't responding to inquiries?

Carsncoffee wrote:OK so I am confused-
Did you find that the stock settings of the Profi 700 are off significantly or not?

Not sure if you were asking Eric or myself. As for thermal image measurements, there are a few caveats: measurements could be off for a couple of reasons.

First reason would be calibration of the unit, but it's a self-calibrating camera and should be rather accurate and precise.

Second reason is that to determine temperature using IR radiation, one has to know the emissivity of the material you're looking at. Using the wrong emissivity can make a material appear of a totally different temperature, such as the stainless steel parts of the machine in the pictures above. I tried two different values for the emissivity of water (0.95 and 0.98) and they gave results about 0.5-1.0 C apart depending on the temperature measured. So I'd guess a 1C error would be the maximum error due to incorrect emissivity for water.

Third reason would be that water is opaque to infrared radiation, so I was only measuring a very thin layer of water that is exposed to air. It could be that this thin layer has a significantly different temperature than water deeper into the stream. Also, steam around the stream could partially hide the stream from the camera, but if so, it's not really visible on the image.

I adjusted the offset on the machine so the hottest part of the image while looking at water streaming out of the group was at the desired temperature (93 C). This gave an offset of 12-13 C, which is in my case slightly lower than the default offset in Celsius on the machine (15 C if I remember correctly). 2-3 C off the default is not huge and could be due to ambient temperature and local peculiarities (e.g. it's installed in a corner and under cabinets so it's not in a well-ventilated area and thus the machine could be keeping its heat better).

This means 6-7 C off from the default offset in degrees Fahrenheit though, which is a bit high, but it's not 15-20 C off as I feared.

That being said I found that even using water temperatures below 90 C, the results were not that offensive. I haven't lowered the temperature enough to taste frank sourness in my shots.

I did a very quick and probably not very accurate measurement with a Thermapen cooking thermometer (supposed to measure temp after 2 seconds):
with PID saying grouphead is at 202 F, water coming down from head was measured at 198 F.

The problem with digital cooking thermometers is that they take a while to stabilize (or they need to be covered in water). The best way to do this is to let the water fall into a thermally isolated cup (e.g. styrofoam) that is allowed to overflow, and put the probe in the cup. The other problem is that when just letting water flow into the cup, a large amount of water is drawn from the brew boiler compared to when pulling a shot. Consequently a large quantity of fresh water is drawn into the boiler from the reservoir or line, and the water temperature drops much faster than during a shot.

Eric's thermometer is probably the best way to take the measurement short of a Scace as the measurement offset is smaller and probably less variable than the boiler offset. If I had a Scace or Eric's thermometer I could correlate those measurements with the IR ones but it's not really in my plans for now.

Unfortunately, I too, have been trying to sort out why my espresso tastes bitter/over extracted.

As I said above, it's not illegal to use colder water and it seems to help a bit.

One thing I noticed is that pulling shorter shots (on the ristretto side) helps quite a bit. I suspect that the ultimate cause of my problems is uneven extraction when going with a coarser grind. Timing is right and I get a mix of over-extracted and under-extracted coffee in the cup. With the ristrettos I think the extraction is more even. As soon as I get a bit of channeling the shots taste significantly worse.

thanks for the replys neutro and erics-
I will try to just lower the brew temp by a couple degrees at a time to see if this helps.

I tested my Profitec Pro 700 last night with a Scace 2 device.

With the factory setting, the temperature at the brew head was 220 F. I adjusted it down to the lower limit of the adjustment range, and achieved 206 F.

Did you attempt to calibrate the thermofilter and the meter? What meter are you using?

Based on what you have said, something seems really amiss.

I should have written more accurately. It was not I who took the measurements. It was someone who has years of experience using Scace devices and has calibrated machines for many barista competitions.

The only conclusion that I can come to is that my machine did not undergo any kind of factory test. I would also speculate that the thermocouple ECM is using is not matched to the display.

And the reason your numbers are so far off,I BELIEVE, is that you have the F and C scales mixed up.

If you want the machine to operate in Celsius mode, all the settings must be in C. Likewise for Fahrenheit mode. See this and compare your settings to it: /downloads/ ... ttings.pdf and/or /downloads/ ... elsius.pdf

The Profitec and most other machines fitted with a GICAR controller, utilize a 50K ohm negative temperature coefficient (NTC) thermistor to measure temps.

Eric and I are having a conversation about the meaning of the t1 and E1 parameters.

I think t1 is the setpoint of the brew boiler. The PID controller will drive the error between t1 and the temperature measurement in the boiler to zero.

The reason I think t1 is the setpoint of the brew boiler is the description in the manufacturer's documentation that says t1 is the "desired value in the brew boiler - desired temperature value at the measuring point".

I am interpreting E1 as a parameter which sets the temperature displayed to the user when the unit is in normal operating mode, to indicate the assumed temperature of the water in the group head. The displayed temperature, in my theory, is (t_m-E1) where E1 is the assumed temperature difference between the boiler and the group head and t_m is the measured temperature in the boiler. (t_m=t1 in the steady state.)

There is an alternative view, which is that the setpoint of the boiler is actually (t1+E1). So t1 is the desired temperature at the brew head, and the PID controller controls to (t1+E1) since the boiler has to be hotter than the desired temperature at the group head. In this view, I think one is assuming there are translation issues in the manufacturer's documentation.

At least one of these views is wrong. Can anyone shed any light?

I am pretty sure the offset E1 is added to the user-input temperature T1 to get the brew boiler temperature T1 + E1.

I agree the wording in the parameter charts is confusing (or perhaps plain wrong). But if you set E1 to 0 intially, and then change it, you can actually hear the boiler go into action. (Especially in the case of T2 and E2 -- the steam boiler is easier to hear). If T1 was the brew boiler temperature, then changing E1 would only change the display and wouldn't start the boiler.

Plus, the default for T1 (199F or 93C) is typical for the group, not the boiler.

I may be wrong here, but there's an obvious way to be sure: just leave E1 to 0.

Note that there's also a way to factory-reset the PID which is nowhere in the instructions. If I remember correctly, you have to hold the right (up arrow) button while turning the machine on. I wonder if the controller can be put in a weird state after many modifications to the parameters.

I hear you.

When they define E1, they give a definition that is consistent with my interpretation: "E1
Temperature difference measuring point to the display
Example: 10°C: measuring point 100°C => display 90°C"

i.e. when the temperature sensor in the boiler says 100, the display says 90 to compensate for the path loss.

Note that this has nothing to do with the setpoint.

It would be really nice to get some clarification from the manufacturer. I sent them a message.