The Vibiemme DD version 3 with PID control on the brew boiler has been out for some months now, but there's nothing I can find on users experience with temperature control, hence this new topic. I've had my DD now for a bit over 2 weeks now, so I'm still working out the kinks. For a great introduction to the machine and a detailed first look check out http://www.espressomyespresso.com/. In a follow-up article I'll discuss my fun with adjusting the P, I, and D of the PID, but here I want to relate my experience with the temperature offset, also a feature of the PID.

The stock setting of the temperature offset is 19F, which means that the boiler will be 19 degrees Fahrenheit hotter than the water entering the basket. My wife and I have been drinking three different beans lately, Blue Bottle Roman that likes 184F, George Howell Matalapa that likes 195F, and Blue Bottle Retrofit that prefers 200-202F, so I've been all over the temperature spectrum. To measure temperature during the shot I drape a very fine gauge thermocouple over the 15-g basket and bend the wire it so that it is above the puck. I do notice that measuring this way produces a 2-5F droop during the shot. I attribute this to coffee swelling over the TC tip for two reasons: When I start with the tip under the coffee it reads lower than is usual, and when I measure the shot temperature without coffee but with a poor-man's Scace TC it starts at the same value and never droops more than 1F. I use these non-drooping values for the offset measurements quoted below.

Just a few words on the poor-man's Scace TC so you know how these measurements were made. Turns out that a hole in a blind back-flush basket 0.008" in diameter allows flow of about 2oz in 30 seconds with a pump pressure of 8-9 bar, just like a typical shot. So the same TC over the edge of this basket measures the temperature of the water in the basket.

So, on to the offset measurements. Unfortunately, the offset differs with set temperature and with whether the steam boiler is on or off. Surprisingly, the offset decreases (the water delivered to the basket is hotter!) with the steam boiler OFF. I don't have any hypotheses why this would be. All measurements were made after at least 45 min to stabilize and then after a 2-oz. flush. Here are the results:

Set Temperature(F), Delta-T(F) with Steam Boiler on/off
185 25/23
195 30/27
200 31/28

The fact that Delta-T increases with increasing boiler temperature is not surprising, since the heat loss from the exposed E61 group head increases. And knowing these variations it's easy enough to choose the set point to account for them. What I do find surprising, in addition to the behavior with the steam boiler, is that these offsets are all significantly larger than the factory value of 19F.

I hope these values will be useful to users of the VBM DD version 3. I look forward to measurements from others that can verify or improve on them. For the purist, room temperature was about 73F.

....................................
Gerry

While I have not got any numbers to share specifically, I can add some subjective observations. I am using Eric's group thermometer and when I can take my eyes off the show produced by the bottomless portafilter (its visual allure drawing your attention away from the rest of the world for around thirty seconds being its only possible negative aspect), my offset at 19 seems to be working just fine. The group thermometer shows about 203 and will actually go up about one or 1.5 F degrees if there is any change during the extraction. The DS would tend to drop by that amount or a bit more, FWIW. To me this slight rise in the DD would seem to point towards a fairly flat temperature profile throughout the extraction. It is no secret that I am working on the owners manual for the DD and so at this point I have made no other changes to the PID settings other than moving the setpoint up or down a degree or two as I learn the machine.

The offset that works in any given circumstance can vary depending on such things as ambient air temp in the area around the machine and even if an air conditioner blows in the area, or air currents which absorb more thermal energy from the group than in a still, quiet-air spot.

It will be interesting to hear from users who come up with PID settings that work differently. Right now I am happy with the "factory" settings and offset that, if I have heard correctly, were devised by Jim in his testing. The manual probably will not have much that will assist in this advanced area, aimed mainly at the new user. I just finished an illustrated explanation of how the internals an E-61 group work (I have spent three days on that alone). My wife read through it, and even she understands the detailed inner workings of the group now (I love her, but she is not at all mechanical)!

If anyone has numbers concerning length of cooling flush in relation to how long the machine has been idling I would appreciate it if you would send those to me. I assume that there is some variance and I would like some numbers to either average out, or at least compare to what I am finding. Such as - flush time after idle for 30 minutes, one hour, two hours, three hours would probably do it.

But once you pair this thing with a decent grinder, and if you have a good set of preparation parameters in hand, this thing is really quite amazing. This morning I made a couple from a batch of beans that were ejected too early (attention off the Hottop as I talked on the phone, finger on wrong button, it beeped the warning beep safety point, I pressed wrong button, ejected beans about ten seconds after end of first crack). Anyone for an unsweetened grapefruit cappuccino? A Sears Diehard Battery would be jealous of the acidity I had in my cup this morning! PERUVIAN PUCKER POWER! The machine took what the beans had to offer and put it in the cup. I think I wil drop the temp tomorrow a degree or two and see how that goes.

"KISS ME! I just had an espresso, and my lips are pre-puckered!"

Gerry wrote:The stock setting of the temperature offset is 19F, which means that the boiler will be 19 degrees Fahrenheit hotter than the water entering the basket.

No, the boiler will be regulated at 19°F higher than the displayed temperature. The temperature of the water entering the basket will vary significantly from this displayed temperature depending on boiler temperature, any flush regimen and environmental conditions.

Gerry wrote:Surprisingly, the offset decreases (the water delivered to the basket is hotter!) with the steam boiler OFF.

IMO, it makes sense. With the steam boiler ON, the interior is hotter and presumably the temperature differential between the top and bottom HX pipes is lower, resulting is a lower flow rate. A lower flow rate puts less heat into the group. A cooler group lowers the boiler water temp.

Randy G. wrote:The group thermometer shows about 203 and will actually go up about one or 1.5 F degrees if there is any change during the extraction. The DS would tend to drop by that amount or a bit more, FWIW. To me this slight rise in the DD would seem to point towards a fairly flat temperature profile throughout the extraction. It is no secret that I am working on the owners manual for the DD and so at this point I have made no other changes to the PID settings other than moving the setpoint up or down a degree or two as I learn the machine.

The shot profile varies with the boiler temperature, flush regimen, ambient conditions AND the position of the probe in the group.

Also IMO, based on 5 years of shot profile measurements, group offset temperature on exposed group machines is next to useless. A far more useful figure is obtained with EricS' adapter. On my HX and v2 machine, I can tape over the display temperature and adjust using EricS' adapter by taste. The boiler offset feature is pure marketing hooey.

CafeIKE, Thanks for the thought that convection through the group head slows down when the steam boiler raises the ambient temp. That makes a lot of sense and solves my mystery.

Your first point is, of course, correct. Still, my goal is to make my statement correct, that is, making the displayed temperature be equal to, or offset by a known amount from, the water temperature going into the basket. I think it's important to know the temperature at the top of the basket because I believe this is the temperature people have in mind when they say things like

Pulled on days 6-10 @ about 198°F to about 80% from a serious updose, this blend is stunning.

If a temperature other than basket temperature is meant by this statement, please let me know. I also measure the group temperature with something similar to EricS's TC, but I find the difference between group and basket temp to vary depending on flow rate, probe position, and boiler temp. So quoting group temperature is not very helpful, except perhaps for reporting taste changes due to temperature changes rather than absolute temperatures.

There seems to be a spectrum of opinions in these pages between those who want to measure everything (extracted solids, for example) and those who would rather do everything by taste. Either extreme is fine for getting the best shot for the individual doing the pulling, but IMHO to communicate what works, and what doesn't, requires some measurement. Sure, it's essential to optimize the temp by taste, but finding the place to start from helpful suggestions like the quote above can shorten the process considerably for someone unfamiliar with the beans.

Randy G, thanks for your feedback too. I'm looking forward to your more extensive report and manual.

Gerry wrote:If a temperature other than basket temperature is meant by this statement, please let me know.

I try to use the end shot temperature as measured with EricS' adapter. It's the most repeatable. When I insert the probe, I remove the shower screen and the water spreader, insert a drill to block the passage and position the probe so it just intrudes into the angle in the e61 water path. Only by doing this are measurements repeatable and transferable between my machines.

As example of inutility the boiler offset temperature reading : Today I changed from a serious updose lungo to normal dose ristretto. Boiler temperature +4°F, shot temperature no change.

Thanks for the note. Do you have any evidence that the offset between your group and basket temperatures is independent of flow rate? I haven't looked at that with the VBM, but on the Tea (both E61) I noticed a dependence.

As promised, here's a very brief summary of my journey into VBM PID land. I noticed that the stock setting of the PID caused an oscillation between 198 and 201 F, when set at 200 F, with a 2.5-minute period. Using the Wikipedia tuning process http://en.wikipedia.org/wiki/PID_controller and the Ziegler-Nichols method outlined there, this suggests that the "P" was set about right, the integration constant "I" should be 0.01 instead of zero, and "D" should be 15 instead of 0.5. I have been using these PID values for a week now, and notice that without disturbance the temperature stays within 1F of the set point, and overshoot is somewhat less than with the stock values. That said, it's not worth the bother changing them unless you are particularly compulsive or just like to play because the minor improvements are doubtless invisible when it comes to temperature stability of the shot. This is dominated by the large mass of the group head, as has been often pointed out. The bottom line is that whoever determined the original PID settings for this machine did very well indeed.