Eric S' E61 grouphead thermometer - calibration

Need help with equipment usage or want to share your latest discovery?
mad1
Supporter ♡

Postby mad1 » Jul 03, 2019, 7:21 pm

Hi

I have Eric S' grouphead thermometer installed in my Quickmill Alexia with PID. When I run water for about 5+ seconds, the temperature readout of the grouphead thermometer stabilizes at 201F. However, if I have a puck in, the temperature stabilizes lower, about 197-198F according to the thermometer. Let's assume that I'm going for a puck temperature of 197F. I would want the thermometer to read 200-202F through the shot (given a 3-5 degree loss from thermometer to puck).

So, my question is - is the grouphead thermometer thrown off by the reduced flow rate, or is it that the water is actually colder. If that's the case, calibrating without a puck is pointless, correct?

Thanks,
Madhu

Nunas

Postby Nunas » Jul 03, 2019, 11:38 pm

Short answer, the water has to be colder. When the machine is siphoning, water flows from the boiler via the top pipe, cools in the group, then returns via the lower pipe. When you are pulling a shot, water is drawn from both pipes, hotter water at the top and cooler at the bottom.

Long answer, I suggest that 'calibrating' is really not the right verb here, but I think I understand what you mean. The thermometer itself is dead nuts on, as delivered by Eric. You can test it by putting the probe into rapidly boiling water. It will read 100 C (or at least mine does). So, the issue has to do with compensation for the difference between measuring the water in the puck and the water on the way to the puck, where the thermometer is located, not calibration. I suspect this varies from machine to machine. In the case of my HX, the flush water comes out at less than 100 C at the onset, then quickly rises to exactly 100. As the flush progresses, the temperature keeps dropping, slowly at first,then more rapidly. I generally stop at 96, which I've measured gives me about 93 in the puck. Also, I've noticed that at this temperature, the reading stays fairly constant, only dropping very slowly, as the shot is pulled. To achieve all this, I had to practice a lot until I developed a sense of what my machine was doing. It takes my fully warmed up machine, an unrestricted dragon, about 20-seconds to do the flush.

On the other hand some machines either have a different tube geometry, or a restrictor in the thermosyphon. Users of these machines (and some identical to mine which have had added a restrictor) report much less need for flushing. They also report totally different dynamics in the movement of the temperature. You can do a search here on H-B "restrictors' and you'll see what I mean.

I have not tried this, but I suspect that even between identical machines a little thing like changing the boiler pressure by a few tenths of a bar would make a difference. I consider Eric's little thermometer an indispensable accessory for any HX machine. But, you have to use it every time, you have to do at least some rudimentary testing of puck temperature, and you have to get to know your machine, for it to be truly useful.

I know I've not answered your question...I'm not sure one can answer it as there are so many variables.

Bluenoser
Supporter ♡

Postby Bluenoser » Jul 04, 2019, 9:55 am

As an HX owner with Erics thermometer I also have struggled to determine the brew water temp compared to what I read on the thermometer. There is no easy answer to this and if you search HB you will see many discussions on this. If you do a flush, there is a "flush n go" and a "flush n wait" technique. The thermometer is dead on in the water at that point of the head, but the E61 mass is the most important influencer of how that water temp gets modified. I have tried to use external thermometers and thermocouples with very limited success. I think the only true way to measure brew water temp is to use a SCACE. Too expensive to buy, but if you can rent one, you will learn a lot about the thermodynamic properties of your machine.

If you notice a too cool in the end temp of the shot (say last 5 seconds) when you put in the portafilter then your portafilter may be cold. It is important to put this into the group as it is heating. A cold nose (cold portafilter) will cool the group head, cooling the brew water as it reaches the puck. A flush through the empty filter before the shot can help heating up the filter (and the group head)

In talking to some owners of Profitec HXs, we find that Eric's thermometer will increase up to 5 degrees on as the shot pulls and will then decrease after about 5 seconds and remain relatively constant. I am assuming this final point is close to the brew water temp the puck will see.

I am not very knowledgeable on HX flush techniques and there are many here who can add much more accurate info.

One huge benefit I find with Eric's thermometer is to determine when the water in the TS is hot enough to pull the next shot. In my PID HX, the restrictor is poorly designed if you want quick, back-to-back shots. After 2 shots, my TS is too cool to pull that 3rd shot, but the thermometer will let me know when the water is back up to temp (it can take between 10-15 minutes). Each manufacturer's design is different; and I suspect a restriction with a very tiny opening might mean that even within identical models, manufacturing tolerances might cause slightly different thermal characteristics.

mad1
Supporter ♡

Postby mad1 » Jul 13, 2019, 5:48 pm

Thanks for the replies. I guess I'll have to live with some inaccuracy in actual grouphead/puck temperature, unless I can get my hands on a Scace device. However, I'll try to do my best to get a consistent grouphead temperature, and the rest of the post is an attempt at doing that.

To answer some questions above - I always use a bottomless PF and keep it locked into the group. I dose into the basket and only remove the PF to put the basket in. I rarely do multiple consecutive shots - I pull maybe a shot every 2 hours, with a cup warming flush before the shot. No milk steaming.

I guess my main concern is that the delta between the PID readout (boiler temperature) and the thermometer is not consistent. I made two videos - one with no PF, and the other with a puck in.

Both shots had a PID setpoint of 216F.

No PF:
With no PF, the grouphead temperature stabilized at about 203. It does start to go down at the end - I imagine that's the boiler running out of hot water - you can see the drop in PID readout a few seconds before that. I cut the shot at that point.


Pulling a real shot
The second shot doesn't really stabilize - the grouphead temperature bottoms at 197 before slowly rising to 200.


Here is a second-by-second graph for both shots.
Image

Based on the above, my working theory is that a lower flow rate when a puck is in leads to a higher heat loss between the boiler and the group, ergo, lower grouphead temps. This is somewhat corroborated by the slight increase in grouphead temperature when the flow rate increases towards the end of the real shot. It's strongly corroborated by the fact that a blank shot shows a 5F higher grouphead temperature.

So, if I want a grouphead temperature of 200F during an actual shot, I must adjust the PID higher. Basically, what I'm saying is that calibrating grouphead temperature without a puck in is pointless.

Thoughts?

User avatar
GC7

Postby GC7 » Jul 13, 2019, 6:17 pm

Hitech should have calibrated your machine when they installed your PID. You should have been told the exact offset from the boiler set temperature to the puck. The $350 you spent for that was more than expensive enough for what looks like their standard delta PID without shot timer or other new features of more modern espresso temperature controllers.

RyanJE

Postby RyanJE » Jul 13, 2019, 7:05 pm

I had an Alexia with the group thermo as well as a Scace... your group should idle after a 50 minute warmup with the PF in at 198 ish... the group thermo reads about 3-5 high about half way into a 25-30 sec shot..

For me, my machine needed a 24-25f offset rather than the stock 18f it came set with.
I drink two shots before I drink two shots, then I drink two more....

maximatica

Postby maximatica » Jul 13, 2019, 9:57 pm

Unfortunately, due to time, I can't address each issue in the prior replies.

But I can say, that I had an Isomac Millenium and put a PID on the boiler with 2 thermocouples and then put Eric's thermometer in the group head and then had a Scace in the portafilter.

The temp at the Scace was 10F lower than the readout at the thermometer.

The boiler was insulated and I had some sound damping foam in there to try and quiet down the pump. I wrapped the copper pipes in insulating metallic tape to try and keep the cooling from the exposed copper tubing to a minimum.

But over time I had to develop a bit of a gain ride technique as I had the side of the machine off due to always futzing with it.

I had a switch that changed from the PID to the pressurestat to add heat to the boiler on the fly for steaming and then I would wait and use that heated water to help maintain the temp through the pour.

But I eventually had had enough with the futzing and got a BDB and I am much happier.

But based on my experiences, a 10F drop from the thermometer is what you would find if you were going to the trouble to try and get an exact temp in the puck (area).

HTH.

FWIW, if someone wants to make their own Scace type measuring device and has an E61 group head. I have an E61 portafilter with a hole drilled in it, a Fluke immersion probe and a Fluke 52 digital thermometer that needs a new home. I was going to get some high temp RTV to use to absorb temp like the Scace device does and have my own DIY Scace but just went ahead and got a Scace instead.

tracer bullet

Postby tracer bullet » Jul 14, 2019, 11:01 am

I had to set my Alexia Evo offset to 28 deg F to get where I wanted to be. 30 minutes is a minimum warmup and almost stable, and by 45 minutes it's there. I have some Excel data but haven't ever posted it, suppose I should someday.

Anyhow, as you know the thermometer is just measuring the water that goes past it plus some residual from sitting in the group head. You already know the water is very hot in the boiler, and cools before it gets to the thermometer. It will continue to cool a bit more before it hits the actual puck, and then within the puck itself. Even that'll change over the duration of a shot.

Other threads here in the past showed a group head temp approximately 5-6 degrees over the puck temp and so that's what I've shot for myself as well. When I've tried a few ways to measure puck temp (portafilter with a dense sponge stuffed in it, a thermocouple in that, and a small needle valve to act as a flow restrictor to allow water out of the portafilter around the same rate as espresso does) - I do indeed get around that 200F mark. I say "around" because every test was a little different but in the end it was in the neighborhood.

After that the idea is to adjust based on taste and to be honest I'm still learning to do that. But at least (I think) it's a good starting point.

mad1
Supporter ♡

Postby mad1 » Jul 14, 2019, 1:56 pm

GC7 wrote:Hitech should have calibrated your machine when they installed your PID. You should have been told the exact offset from the boiler set temperature to the puck. The $350 you spent for that was more than expensive enough for what looks like their standard delta PID without shot timer or other new features of more modern espresso temperature controllers.


I would have liked that, but that's water under the bridge - The PID was put in 6 months ago.

RyanJE wrote:I had an Alexia with the group thermo as well as a Scace... your group should idle after a 50 minute warmup with the PF in at 198 ish... the group thermo reads about 3-5 high about half way into a 25-30 sec shot..

For me, my machine needed a 24-25f offset rather than the stock 18f it came set with.


That's helpful, and sort of consistent with what I'm seeing. The 24-25f offset - was that with a puck in?

mad1
Supporter ♡

Postby mad1 » Jul 14, 2019, 2:13 pm

maximatica wrote:I had a switch that changed from the PID to the pressurestat to add heat to the boiler on the fly for steaming and then I would wait and use that heated water to help maintain the temp through the pour.



That's really interesting - is it possible to program the stock PID to have a more aggressive heating profile? For example, if the temperature drop slope is significant, then we're probably pulling a shot, and we should just max out the heat. I'm noticing that during a shot, the PID pulses the heat - sort of like a 20% heat on, 80% heat off

maximatica wrote:But I eventually had had enough with the futzing and got a BDB and I am much happier.

That's surprising to me What is it about the BDB that does a better job at intrashot temperature maintenance?


tracer bullet wrote:Other threads here in the past showed a group head temp approximately 5-6 degrees over the puck temp and so that's what I've shot for myself as well. When I've tried a few ways to measure puck temp (portafilter with a dense sponge stuffed in it, a thermocouple in that, and a small needle valve to act as a flow restrictor to allow water out of the portafilter around the same rate as espresso does) - I do indeed get around that 200F mark. I say "around" because every test was a little different but in the end it was in the neighborhood.

Thanks for the datapoint. What was the variance between tests - when you measured temperature at the puck, did some tests have a wider grouphead-puck differential, or were most/all in the 5-6F range?

tracer bullet wrote:After that the idea is to adjust based on taste and to be honest I'm still learning to do that. But at least (I think) it's a good starting point.


That's what I've been told to do too. However, the numbers are telling me something is off. Take the shot in the video above for example. That was approximately 198F throughout the shot, so let's say 193-195 at the puck. 30 seconds, 18g in, about 30g out. The coffee was a light-medium roasted SO (I forget which, but most of the SO's I get tend acidic). Based on my understanding of HB, at 193F, I should expect a super sour shot, no? However, it was balanced! And that's been pretty consistent over the last 6 months that I've had the PID. So, I'm "happy" with the balance of the taste I'm getting with the current settings. My cups generally taste fine - balanced and more complex than 90% of the shots I get from 3rd wave cafes here in the LA area. However, I'd like to be able to reproduce the recipes of folks on here who claim to have seen God in a cup (to mildly exaggerate :)). And for that, I have convinced myself that I need to know the puck temperature.

I guess that another question that comes up is - "OK, if my puck temp is 193-195, how come a shot with the above parameters tastes balanced, and a slightly longer shot tastes bitter?" This is with light-medium SO's or blends.