Monitoring Brew Temperature - E61 & Silvia - Page 2

[RegulatorJohnson]

hi my first post on H-B long time lurker.

eric made me join

i have one of these in my expobar pulser. its great!

everyone should get one.

BTW the most entertaining part was ordering a HTTC from omega. it was delivered in a GIANT box. heheh i thought they sent me 10,000 pieces accidently.

[HB]

Certainly I believe installing the probe as specified in the installation manual is the "correct" position... The heat transfer antics in the water flow path within the grouphead of ANY machine are complex and the closer you can get to the actual point of delivery, the better off you are from an information standpoint.

Take this kindly as I know you will and as I believe you know - not everyone is looking for the same information. Consider the chart below:


Four different repeatable measurement points. Which offers "better" information?

This is a composite of various measurements I've posted over the last couple years, representing the typical output of the thermofilter (red), over-the-basket-lip (yellow), your TC adapter (blue), and just for fun, the bottom of the basket (green). Being closer to the point of delivery at a standard position like the thermofilter is better if you want to compare brew temperatures among machines, but an equally interesting question is this: How can one squeak more performance using sensible temperature management from "old school" machines (for those who can't afford GS3's)?

That's what I'm liking about this adapter. The thermofilter makes machine-to-machine comparisons more meaningful. As I documented in Developing Better Temperature Control Skills of the Elektra A3 writeup, it also gives wonderful hints about how one can optimize performance. But if you already have a good handle on a machine's operation, your TC adapter offers more revealing information since it provides real-time data during the shot, and more importantly for HX machines, information about the flush preceding the shot.

Metacomment... During the A3's Bench review, almost a month went by before I bothered measuring the brew temperature. What was the point? I already knew what the meter would say based on the shots' taste. But offering me help on getting that 4th shot quality out of the 1st or 2nd shot? That got my attention.

[BobY]

Intro:
I preface this report with the caveat that I have only used this adaptor for one day and never with coffee in a PF. Yesterday I ran tens of measurements on my Andreja Premium (25-second, 60ml shots), comparing Greg Scace's Thermofilter to EricS's adaptor. Most of the shots, except for a few to check out the initial flush, were a simulation of my daily routine: (1) After an initial warmup of about an hour, flush about 4 ounces and almost immediately, pull the first shot. (2) Right after the shot, another two short flushes to clean the showerhead first without and then with a brush. (3) Then a 4-minute wait while I prepare the next shot (without the PF, or in this case without the Thermofilter locked in place). (4) Lock and load and pull the next shot. (5) repeat from (2) on.

The only changes I made to the above routine was to vary the waiting time between shots to see what influence it would have on the shape of the two curves generated and on their relationship to one another.

Comments:
Eric's instructions are very detailed and generally excellent. There are a few small points that I would expand slightly but unless you have the adaptor, they're of no interest here so I'll pass them on to Eric in an email. But for those of you with an Andreja, the distance from the seat of the M6 allen-head capscrew plus the crushed copper washer, to the point of measurement for the TC probe tip was 0.630". It will no doubt be different on other machines. Eric calls this "the adjusted value of D". At this length, the probe tip is just slightly into the intersection where the water takes a 90-degree turn on its way to the showerhead.

As Eric has already noted, the Omega HTTC probe has very fine wire attached and I'm now trying to figure out how to apply some strain relief to keep it from breaking. But in general, the accuracy (checked against a lab standard) and the response time is amazing. If I put the probe tip between my fingers, it goes to temperature in less than a second and immediately stabilizes there. So I guess it's a toss-up between fragility and function.

The adaptor fit perfectly on the Andreja (as opposed to another user's experience with, I suppose, another type of machine) and Eric's "paperclip method" of determining the correct point for the tip position, worked just fine.

Observations:
It appears, looking at my data and the generated curves where I recorded both temperatures (T1 and T2) and the difference between them (T1-T2), that no matter how long I wait between shots, anywhere from 1 to 5 minutes, the shape of the two curves and their temperature relationship to one another varies very little. At the beginning of the shot they are about 5 deg F apart at the point just after the initial buildup to maximum temperature and converge to less than 1 deg F at the end.

Actually, this is good news because, using the adaptor for a flushing guide (its primary purpose for me), I can be relatively certain that if, I flush the machine until the I'm about 5 to 6 deg F above my intended "starting temperature", I can quite accurately predict the temperature level and profile of the entire shot, given that the Andreja holds the shot to within a 2-degree F range from the maximum temperature at the beginning of the shot, to the minimum temperature at the end. Again this is only after one day of use and I didn't play very much with flushing routines vs. time between shots to attempt to change the shape of the temperature profile. The profile illustrated is the one I use for my daily drinks.

The curve that I have posted is very typical of all the curves generated yesterday. The main difference between them is the temperature on the Y axis. I could almost go to the graph and change the values (within a reasonable range) and the curves would look the same. I said "almost". But again, what this tells me is that the adaptor can be a valuable tool for judging how much (or long) to flush and also gives some usable information about the shot temperature because at the end of the shot, the two temperatures converge.

BTW, it may seem, looking at this curve, that there's a huge difference in temperature between the two devices. But consider the fact that the temperature range (axis) is very small - only about 15 degrees F so the distance between the temperatures seems exaggerated. The difference between the two, when the shot is at its maximum point in the Thermofilter is only about 4.5 deg F.

So far I'm a happy camper!

BobY

[BobY]

... That's how I arrived at 206F as the "right" temperature to stop flushing....

After yesterday's testing, I'm also thinking that somewhere between 206F and 207F will be the ideal stopping place for a flush.

Edit: After today's testing, I now must qualify the above. This is only true when my machine has been active and has not been idling for a more than about 5 minutes. If the machine has been idling for, say, 10 minutes, then a flush, stopping at 206F will produce too hot a shot. The machine needs a lower flush-stop temperature after a long idle. I'm still checking this out to zero-in on the correct rule of thumb.

[erics]

OK, as previously promised, here is my attempt at loading a decent chart on HB. I attempted to duplicate Bob's starting temperature for the adaptor thermocouple and Silvia was reasonably cooperative. The variation in starting temps for the Thermofilter was due to me forgetting about the heating coil being on and having to repeat the shot so soon after the "error".

Eric S.

[HB]

If any manufacturers / vendors / entrepreneurs are reading this thread and thinking how to productize the TC adapter, allow me to offer a suggestion: Whatever you do, make it "idiot proof." One way of reducing the thinking requirement is eliminating the temperature readout. Not to insult the fictitious Joe "Sixpack" Barista, but if he sees readouts in excess of 212F, all sorts of difficult questions could arise. If one can provide step-by-step instructions, Joe will be happier if you don't require that understanding proceed use.

So I propose the HX Ready-Temp Readout (tm):


HX-RTR, patent pending

The scales aren't necessarily linear. The red LEDs could be 216, 213, 210, yellow LEDs 209, 208, 207, and green LEDs 206.5, 206.0, 205, and 204. Instructions on getting temperatures "just so" could be easily stated:

• If the machine has been idle for 10 minutes or more, flush until the rightmost yellow LED illuminates ("warm up" flush). Note: The empty portafilter should be locked in.
• Remove the portafilter. Dry the basket. Dose, distribute, tamp. Flush until the second green LED illuminates (adjusted to taste).
• Lock in the portafilter.
• Immediately start the extraction or
• Wait n seconds, then start the extraction. [ed: depends on choice of flush-n-go or rebound; best method depends on espresso machine]

While I am a gadget junkie, one of the most appealing aspects of the RTR is its lack of digital readout. It encourages the barista to adjust temperature by taste, not a number.

[Ozark_61]

I would opt for the digital readout - shouldn't add much expense for a purpose built display and someone buying this class of machine *should* be able to understand the concept easily enough. I need to flush to 202' to prevent the brew temp from climbing too high during the shot. Maybe I should go back and lower my pstat to get the idle temp below 220'...
Geoff

[miKe mcKoffee]

I would opt for the digital readout -

So would I, and preferably with (adjustable) offset.

[annp]

I've been looking at Eric's adaptor and the Scace device.

I can see that Eric's adaptor is a better adaptation for Anita (cost wise) and I've looked at his instructions and read enough of the posts here to have developed the following concerns.

As I understand it, installing Eric's adaptor in Anita means removing the allen bolt in the E61 group head. Now this is a semi permanent modification, the thermocouple bolt is not designed to be installed and removed from the group head at whim - it's in there, or it's got the stock bolt in there.

This are my concerns:

Apparently the wire which would run to a meter of my choice from the thermocouple is fairly thin and fragile, so if I buy a meter that is super flexible and usable for lots of different temp applications, like a Fluke 50, it's going to have to sit on the counter next to my machine. My counter space is small and Anita gets moved around a lot for filling, cleaning and access to other things on the counter.

I've got slidey plastic furniture feet on Anita, so that's not an issue, but anything else attached to her with a thin wire becomes a problem. What is a concern is that I'm not the one always doing the moving and I'm not sure that the person(s) doing the moving are going to be as careful as I am.

Clearly, the solution would to be not to use a Fluke 50 meter, but mount a PID temp controller on the machine (as Eric suggested in his instructions) and just leave everything set up that way. No one else uses Anita but me and if the wire from the thermocouple adaptor wasn't floating around, I'm sure I'd be fine.

However, that does pretty much make Eric's adaptor and meter specific to Anita - or another E61.

My other concern is that I know I'm going to own another machine in the future and it may not be an E61 machine. So from that aspect the Scace device is appealing, because as I understand it, it'll work with anything that is 58mm compatible. I'd like to think that a future machine would have some built in temp monitoring and the Scace device and Fluke would let me quantify it's functioning. Or, it might be handy to someone else who'd want to use it.

Finally, - and please correct me if I misunderstand - I believe temp accuracy in relation to this application is a function of two things; The placement of the probe in relation to where coffee is in an espresso machine and the accuracy of the meter doing the reading.

I assume the Scace device is more accurate than Eric's device with a high quality meter simply because it is the same place relative to where the puck would be, and I could check temp in all my operating conditions and go from there.

The advantage to Eric's device is that it adds full time monitoring capabilities - the Scace device is a diagnostic tool only.

So, what do all you experts think?

Ann

[erics]

Hi Ann -

To try to respond to some of your concerns:

The thin, relatively fragile wire (30 gauge, teflon insulated) is only found on the 1/16" HTTC series thermocouples. The TJC series thermocouples utilize 24 gauge wire and, if you talk VERY nicely to Omega, they may wrap them in stainless braid for an additional $3. The TJ series thermocouples come with 24 gauge teflon wire and a spring strain relief (identical to the Scace Thermofilter). They will do the SS braid on those just for the asking (and $3). The HTTC is a very responsive thermocouple but that's not to say the others are bad.

A disadvantage to using a PID as a meter is that it needs power and should really be in a housing to "look right".

As far as temperature accuracy is concerned, any thermocouple from a reputable mfg and a good meter such as the Fluke 50 series, Omega HH506RA, Extech will give good results and can be improved by calibrating the temperature measuring components as a system. But the idea behind measuring temperature is really just to correlate with your tastes for a particular coffee. The thermofilter provides you with data that is more representative of brew water temperature than anything currently available.

Eric S.

[HB]

I assume the Scace device is more accurate than Eric's device with a high quality meter simply because it is the same place relative to where the puck would be, and I could check temp in all my operating conditions and go from there.

But the idea behind measuring temperature is really just to correlate with your tastes for a particular coffee.

Eric's right, the name of the game is correlating taste and temperature, which assumes you can reproduce a given temperature as desired. To demonstrate this point, I've created a series of videos showing the temperature reported by a thermofilter and the E61 adapter. These videos were made in sequence at different time intervals and none were omitted so you'll see real results and the limitations of the equipment. The target temperature was 201.5F. Most of the time I got it right, but once I was off by as much as 2.2F.

The technique to obtain these results is specific to the espresso machine, boiler setting, and probe placement. The machine was regulated for heavy duty use, so the pressurestat is at the top of the recommended setting. Subsequently the flush amount is slightly more (flush stops when E61 adapter reading passes 204F) and the delay before starting the extraction is short (approximately 10 seconds, also known as "flush and go").

Short idle period (< 3 minutes) after warm up flush (not shown). Result: 202.1F

Another short idle period. Result: 201.2F

(This video is a blooper - the phone rang, the Fluke auto-shutoff mid-shot. I was waiting for the local fireman to burst through the door next... )

Very long idle period (10+ minutes). Result: 199.3F

I intentionally did not do a "warm up" flush prior to this test and subsequently overshot the flush amount.

Medium delay (5-7 minutes). Result: 201.3F

[

Medium delay (5-7 minutes). Result: 201.5F

Long delay (10 minutes). Result: 200.0F

Long idle periods without a "warm up" flush are easy to overshoot the flush amount.

[annp]

Eric and Dan,

That is very helpful information from both of you, thanks for putting the time into your responses.

The information in comparison between both solutions is really helpful.

One thing I know for sure, which was especially emphasized watching in Dan's videos - I suspect I'm not flushing by "feel" long enough.

I'm not 100% decided as to which solution is right for me, but I do need to get some numbers. I'm not getting as consistent results flavorwise as I'd like and it points to temp.

Ann

[HB]

I suspect I'm not flushing by "feel" long enough.

You're probably a lot closer than you think. The pressurestat of the espresso machine depicted in the videos is cranked up because the last experiment was recovery testing. Your Anita is likely at a more reasonable setting, plus the Quickmills that I've used aren't known for mega-flushes (I would create a video of the Vetrano, which is similar to your Anita, but it shipped out shortly after the review was complete).

[annp]

Dan -

I was kind of in the middle.

I let Anita run a little longer on the flush this weekend, like 7 ozs instead of 4 (yes, I adjusted the pressure down the first day I had her) - and something is certainly different. I cannot identify it exactly as to what I tasted though, but almost as if there was a metallic taste and it's gone and another more pleasant taste is in the forefront.

I've got a new coffee though and I wish I didn't, because it is presenting all sorts of other hassles.

Still, I need to get some temp info.

Ann

[barry]

Apparently the wire which would run to a meter of my choice from the thermocouple is fairly thin and fragile, so if I buy a meter that is super flexible and usable for lots of different temp applications, like a Fluke 50, it's going to have to sit on the counter next to my machine. My counter space is small and Anita gets moved around a lot for filling, cleaning and access to other things on the counter.

fwiw, i run 30 or 32 gauge thermocouple wires with plugs connected to a variety of instruments/dataloggers which get moved around a fair amount. as long as there is no strain on the wire and no sharp bends, there doesn't seem to be any problem with wire longevity.


the smaller the thermocouple, the better the response time.


--barry "size does matter"

[arossphoto]

I kind of share the earlier poster's concerns about the thermocouple and external meter. Would it be possible to use this device to mount a modified digital thermometer a la Dave Stephen's No More Guesswork! Naked Truth of E61 Temperature Revealed?

Cheers,

Andrew

[barry]

I kind of share the earlier poster's concerns about the thermocouple and external meter.

i really think this is a non-issue provided some basic care in handling is exercised. don't jerk on the wires, don't pinch the wires. it's not that big of a deal.

[HB]

it's not that big of a deal.

And Dave's thermometer wouldn't be nearly as responsive as the HTTC thermocouples.

[arossphoto]

i really think this is a non-issue provided some basic care in handling is exercised. don't jerk on the wires, don't pinch the wires. it's not that big of a deal.

I agree that it's not that big of a deal and would certainly be more accurate, but having a small digital display attached directly to the group head would certainly be nicer aesthetically, and would make the kitchen look less like a workshop.

[arossphoto]

I was just looking at the dimensions of the adapter, and it doesn't look like it would even be possible to mount a thermometer. The hole in the adapter for the thermocouple is 1/16" and the thermometer I have here is 1/8". I imagine this is pretty standard and I doubt it would be possible to find a thermometer with a 1/16" diameter.