Monitoring Brew Temperature - E61 & Silvia - Page 4

[randomperson]

You're right, alto. If I thought I could install Eric's adaptor myself I would. If the Scace system were less than several bills, I'd do that. In the meantime I'll stay blissfully ignorant, and be happy with it.

[altoCalgary]

RP,

The installation of the adaptor kit is quite easy. Eric's directions are very clearly written, as is his FAQ. I used a 5 mm (millimeter) allen hex wrench to remove the capscrew. After finger-tightening the adaptor body, I used a small Crescent-type adjustable wrench to finish the tightening of the adaptor, the adaptor closure plug, later removing the plug and tightening nut and ferrules onto the thermocouple. I used a paper clip to determine the depth of the probe to the upper edge of the intersection and then used a digital caliper to measure this depth. I thought I might need a 7/16" combination wrench to hold the adaptor body while I tightened the nut and ferrules, but the body didn't move at all.

Finding a supplier for the custom probe proved more of a problem for me, being in Canada. But, as I noted in a previous post, I found one. I had a Fluke thermometer for measuring espresso in the portafilter, so I used it. If I was starting from scratch, I would buy a Fluke 53 II with data-logging capabilities or an Omega data-logging system suggested by Eric.

[HB]

Still have no idea what temp I am achieving though. Routine is now flash plus 9 count (6.5 ounces), 20 second rebound. Pstat at .95 topping out at 1 bar. Any idea what temp I might be getting?

That's longer than the 10 second rebound than I normally use to yield 201.5F; I would guess that you're in the 202-202.5F range.

[randomperson]

Thanks Dan -- that's just what I needed. I thought that flash plus 9 would get me to 197 before the pull and with a 20 second rebound I'd be up around 201 -- then again, I'm making up all these numbers!. I'll back off on the rebound and see how things taste after that. (I have experimented with flush, but not rebound. That's next.) That said, I'm really happy with my results -- go figure! But the analytical side of me craves some numbahs! Gotta have numbahs!

Bah. I know there is an expensive temperature measurement system in my future! It's just a matter of time.

[altoCalgary]

I have been collecting Andreja Premium/Vetrano performance data since Eric's adaptor arrived. These provide confirmation of the consistency of the adaptor system to allow dialing in a starting temperature and its corresponding average temperature. I have sorted the data according to starting temperature for ease of comparison. I have selected data of only pours with great taste.
And, yes, I am a half-blind one-armed barista suffering from a moderate case of ADHD/ADD/AADD. ;-) Well, at least the last part is true - witness my edits since posting.

(Equipment: Plumbed & filtered Andreja Premium with Sirai commercial PStat, recorded with Eric's adaptor, T-type ungrounded thermocouple & Fluke 51 II digital thermometer. The readings recorded below were made under no-flow, that means in air rather than water, conditions. Brew Pressure 9 Bar (max.) Boiler Pressure (mid-point 1.1 Bar. Please remember that I live at 3627 feet above sea level. This means that water boils in my house at approximately 205.5 degrees F.)

Table:

Pour Beginning Temp.(F)      Final Temp.   Ave. Temp.   Max. Temp
196.1                        200.7        202.6         210.5
197.6                        203.5        205.5         211.6
198.5                        203.3        205.5         211.7
198.5                        205.3        207.3         212.9
198.6                        202.5        204.7         210.9

198.8                        202.7        204.6         210.3
198.8                        202.8        204.9         211.3
199.0                        203.5        205.6         211.7
199.5                        203.7        205.6         211.6
199.5                        205.1        207.2         213.2

199.9                        203.5        205.5         212.0
200.0                        206.3        208.4         213.7
200.5                        204.3        206.3         213.1
200.5                        204.9        207.1         212.9
200.7                        204.7        206.5         213.1

200.9                        205.1        207.2         213.0

'Pour Beginning Temp' refers to temperature my adaptor/thermocouple/Fluke records when I begin my pour, just as I flip the lever up.

Typically, I turn on my APrem. at 6 a.m. (I know, I could be using an automatic timer, but getting up helps me get going in the morning.) I usually get ready to pull my first shot around 6:45 a.m.

Here are data of my APrem from turning on until hitting maximum temperature that I made yesterday. (This is with the Boiler Pressure set at mid-point 1.2, rather than 1.1 above. I made this pressure change recently because I wanted to test my espresso at brew temperatures including those in the 200 to 203 range. With a mid-point of 1.1, I had to wait too long for my system to rebound. In any event, a boiler pressure curve with mid-point of 1.2 has the same trend line as 1.1. I should also note that the adaptor/thermocouple/Fluke thermometer system is reading under no-flow conditions, that is, gas rather than water conditions. Further, I have not calibrated the Quick Mill Boiler Pressure gauge for accuracy. It is performing as it came from the factory.) As you can see, the temperature initially rises quite quickly and then levels off near the 25 minute mark.

Table:

Minutes     Degrees F

0              70.8
5              93.1
10            153.7
15            180.5
20            193.9
25            201.2
30            205.5
35            208.4
40            209.5
45            210.5
50            211.3
55            211.6
60            211.7

First, I have to cool down the water in my boiler.

When I flip the lever up, the water from the boiler, within three seconds, causes the thermocouple's temperature to rise past the pre-flush group's temperature (causing the HX hump profile) and then it begins dropping. I usually stop the cooling flush just before my Fluke reads 196 F when I use a 1.1 bar boiler pressure setting and 206 F when I use 1.2 bar. (Of course, I could flush until my Fluke reads 198, 200, etc. I try to find a consistent begin-flush temperature that results in a reduced-rebound waiting time. See also Geoff's quote below. )

Then I begin my espresso preparation routine. I have already put the beans in my MM E-B. I grind, using the WDT (Weiss Distribution Technique), tamp and lock in the portafilter. I glance over at the thermometer, wait if I need to or, if I am quite close to my beginning temperature pour, start right away. I flip up the lever and, at the same time, press the MIN/MAX button on the Fluke. Quickly, I push the MIN/MAX two more times to put it in AVG mode. That way, I can watch the temperature changes in close to real-time - my ungrounded T-type thermocouple is a bit slow. Then I lean over and watch the pour just beginning to come out of my 'naked' portafilter. When I feel my pour is done, I flip the lever down and at the same time press the HOLD button on the Fluke. All of the data is now stored and waiting for me to retrieve. I record the data as well as several of bits of other information on a data sheet I have created. After my pour, I do a wriggle rinse and shut it off at 196 or 206 F. This acts as my next flush. My whole process begins, again. I don't care much about time as long as I can complete my preparation routine. My brew temp and routine leading up to it allow me to make consistent shots.

BTW, If I want to be precise, I wait until the exact beginning temperature is reached. From the above series of pours, you can see that I was aiming for about 198, 199 and 200 F, but I was focusing more on the flow of the process than the exact beginning temperature.

If you refer back to Geoff's (Ozark-61) post of August 2005, this might explain some of the variance in my shots.

If you do your cooling flush to 201', then pull your shot after a little pause, then you will have a stable shot, maybe around 200' at the finish of the shot - pretty good! But, if you then prep your next shot, and then do the cooling flush to 201' again, then you will not have the same stability through your shot, due to all the cooling you've done (reduced heat capacity in the grouphead, I believe). It could even get down to 198'. I find that I might compensate by not cooling as much on the second / third shots, maybe 202' instead, and then you will finish up around 200-201', and you probably want to lengthen the 'post-cooling-flush' as well.

See also Bob Y.'s

if I've pulled a few shots, the machine settles into a stable pattern and I can pull shot after shot with no more shot-to-shot variation than a fraction of a degree F. I just have to start the shot at the exact same temperature each time, catching it as the temperature creeps back up after the last shot.

But when the machine has been idle for a while, and the temperature creeps up to 200 deg F and beyond, the first couple of shots after a flush of any length have a sharply dropping temperature profile compared to the shots that stay within less than one degrees if they are pulled after the machine has stabilized.

[pfieber]

AltoCalgary

In regards to the data you provided using the adaptor. Are these readings with the machine at an idle for a certain period of time or when you first start it up after an hour or so warm up period?

Have you done any similar testing with consecutive shots? I find my readings are different using consecutive shots.

Paul

[altoCalgary]

Paul,

I hope my additions provide more clarity and answer your first question. Also, I would expect your readings to be different for consecutive shots, just because the process occurs over a shorter period of time. I would like to see your results. If you want you can post them, or you can email them to me. Thanks for your reply.

Ron

[stevendouglas]

I was hoping to add a temperature probe to the Elektra A3. However, since it's not an E61 group, I don't if this device (or one similar) would work. There is a screw in the group head that is pressurized when pulling a shot (ask me how I know ). You can see it in the picture below:


However, I haven't seen any technical drawings to know how this screw goes into the group. Nor do I know how this screw compares to the E61 allen head screw in terms of length, diameter, or pitch.

Has anyone tried to add a adaptor to an Elektra A3?

[erics]

Hi -

The Elektra A3 has, if I remember correctly from the CoffeeFest experience, an M5 allen hd screw vice the usual M6x8 mm found in the great majority of E61 (and Silvia) groupheads. In addition, the screw in the Elektra is sealed by means of a thread sealant vice a copper sealing washer or teflon O ring (again from memory).

The adaptor described and discussed in this post has M6 threading and thus would not fit. However, first you need to confirm the threading dimensions of your Elektra and we can go from there.

Eric S.

[Bob Barraza]

Eric's memory is impressive! We looked at the screw on my machine and it is a smaller diameter than in the E61 group. It also appeared to me that tapping the threads to the larger diameter would also be a problem because there appears to be little room between the screw head and the flat part of the group. So I abandoned that approach and sought another.

In the Buyer's Guide to the A3 (The Bench Forum) Steve Robinson posted pictures during a partial tear down of the group http://www.home-barista.com/forum...ktra-a3-t7-20.html



Once you remove the lower portion of the group by removing the three set screws from the top of the group, you will notice that there is an O-ring type of gasket that seals the two halves. I took a chance and ordered 30 gauge thermocouples that I intended to lay across the gasket and guided it down one of the four holes in the lower half of the group (the center hole is used to screw in the shower head and screen).

It worked! There was just enough room to run the wire between the two halves and still be able to get a good water seal. It may have required a bit more torque than without the TC wire, but it does not leak. I plan to order a new gasket to have on hand because if I ever have to disassemble for any reason, I suspect that the gasket will have deformed and create leaks. In the end, a small price to pay, and much easier than drilling the group.

By running the TC wire towards the back of the group, I was able to route it through the openings where the 3-way valve is located. From there, it runs down under the machine and along the countertop to the meter. Basically, unnoticeable.



I also placed another TC under the screen by attaching it to the center screw. From there it simple lays between the lip of the portafilter and the group head gasket. By firmly seating the portafilter, it does not leak, although there is a marked increased tendency to channel. My meter is dual channel, so it was interesting, but I quickly removed this TC and just use the one in the group to monitor brew temp.



When fully assembled you really can't see the TC going to the group.



I am still learning about the best combinations of final rinse temp and rebound times, however, I am settling in on a 10 sec rebound time because it seems to yield tighter brew temp profiles. Intuitively, this seems to be consistent with the large diameter of the heat exchanger and the laminar flow characteristics of water under pressure.

The A3 is a very forgiving and an overall excellent machine that is far more capable than my skill level. However, adding the temp monitoring in conjunction with the WDT dosing results in very few sink shots.

[erics]

Thank you for the compliment Bob.

Here is another potential approach for Elektra owners - the Philips head screw that seals the water passageway drilling has, I believe, M5 x 0.8 threading. A US screw thread that is pretty darn close to this is the common 10-32 thread. If I were to designate the 10-32 thread in metric terms, it would be an M4.826 x 0.79375. This company specializes in all kinds of "fittings" which have potential:

http://www.beswick.com

In particular, this fitting appears to be a candidate if one were to bore the hole through to a 1/16" dimension vice the stock dimension of 1/32".

http://www.beswick.com/products/p...details.php?pid=86

There is another fitting on the list that is bored through 1/16"+ but the 5/16" hex size may be a little too large for the Elektra.

Eric S.

[expressotaglong]

FYI for you espresso instrumentation fanatics. A company called Small Parts Inc. sells compression ferrules which will allow you to use very small diameter thermocouples with conventional 1/16 diameter fittings. Here's the link:
http://www.smallparts.com/products/descriptions/fp.cfm

These ferrules will let you play with 1/32" and smaller diameter sheathed thermocouples. These smaller thermocouples should give you a boost in response time, plus they are nice and bendable. With the addition of a few nice bends you can route the end of the thermocouple back inside the sheet metal for a nice clean look. The small thermocouples can also be fed through plumbing and passages to take temperatures exactly where you want.

Lee

[edwa]

Eric,

I've read and re-read the posts and the downloadable instructions so please forgive me if you've answered this before.

When you describe the dimensions of value "D", the length of the paper clip from the copper seat to the intersection, they are all in decimal points. ie. 0.563". Yet the coat hanger method in the FAQ section doesn't seem as precise. That is a little confusing. My household tools can't measure anything that accurately and I'd have to check if my reading glasses could see anything beyond 1/32 of an inch. What do I need to purchase to correctly make the measurement or will a 1/16" scale ruler be sufficient?

I have no qualms about purchasing the Omega meter but the FUJI PXR3 digital display interests me. Does anyone know its read rate? Do you need a box to house it in? I assume it needs to be plugged into an outlet. Do they have on/off switches or do you just unplug them? Do the Omega thermocouples need special adapters to plug into the Fuji?

Sorry again, I've never been around thermocouples and temp controllers but I am mechanically adept, so I'm sure I can do the installation.

Thanks in advance,
Ed

[erics]

Good Morning Ed -

0.563 is simply 9/16", actually 9/16" = 0.5625". I will re-read the instructions myself from your vantage point and see what I can/should do.

The "paper clip" method would be suitable for, amongst other things, someone with normal to small size hands. The "thin coat hangar" method would be really suitable for anyone and might be easier to do because of its size. Both methods are simply a "poor man's depth gage".

"I have no qualms about purchasing the Omega meter but the FUJI PXR3 digital display interests me. Does anyone know its read rate? Do you need a box to house it in? I assume it needs to be plugged into an outlet. Do they have on/off switches or do you just unplug them? Do the Omega thermocouples need special adapters to plug into the Fuji?"

Answers:

I tried to find the read-rate specs on the PXR3 but was unable to - I would think 2/s and maybe even a little better. Heck, the inexpensive PID's claim 4/s and the HH506RA is 2.5/s.

You don't need a box but it does make the installation aesthetically pleasing. Boxes can be configured with a small toggle switch for on/off purposes or can be plugged in and left on (no problem there) or can take power from the machine. The PXR3 typically takes 1/4" stripped leads or it can be easily modified with a female miniature plug to accept a thermocouple with an attached male miniature plug.

Email me at erics@erols.com for any additional info pertinent to your particular situation.

Eric S.

[major4579]

I have a Fuji PXR3 controlling my Rosto roaster. According to the literature, the "Input Sampling Rate" is 0.5 secs. - is this what you were looking for?

-Marc

[edwa]

Thanks major4579 and congratulations on your big win.

Did you put the Fuji in a housing? If so where did you buy it or did you build it?

[major4579]

Yes the Fuji was mounted in a project box - remember I'm using the Fuji to control my home roaster, not my espresso machine. ANy inexpensive project box from Radio Shack would work fine. The Fuji PXR3 can be ordered with a number of options - one is power supply, either 100-240 VAC or 24V AC/DC.

One of these days I'll post some pictures.

As Eric S. noted there is no on/off switch (I use a separate one).

Hope this helps.

Marc

[JimG]

FWIW, I have an Eric-style sensor wired to a Cal 3200 PID controller (used as a display only) on my Silvia. This particular controller is cheap, can be configured for C or F, can display 0.1 degrees, and allows the user to easily calibrate both sensor span and offset. I have the 3200 mounted in an external enclosure right below the "real" PID for my machine.

I have no idea what its read rate is, but in combination with the sensor, the 3200 gives me valuable real time temperature feedback for pulling shots and for knowing the all-important group temperature. I have good reason to believe the readings are accurate - I just don't know how quick or slow the response time is.

If you are budget-constrained, and only looking for a display, you might want to consider the Cal in lieu of the Fuji.

EDIT to original December 2006 post: After doing some more testing on the Cal 3200, I have learned that the display changes very slowly with input from the t/c. Although I still have the Cal on my Silvia, and still find it very useful, I think I would use a different controller next time.

Jim

[stevendouglas]

In the Buyer's Guide to the A3 (The Bench Forum) Steve Robinson posted pictures during a partial tear down of the group http://www.home-barista.com/forum...ktra-a3-t7-20.html

http://i6.photobucket.com/albums/y230/home-barista/DSCN0507.jpg

Once you remove the lower portion of the group by removing the three set screws from the top of the group, you will notice that there is an O-ring type of gasket that seals the two halves. I took a chance and ordered 30 gauge thermocouples that I intended to lay across the gasket and guided it down one of the four holes in the lower half of the group (the center hole is used to screw in the shower head and screen).

It worked! There was just enough room to run the wire between the two halves and still be able to get a good water seal. It may have required a bit more torque than without the TC wire, but it does not leak. I plan to order a new gasket to have on hand because if I ever have to disassemble for any reason, I suspect that the gasket will have deformed and create leaks. In the end, a small price to pay, and much easier than drilling the group.

By running the TC wire towards the back of the group, I was able to route it through the openings where the 3-way valve is located. From there, it runs down under the machine and along the countertop to the meter. Basically, unnoticeable.

<image>

Bob,

After a very long time, I purchased the Omega H506RA. Just as you suggested it's a pretty simple set up to run the thermocouple into the group head as you show above.

HOWEVER, I don't believe that I used the correct thermocouple for the application. I used the one that came with the meter (SC-GG-K-30-36), which uses a glass braid insulation. The insulation shredded during the first uses and now the indications are erratic (it peaks at 208-209F even after the A3 has been idle for an hour or more, hangs out there and then drops as brew temp drops). I've ordered new thermocouples with a PFA insulation which should be more resistant to water and abrasion. Couple of questions:

1. Do you recommend the 30 gage or 36 gage wire? (I saw you used 30, but I thought the 36 might allow the o-ring to seal with less torque)
2. Did you use PFA insulated thermocouples or some other insulation?
3. Any other advice on this set up?

Thanks for your help. It takes me a while, but I do get there! I think that I'm really going to learn a lot with real-time temperature monitoring.

Steve

[Bob Barraza]

Steve,

The K type thermocouple that comes with the meter is inappropriate for use in the water path (as you have unfortunately found out). I used the 5SRTC-TT-T30-36 which is teflon coated. Unfortunately they only come as a 5-pack for $59. I agree that the 36 gauge would require less torque to get a good seal, but I was concerned that they might be more fragile.

Keep in mind that you will have to re-program the meter for the T type TC.

Since the meter is capable of simultaneously reading two channels, I took one of my spare K TCs that came with the meter, and I literally wedged it between the back of the group and the 3-way valve. This allows me to monitor both the temperature of the group as well as the brew water temperature. I did this on a lark, and I have found that knowing the temperature of the group is as important as knowing the temperature of the brew water.

This meter has three areas for display. I have set the top larger display as 'T1', and use this to display the brew water temperature. The lower left display as 'T2', group temperature. Finally, the lower right display as 'T1 - T2'. This lets me easily see the two temps and the difference.

I have been surprised by how much the group temperature can vary depending on idle time, frequency of shots, amount of rinsing, etc. Now I not only rinse to purge super-heated brew water but also to either cool or pre-heat the group which I like to have at about 200 deg. F. As you know, this group is a big chuck of metal that has a very significant heat capacity. As such, it is designed to buffer the brew water temperature. Hence, nailing the brew water temp with a cool group results in a cold shot which you can taste. Perhaps this only applies to this machine, but the group temperature is a 'loose cannon', and controlling it is well worth the effort. It may not be a problem with other designs, but I would not assume anything until I measured it.

Since you already have the dual channel meter and extra TCs, I would encourage you to try this simple setup. Let me know what you think.