Scace Thermofilter Temperature Device - Page 8

[wogaut]

Moved from Synesso Cyncra Single Group - A Home Barista's Perspective by moderator...

I wish that were true, it would have made life easier for all of us. But in my experience, the intrashot result as measured with the Scace is sharply different than when measured with a T/C in the coffee puck a-la Schomer or by snaking the T/C over the lip. Scace will show a much wider temp variance than the coffee method. The coffee itself attenuates the reading. As an example, a stock home machine will show a 5f intrashot with Scace, while with coffee, it will show around 2-3f.

Maybe this discussion is more suitable for the Scace device thread, but as a response to Abe and Chris and my thoughts about the use of the Scace device method vs. Coffee Basket method:

What I observed with the 'Schomer method' in comparison to the data published at HB about the Scace device is a difference that's primarily in the first 5-6 seconds of the shot or so. In fact, the coffee method (if done properly, so that the probe is level and visible on top of the packed coffee) seems to get to target temp faster than the Scace method.

I was wondering: With the coffee method you have just a small room between shower screen and coffee puck to be filled with water until the probe is fully submerged in water. With the Scace thermofilter, in order to have the same water debit as the coffee method while having a solid 'puck', there's a larger volume to be filled up before the probe is completely submerged in brewing water. In your opinion, is there any chance that the slower growth with the Scace device (e.g. the graphs in Bob's post http://www.home-barista.com/forums/scace-thermofilter-temperature-device-t453-101.html) is related to filling up that volume, while the coffee method fills up the empty volume very fast and then starts soaking the puck with the rest of the water debit difference?
That would also mean, that a significant portion of the smaller differences in variability and higher average temps with the coffee method that some people observed can be attributed to what the probe 'sees' in these first few seconds. Which would further mean, one can't draw the conclusion, that the Scace method is more sensitive than the coffee basket method based on higher observed variabilities (as it 'picks up' wider variabilities), but this wider variability at the start of the shot is primarily a product of the used method?

And Chris, I'm writing this not to defend the relevance of the data I showed (you seemed to get impatient with me ), I also do think one has to compare apples with apples (maybe I'm not the only one still using the dinosaur method). I'm merely wondering, how the differences in variability can be interpreted, as I believe they might not be a true indicator of higher test sensitivity, as Abe's last message seems to conclude (maybe he didn't even imply that). I do like the Scace device as such, it reduces variability in the used method, since setting up an experiment with it is better controllable across testers than with the coffee basket. (Also the coffee basket is more messy )

Cheers,
Wolfgang

[Abe Carmeli]

Maybe this discussion is more suitable for the Scace device thread, but as a response to Abe and Chris and my thoughts about the use of the Scace device method vs. Coffee Basket method:

wolfgang,

You are making very good points here, and to make things even more complicated, Dan's experience with the Elektra, contradicts mine with the Brewtus. In his case Scace was flatter than coffee. Once you get coffee in the picture, the machine internals, pressure profile, preinfusion profile etc. all start playing a bigger role and affect the reading.

[malachi]

The primary trouble with the coffee (schomer) method is that it introduces potential variance - not only test to test but in particular from tester to tester. This means that comparing results becomes nearly impossible. The Scace may not be better but is both consistent and a real potential standard.

[gscace]

I wish that were true, it would have made life easier for all of us. But in my experience, the intrashot result as measured with the Scace is sharply different than when measured with a T/C in the coffee puck a-la Schomer or by snaking the T/C over the lip. Scace will show a much wider temp variance than the coffee method. The coffee itself attenuates the reading. As an example, a stock home machine will show a 5f intrashot with Scace, while with coffee, it will show around 2-3f.

That makes sense. In the coffee method, the thermocouple contacts coffee which is being heated by the water. So the time constant / response time is a combination of time constants for the coffee and for the thermocouple wire. Since the wire time constant is near to zero, the response time of the system is driven by the coffee. Temperature fluctuations are likely to be substantially attenuated by the coffee cake as well, so it is reasonable to expect fluctuations to be less in the method of placing a thermocouple wire on top of a coffee cake. The Ewe-no-hoo device prolly does a much better job of measuring actual water temp as it leaves the group. Whether or not one that is better than measuring the attenuated value is a reasonable question. On one hand, the attenuated system is real. On the other hand, isolating the temperature from effects of the coffee is more reproducible, easier to do, and is probably a better test of machinery performance, since reproducibility of temperature at the group means reproducibility of temperature in the coffee.

They're different measurements.

-Greg

[wogaut]

The Ewe-no-hoo device prolly does a much better job of measuring actual water temp as it leaves the group.
They're different measurements.

-Greg

Yes, there are different measurements, indeed.
I also agree that that device picks up less variability due to the TESTER than the coffee method as discussed earlier, since the procedure has less possible variables.

But I disagree, that the Scace Thermofilter (guess that's that Ewe-no-hoo device you're calling it????) must be better in measuring the "unattenuated value". I want to refer to my earlier post, that with the coffee cake in place the space between the group screen and the coffee puck is filled up within 2-3 seconds at most, while with the Scace device almost the complete difference in the water debit has to be filled up in the space inside the simulated puck before the TC is fully submerged. How else do you explain that several coffee cake measures (if properly done with the TC flush to the puck surface) reach a flat temperature state faster than the Scace device?!
If attenuation where the only possible explanation, then one should always see a slower growth with the coffee puck, since the "cold coffee" at the start of the shot should have the most attenuation. I rather think that as the heat capacity of coffee is significantly smaller than that of water there's not that much attenuation if the probe is placed on top of the puck. (and not touching the group screen!).
With that line of thought the higher variability observed at the beginning of the shot with the Scace device is less 'real' than the one with the coffee method, and doesn't necessarily reflect a more sensitive test.

Wolfgang

[gscace]

Yes, there are different measurements, indeed.
I also agree that that device picks up less variability due to the TESTER than the coffee method as discussed earlier, since the procedure has less possible variables.

But I disagree, that the Scace Thermofilter (guess that's that Ewe-no-hoo device you're calling it????) must be better in measuring the "unattenuated value". I want to refer to my earlier post, that with the coffee cake in place the space between the group screen and the coffee puck is filled up within 2-3 seconds at most, while with the Scace device almost the complete difference in the water debit has to be filled up in the space inside the simulated puck before the TC is fully submerged. How else do you explain that several coffee cake measures (if properly done with the TC flush to the puck surface) reach a flat temperature state faster than the Scace device?!
If attenuation where the only possible explanation, then one should always see a slower growth with the coffee puck, since the "cold coffee" at the start of the shot should have the most attenuation. I rather think that as the heat capacity of coffee is significantly smaller than that of water there's not that much attenuation if the probe is placed on top of the puck. (and not touching the group screen!).
With that line of thought the higher variability observed at the beginning of the shot with the Scace device is less 'real' than the one with the coffee method, and doesn't necessarily reflect a more sensitive test.

Wolfgang

Yeah. That's quite possible. In addition, the water flow rates may not always be the same. In the case of the coffee, the water is saturating the coffee and space between the grinds, and building pressure differently than with the thermofilter. So they are not the same here either. No matter as long as one compares apples to apples and as long as reasonable care is given to minimizing deviations from what is "real". I do find it very interesting that the techniques don't appear to produce the identical answer.

-Greg

[erics]

Greg, I would suggest that the instructions for your thermofilter device include some words re calibration of the thermofilter and the attached meter in boiling water (distilled?) for the barometric pressure at the time of measurement. I would hope that the Fluke 50 series meters can be adjusted in 10ths of a degree correction but am not sure of this.

Who is drilling the orifice in your device and is it 0.013" ?

Is Thermofilter II on the drawing board and does it include a pressure tap?

Eric S.

[hailseeszer]

I have been very interested in this thread and so I recently purchased one of these devices from EspressoParts.com. The device arrived in great shape, but the instructions were reproduced on a terrible Xerox. The pictures were blacked out and a good portion of the text was either smeared or missing. Luckily I printed the instructions from this thread although I am unsure if those are current, because the last post to this thread was some time in Jan. 2006.

I don't mean to be picky, but maybe some kind of case and a good quality set of instructions and support contact information would have made for a more professional type of a package (especially for the cost).

So now I come back to this thread and it seems it died some time in Jan. 2006 or maybe it has moved. I was just hoping I could continue to be involved with new posts if they exist. Please reply and let me know either way. I am concerned I purchased a device that might not have any after market support. Not only did I purchase this device, but I also purchased the Fluke 54 - 2. I've invested heavily in this so I just hope there is continued support out there somewhere...

Thanks!

Jim

[HB]

I believe the instructions and WBC protocol in this forum are up-to-date.

A little background on HB... The threads of the Bench forum are for review purposes. Some of the threads are research materials for upcoming articles (e.g., the recently completed Quickmill Vetrano); others, like this thread, are reviews in their own right. The thermofilter is an integral part of the calibration procedures for SCAA competitions and a key part of the HB review process. Greg Scace and Barry Jarrett designed and tested it; both gentlemen are active members on HB.

By the way, I have the same exact setup as you (thermofilter, Fluke 54-II). Rest assured, you've made a good choice.

[hailseeszer]

Thank Dan, for the information.

I guess I really found this discussion to be quite interesting so I investing in the device and wanted to be a part of any ongoing discussions. I'm sure I, too, would have had questions and possibly some input.

Oh well, maybe it will pick up again some time.

Thanks again!

Jim

[gscace]

I have been very interested in this thread and so I recently purchased one of these devices from EspressoParts.com. The device arrived in great shape, but the instructions were reproduced on a terrible Xerox. The pictures were blacked out and a good portion of the text was either smeared or missing. Luckily I printed the instructions from this thread although I am unsure if those are current, because the last post to this thread was some time in Jan. 2006.

I don't mean to be picky, but maybe some kind of case and a good quality set of instructions and support contact information would have made for a more professional type of a package (especially for the cost).

So now I come back to this thread and it seems it died some time in Jan. 2006 or maybe it has moved. I was just hoping I could continue to be involved with new posts if they exist. Please reply and let me know either way. I am concerned I purchased a device that might not have any after market support. Not only did I purchase this device, but I also purchased the Fluke 54 - 2. I've invested heavily in this so I just hope there is continued support out there somewhere...

Thanks!

Jim

Jim:

Email me with questions or call at two for zero two five two zero two for two. Sounds like you have the right components. Instructions from this site are current - same as the Xerox copy.

-Greg Scace

[hailseeszer]

Greg,

I really appreciate the support and I will email you if I have any questions. Don't have any right now, but that's mainly due to my waiting for my new Fluke to arrive.

Thanks again,

Jim

[hailseeszer]

Greg,

I know I just purchased a 58mm model for my ECM Giotto Premium, but I am seriously considering the purchase of a La Spaziale Vivaldi S1 which has a 53mm portafilter. I was wondering if you made that size? I don't see it listed on EspressoParts.com. If I do purchase the S1 it will be my primary machine. I may consider selling the Giotto, but not sure at this point. I also am not sure if I would want to keep the 58mm device. I wouldn't need a naked portafilter for the S1, because I'll get one with the S1 purchase.

Do you sell direct and would you consider a swap? I just purchased the 58mm this past week. No matter what I will want a 53mm if the S1 becomes my primary machine.

Thanks again for your support!

Jim

[gscace]

Greg,

I know I just purchased a 58mm model for my ECM Giotto Premium, but I am seriously considering the purchase of a La Spaziale Vivaldi S1 which has a 53mm portafilter. I was wondering if you made that size? I don't see it listed on EspressoParts.com. If I do purchase the S1 it will be my primary machine. I may consider selling the Giotto, but not sure at this point. I also am not sure if I would want to keep the 58mm device. I wouldn't need a naked portafilter for the S1, because I'll get one with the S1 purchase.

Do you sell direct and would you consider a swap? I just purchased the 58mm this past week. No matter what I will want a 53mm if the S1 becomes my primary machine.

Thanks again for your support!

Jim

Well I'm supposed to get off my dead ass and get the 53mm version out the door. I've built several and I'm gonna be supplying Espresso parts NW with them. Gotta get a few free hours to get it jumpstarted. So they're coming, and Terry will be the guy from whom to buy them.

-Greg

[barry]

Merged from Exercises for tuning your barista techniques by moderator...

Bob Yellin led the review of the Scace Thermofilter Temperature Device back in August 2005 and I've been using it extensively ever since. Its strength is comparing brew performance between machines because of its exacting measurement standard (the temperature probe is precisely in the same location, same buffer medium, flow rate, etc.).

while the probe might be in the exact same location within the portafilter, the probe is not necessarily in the same position relative to the brewhead (unless the device is locked to the portafilter body with a retaining screw or other means), as the device (as with any portafilter basket) may rotate on the portafilter body rim as the assembly is locked into the group.


--barry "have i really been gone for two months?"

[HB]

I wouldn't expect differences in grouphead clearance to make a significant impact because the probe is very close to the pseudo-puck, which buffers the temperature variance, and what's above the puck is a pure water cushion. The WBC protocol doesn't note the minimum water cushion depth needed for accurate results (i.e., if the probe is in too close proximity to metal, the brew water temperature readings will probably be too optimistic); perhaps it should. As noted in the thermofilter writeup, Schomer-style and over-the-lip probe placement results using real coffee vary considerably depending on the depth of the probe and the puck preparation.

PS: You have been gone for awhile, glad you're back!

[cannonfodder]

... (i.e., if the probe is in too close proximity to metal, the brew water temperature readings will probably be too optimistic)

If you are taking measurements on a well used piece of equipment, the location of the probe in relation to the dispersion screen will change. As those lug raceways wear, the locking position of the PF changes elevating the probe location. How much of a change would that make V.S. a new machine of the model, or are we just splitting hairs?

[HB]

...are we just splitting hairs?

Based on Andy's comments in Scace Thermofilter Temperature Device, I don't think so:

I recently got one, but I noticed some weird things happening. I think I know what the problem was.



If you look at this image you'll see the tip of the thermocouple sheath in the area of the red arrow. The tip is where the temperature is actually sensed. You'll also notice that the sheath runs parallel to the rim of the basket. That is good.

When I originally received The Device, the sheath was bent down slightly. The sensing tip was about a millimeter or so lower, just about touching the plastic. I believe that silly millimeter made a big difference. It meant that the probe had more of a tendency to report a composite temperature, part hot incoming water and part slightly cooler water that was running along the surface of the plastic. This made the readings artificially low, and it meant the readings were highly sensitive to how warm the plastic was before starting the shot.

I carefully bent the sheath back up so that it was parallel to the basket rim (the photo shows The Device after corrective rebending). My temp readings went up, and they became much less sensitive to the temp of the plastic before commencing the extraction.

His observations indicate that close proximity of the simulated puck to the probe changed measured results, so it suggests to me that a group with a dispersion screen that sits deeper in the basket could influence the reported water temperatures. If the WBC uses thermofilter results as a key measure of performance (and ultimately sponsorship), I wonder how long it will be before manufacturers "tune" their equipment solely to improve the reported results (e.g., by using a thicker dispersion block, specially shaped dispersion screen, etc).

As those lug raceways wear, the locking position of the PF changes elevating the probe location.

An older grouphead gasket may compress and change the vertical distance between the dispersion screen and probe, but lug wear would only affect the handle's lock-in position. But now we are splitting hairs...

[barry]

fwiw, i wasn't talking about vertical probe distance from showerscreen, rather the placement of the probe within an arc of all possible probe positions under the showerscreen.

[HB]

Understood. I assumed (?) Greg considered placing the probe precisely on-center so rotation would not matter, but ultimately decided the placement off-center nearer the pseudo-puck surface was more important for consistent readings.