kanoyu, greetings and thanks for sharing your trials. I have been flushing about the same amount as you and our pstat settings are very close, though I need to wait considerably longer before I pull.

I'm jealous of the ability to have a Scace device on hand as I have no way to measure the temp exiting the group head. There was once talk of a Scace kit that could be rented but that probably proved to be too troublesome to pull off.

kanoyu wrote:I've had up to 2.5 deg F differences, but nothing approaching 5-10 deg F. Of course I could be measuring or monitoring incorrectly....

You underscore the variances of different machines. I have, based on communications with Eric, assumed about a 4 F degree difference give or take. Eric's installation manual in the first paragraph says,

The difference between the temperature that you will be measuring using this adaptor and measured temperature of the water hitting the puck is approximately 7-9 degrees F and may vary with different machines and different brew conditions.

Although I've seen Dan's graph before I can't be sure that my machine is doing the same as my eyes flick back and forth from readout to watching the extraction and without data logging I can't be a reliable source.

I feel your pain with the hassle of the pour-over. Although I'm plumbed I'm burning through Everpure cartridges, my next step is to take the Everpure off the line and just let the water softener be the only intermediate step. Although my drain is also plumbed I've started collecting the flash water and using it to rinse out the PF after the shot.

edwa,

One of the challenges for me has been the (lack of) consistency, or repeatability, of what I observe my machine is doing. This morning, for example, I followed the protocol that I outlined to you, got no hump (from the tc device), and went from 205.7 deg F to 203.8 deg F at a fairly regular pace (didn't record, so I don't know how linear). Good pour, nice crema, rich taste at first, followed by an unpleasant finish--may have soured on me (or it could just be OLD coffee!). In terms of variance between tc device and Scace, even that varies, as Den mentioned, during the the shot (intra-shot), but also (for me) from shot to shot (inter-shot). For example, I will get greater intra-shot variance on the first pull, and less intra-shot variance on the next one (30 seconds later). I suppose it could be that the Scace device is already super-heated from the first pull. I also get greater stability with successive shots (as registered by the tc device).

Taking a chapter out of other postings concerning method, try to work on one thing at a time. If stability is what you're after, focus on that. Might I humbly suggest the following:

- give your machine plenty of time to warm up
- prepare 2-3 baskets with the same coffee, weight, grind, dose, and tamp
- flush to target temp (on tc device), say 205 (if 4 deg F is your variance, 201 maybe good to shoot at)
- lock first basket and wait 35 (or 40, 45, 50-but try to be consistent) seconds
- pull shot and observe temperature fluctuation
- use the next 35 (or 40, etc.) seconds to remove portafilter, remove basket (helps if you have ridgeless or no spring clip), and lock second basket
- pull second shot and observe temperature fluctuation
- use the next 35 (or 40, etc.) seconds to remove portafilter, remove basket, and lock third basket
- pull third shot and observe temperature fluctuation

Don't worry too much about watching the pour, focus on temperature stability (if that's what you're after). Once you have your routine down, you can check the pour out. Also, there will be a lot going on, so you might want to video the thermometer so you can check it later (unless you have data logging and plotting capability).

It might be easier, and you'd burn through a lot less coffee and water, if you had a Scace, but I'm not sure it would give you any more information on stability. Once you have addressed the stability issue (and I hope you will meet with success here), then you can adjust flush amounts to zero in on temperature (tc device reading relative to taste).

Finally, if I had been nearly as methodical as I am suggesting, I probably wouldn't have struggled as much as I did (am?). I'm sure you will be wiser than I have obviously been.

BTW, that's exactly what I do with my flush water!

Good luck.

edwa wrote:I'm jealous of the ability to have a Scace device on hand as I have no way to measure the temp exiting the group head. There was once talk of a Scace kit that could be rented but that probably proved to be too troublesome to pull off.

Here's a cheap approach - this measures it EXACTLY, and in actual brew situation, and costs almost nothing.
However, it's not a Scace replacement as you can't bang out a bunch of tests with the same coffee puck (it gets cranky... )

the hook shaped "wire" to the right isn't a wire but a reflection of the wire



Side view of tc wire - fine gauge K-type - doesn't last forever but when you want just nip it off and make a new tc by baring the wires and tightly twisting together. is calibrated quite accurate.

Hey Kristi,

I've got a Type K and a cheap reading Celsius multimeter but its read rate is too slow. I have it snaked up the single spout PF an through a sponge for "some" flow restriction. The cost factor would be getting an Omega or Fluke. Your over the lip method would probably work better.

kanoyu, I have similar consistency problems from day to day. I'm pretty happy with the way things pull and taste and my procedure gets me close. I've come to the attitude that, in terms of temp, "close enough" works as long as I'm not hitting the "shoulders" and coming up bitter or sour. Kind of like tuning a guitar string by ear rather than a meter.

edwa wrote:The cost factor would be getting an Omega or Fluke.

I guess response time, more than "accuracy" is helpful - like a meter that will refresh every sec or half sec...

kanoyu wrote:edwa,

I don't know if this is any help, but I've been struggling to fine tune temperature on my Salvatore for almost a year now. I can hardly believe that others are flushing so little (on other HXs) after the sputtering stops. I need to do a 6-7 second count to get the temp down to about 206, as measured by Eric's thermocouple (it takes at least 10 seconds just to get to 212 degrees and the sputtering subsides). This means that from a HOT idle, I end up flushing about 8-10 oz of water. I then wait 30 seconds for a 203.7-204.4-203.5 hump (I have determined about a 1 degree difference between thermocouple temp and 'puck' temp by using a Scace device simultaneously). If I go about 25-35 seconds between shots, I can keep a pretty stable 202-204 on the thermocouple. But even a 3-minute lapse will require another 6-oz flush...

this seems to use and waste a large amount of water. reading many of the threads about hx machines the flushing routines all seem to do so. is there any alternative while still producing an admirable cup?

kikuchio

kikuchio wrote:this seems to use and waste a large amount of water. reading many of the threads about hx machines the flushing routines all seem to do so. is there any alternative while still producing an admirable cup?

kikuchio

Lower your pstat and work toward short flush and short pause - say 10 sec each. However your steaming capability will probably go to hell in a handbasket...

Attached are some temperature graphs of cooling flushes and subsequent group cooling/heating ONLY as measured on a Quickmill Anita using the thermocouple adaptor and a type T thermocouple. These graphs DO NOT represent any pulled shot. The combination of the thermocouple and the meter (Omega HH506RA) just happen to be pretty accurate as best I can determine - ice bath and boiling water and the thermocouple is positioned fairly accurately at the "intersection" as detailed in the thermocouple adaptor manual.

Each and every morning (Anita is clean and has been on for at least 90 minutes due to timer) I flush to about 185 F and start grinding, weighing, leveling, tamping, etc., etc. I flush into my one favorite cappy cup and, by all accounts, its turns out that I am flushing about 8-9 ounces (the cup does overflow). I am pretty anal about how much coffee goes into the LM ridgeless (17.0 g's) and equally anal regarding the leveling operation. I have timed myself and it takes me about 1-1/2 minutes to fully prepare the basket. This could be hellacious in a cafe.



As the grouphead is heating back up due to the action of the thermosyphon, I drop the basket into the PF (at, say, 197.5) that has been waiting patiently in the group and lift the lever at 198.0. This results in a good shot.

This particular method certainly uses up some water especially if you're making a single drink every 20 minutes or so in the AM but it does result in some pretty stable (and, more importantly, repeatable) temperatures. I have promised myself that I was going to try the shot initiation at GH temps of 197, 199, etc. etc. but the drinks at 198 are too good and I am, of late, lazy.

As always, this MS Excel file is available for the asking.

Skol,

Eric S.
erics@erols.com

And here is a temperature graph of a simulated shot pulled through a Scace Thermofilter starting with the grouphead temperature at 198 F.



Again, the MS Excel file is available for the asking.

Skol,

Eric S.
erics@erols.com

Hey Eric,

I did get your email with these data last Tuesday, but I've been out alot, and haven't had the chance to either respond or run a rigorous simulation on Mondiale. But these profiles look similar in "pattern/profile" to what I see with Mondiale. I don't have a recording device and I'm not gonna have time to "play" until maybe Sunday afternoon.....but in principle, this should be a better approach than surfing the temperature drop during the flush.

The downside as you suggest,. will be the increase in time to complete and increase in water-volume.....which in my case is not too much of an issue really on either count....as I'm usually the only one making an espresso, and I'm direct plumbed.

I'll give your scheme a whorl and make another entry to this thread with-in-the-week-or-so.

Cheers,
dmm