La Marzocco - Linea - PID Question - Page 3

Need help with equipment usage or want to share your latest discovery?
Billc
Posts: 304
Joined: 15 years ago

#21: Post by Billc »

Not sure exactly what settings you are changing and I assume you are using a Fuji PXR3. There are a few settings to change before you run auto tune.

Check these settings:
TC = 2 (cycle time)
Pn2 = 2 (Type of thermocouple 2=J)
P-SL = 32 (lower measurement limit)
P-SU = 220 (Upper measurement limit) (determines full scale reading)
Sv-L = 180 (Minimum set value)
Sv-H = 220 (Maximum set value)
Pdp = 1 (do this last for 1 place decimal place)

PVOF = 0 (reset this to zero for auto tune)

Then run Auto tune again.

This should give you some good PID values then the offset in temperature between the boiler and the group head will be about 3-4 degrees (lower than the display reading).

Also, you don't need to do a flush like you would a heat exchanger. The flush is only to heat the external components like the screen and such before you brew.

User avatar
Euology101 (original poster)
Posts: 184
Joined: 13 years ago

#22: Post by Euology101 (original poster) replying to Billc »


Bill,

I appreciate the info, I'll go over all of those settings tonight, and then run the Auto Tune again. After I finish with the Auto Tune, should I change any of the P,I,D settings? Reason I ask is because when looking at the instructions from EP, it appears they change some settings, run Auto Tune, and then change the PID settings according to Greg Scace's #'s, after the Auto tune has run (almost seems like it cancels out the Auto tune).
I'm not trying to replicate anyone, and I am more than willing to mess around with settings, to get it close myself, just feel that having a 14F offset between the boiler, and the head seems, odd, especially when you are telling me it should be closer to 3-4.

Thanks for your help!
-Justin

mitch236
Supporter ♡
Posts: 1231
Joined: 14 years ago

#23: Post by mitch236 »

From my experience and readings of Greg's papers, the autotune program doesn't do a very good job at setting the P. I. D. You should probably start with Greg's settings and adjust from there. There is a great paper describing how to set the PID by Greg. Look for it and follow the instructions. There's a good reason why EPNW ships the PID with a copy of Greg's settings.

BTW, setting the PID is the most important aspect of temperature stability. I really had a hard time setting the offset because there is so much deviation during the shot. But the truth is who cares what temp your PID is displaying, you will dial in whatever works for whatever you are using and will learn how your machine runs. The PID settings, OTOH, determine how stable your machine is both as you brew and for multiple shots. I strive for consistency. When you have consistency, you can achieve greatness. Without it, you are lost.

User avatar
Euology101 (original poster)
Posts: 184
Joined: 13 years ago

#24: Post by Euology101 (original poster) »

Mitch,

Do you have a link to those papers? I agree that I think the Auto Tune appears to not do as nice of a job with the P I D, but at the same token, I think that Bill Crossland is a pretty reputable source on the Linea aspect of things. :D

I'm just trying to accomplish 2 things at the moment:

1: Figure out why my offset is currently around 14F versus the 3-4F that I am hearing everyone say it "typically" is. <-- This can vary depending on a million things, I know, I'm just trying to understand IF "it is what it is", or if I am setting something wrong.

2: Figure out why am I flushing 6oz of water, to get the right temperature water. It seems like a lot more water, than what should be in the AV system of the machine.

All this being said, I am using a crude method of measuring, which I know is not correct, but I should be able to get "in range" until I can get the funds to rent a Scace.

I've been using the machine and pulling shots daily, it is working out great, but I spend more time looking for water in the correct temp. range than I do, making drinks. :lol:

mitch236
Supporter ♡
Posts: 1231
Joined: 14 years ago

#25: Post by mitch236 »

Here's a link that will lead you to the Greg Scace article. Optimizing PID Parameters in a Modified Heat Exchanger (I can't link the actual article as it resides on a Google site)

Your offset isn't really that important. It would be nice if we all had the same exact brew temps as that makes sharing parameters easier but as long as your machine is stable, you will figure out what settings to use for whatever beans you are using. I contend that it is next to impossible to properly set the PID without the Scace Device. The method you are using is crude and prone to error so I'm not surprised your offset is questionable.

Billc
Posts: 304
Joined: 15 years ago

#26: Post by Billc »

Depending how new your linea is 6oz of water is not that much to get it up to temp. The Linea was made to use used constantly, especially the AV model. During the first few shots you have to heat up the flow meter and the 3 way valve. After that it will be right on.
One more very very very important aspect is the thermal response time of your measrement system. You can do a comparission to see how yours stacks up against the Scace and other fast systems. To get good temperature measurements you need to have a 98% reading in about 2 seconds. You can test this by boileing some water and then measuring the temp. Let the probe sit in there for a minute or so and see what your target temp is. Then cool the probe to something less than 40 degrees F. Then start your timer and place the probe back into the boiling water. Time how long it takes to get within 0.3 degrees of the target temp. I did this the other day with my system and it is less than 2 seconds (about 1.6-1.8).

This is a crude test but it will give you a good idea how your system compares. If it takes over 7 seconds you will probably never get any useable results to be able to adjust your machine.

Bill C

User avatar
Euology101 (original poster)
Posts: 184
Joined: 13 years ago

#27: Post by Euology101 (original poster) »

Billc wrote:Depending how new your linea is 6oz of water is not that much to get it up to temp. The Linea was made to use used constantly, especially the AV model. During the first few shots you have to heat up the flow meter and the 3 way valve. *snip*

Bill C
Bill,

I appreciate you chiming in. The Linea is a 1998, Bolted Group style. I'm glad to hear that the 6oz flush, doesn't sound to out of character. I swapped out the Styrofoam cup for a Fluke 51ii, with a Thermocouple wire over the PF. I'll test it to see what the response on it is, but I would assume that method/thermometer should have a fast response. I've found that after flushing 6oz, I let it sit for 30 seconds, and the first shot is usually cooler, but ends around the temp zone I want, the 2nd shot pulled, usually hits my target temp, around half way into the shot.

I know having a flat temp profile is impossible, so, what do people usually shoot for? Ending with the specific temp? Starting with it? I would say I have on average 3 degrees variance over an entire shot, once the system is warmed up, so should I shoot for say 198 in the first 5 seconds, or somewhere around the middle?

This is something I've always wondered, because you see roasters giving temps on their "recommended parameters" but you can't have a flat temp profile can you? If not, where do you put that temp that they recommend, at the beginning, middle or end of the shot?

-Justin

Billc
Posts: 304
Joined: 15 years ago

#28: Post by Billc »

Justin,
Congratulations! You have now advandced to the next level of temperature measurement. You will find that temperature measurement is not that easy when you want exact measuremnts. Especially on a dynamic system like an espresso machine. There is a procedure for qualifying espresso machine for the WBC and you will find that the data taken during this test measures consistancy of the machine. Essentially you measure the temp profile throughout the shot and plot this. Then do it several times and compre the results. With the equipment you have, a good way to do this is to somehow restrict the flow rate of your measurement instrument. It needs to be close to what you would expect while brewing coffee (about 120ml/min). Then run the machine and let it run. You will see the measurement stabilize at a temperature and most likey stay there for quite some time. The temp will jump around some +- 0.7 degrees. This is normal. Most measuring systems are good to about 0.5 - 1 degree F anyway (this is why data logging is so important). The most important part of measuring temp on an espresso machine is a consistent process. When a roaster gives you a temperature that the coffee is best brewed at it is really an estimate. It will get you close. Then you have to experiment and adjust a few degrees up or down. Unless of course you are both using the exact same machine (calibrated).

Whew! I think it might be good to start some more explanations of temperature measurement process. I will work on this............. may take me a while.


BillC

mitch236
Supporter ♡
Posts: 1231
Joined: 14 years ago

#29: Post by mitch236 »

I've often wondered that as well. If you read Greg's article, he goes to great length to describe how to measure the temps and that's what I've done. Even then I still wonder if I got it right however, my brewing temps seem to reflect the suggested parameters within a degree usually.

mitch236
Supporter ♡
Posts: 1231
Joined: 14 years ago

#30: Post by mitch236 »

Bill, are you saying that I should set the PID displayed temp to match the measured temp on my Scace device after the temp stabilized? Greg recommends running the machine with the exact timing that happens during the shot, with pf removal, grinding, tamping and flush all taken in to account.