How do you know when Auto-Tune has finished

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Flomaster

Well-Known Member
Joined
Nov 5, 2010
Messages
2,084
Reaction score
38
Location
Orange
I ran an Auto-Tune the last two night all night long because I didn't know if the Auto-Tune had completed.

I was thinking the display would stop blinking when it was complete hence why I let it run all night long.

when in normal Auto mode does the display blink every time the element fires and display the percentage of power its getting? I didn't notice it doing that when I brewed last weekend with it.


at any rate I believe my PID is properly calibrated for my RIMS tube system after preforming two auto-tune sessions over night. maybe my auto-tune finished in 20-30 minutes and I just didn't know it.

I went into my settings and my AT was set at 3 which per the manual means its been auto-tuned.



-=Jason=-
 
now after posting I find this post by Kal


IMHO the best way to do an autotune is slightly below the target temp and to do it in the range where the PID will be working. For example, 154F for mash.

To quote myself: said:
At (auto tuning): The first few times you brew the temperature may rise higher or drop lower than the temperature you set. This is normal. The PID will automatically learn the heating and cooling characteristics of your setup and tune itself over time to provide more precise temperature control. This learning can also be forced by having the PID perform auto tuning. Auto tuning only needs to be performed once. If anything drastic changes about your setup, consider performing it again. To perform auto tuning, do the following:

1. Fill your tank with water and set the PID to a typical mash temperature of around 154F.

2. Once the water is approximately 10 degrees below the 154F set temperature, set the At setting to 2. Auto tuning will automatically start after 10 seconds. For the next 10-20 minutes the PID will automatically turn the heating element on and off multiple times to see how fast temperature rises and drops in order to set the proportinal, integral, and derivative parameters automatically. The A/M light will blink during auto tuning. Once auto tuning is complete the At setting will return to the default value of 3.


Kal
 
Wait what?

Yes I leave my PIDs on auto-tune all the time. Why not? The environment changes all the time. My tap water temperature changes, my basement temperature changes, the beer I brew changes. So why not auto-tune all the time and let my PIDs adjust to these changes?
 
Yes I leave my PIDs on auto-tune all the time. Why not? The environment changes all the time. My tap water temperature changes, my basement temperature changes, the beer I brew changes. So why not auto-tune all the time and let my PIDs adjust to these changes?

Not sure that's such a great idea. In Auto-Tune mode, the PID intentionally over-shoots your target value and the cuts off to see how the process cools back down to target value. You are going to experience a high-amplitude variance in your temperatures when running auto-tune as it's basically "bracketing" the target temperature to set parameters. Plus I'd be curious to know how much your P / I / D values change each time.

If you insist on running AT each brew, it may be best to do it before you are actually brewing, where temperatures actually matter. Even then, I doubt you'll see significant changes in the values of P / I / D.
 
the PID while running will learn your system will it not? Auto-Tuning just get it dialed in much quicker yes?

also how does one know when the Auto-Tune has completed? is the only way to go into setting and check if AT=3 or does that little A/M light actually stop blinking? im not sure I want to auto-tune again after doing it 2 times. but im inclined to try it just to see for myself.

-=Jason=-
 
Not sure that's such a great idea. In Auto-Tune mode, the PID intentionally over-shoots your target value and the cuts off to see how the process cools back down to target value. You are going to experience a high-amplitude variance in your temperatures when running auto-tune as it's basically "bracketing" the target temperature to set parameters. Plus I'd be curious to know how much your P / I / D values change each time.

If you insist on running AT each brew, it may be best to do it before you are actually brewing, where temperatures actually matter. Even then, I doubt you'll see significant changes in the values of P / I / D.

I already did what you suggest.

"Out of the box" my PID did over shoot but it also started with factory defaults for the P, I and D settings. Now it no longer over-shoots, I guess because it starts with the last learned settings. I wrote down my P, I and D settings after each brew for a few months and they varied slightly. Now I just don't bother & I let the "smart machine" take care of itself.

I need to write down my P, I and D settings after each brew for a longer period of time and compare the trend. I bet the winter values are grouped together and summer values are grouped together but the groups are different. My basement is a lot colder in the winter than in the summer and my water is a lot colder in the winter than in the summer.
 
I already did what you suggest.

"Out of the box" my PID did over shoot but it also started with factory defaults for the P, I and D settings. Now it no longer over-shoots, I guess because it starts with the last learned settings. I wrote down my P, I and D settings after each brew for a few months and they varied slightly. Now I just don't bother & I let the "smart machine" take care of itself.

I need to write down my P, I and D settings after each brew for a longer period of time and compare the trend. I bet the winter values are grouped together and summer values are grouped together but the groups are different. My basement is a lot colder in the winter than in the summer and my water is a lot colder in the winter than in the summer.

do you use your rims tube to heat your mash water from the tap?

-=Jason=-
 
do you use your rims tube to heat your mash water from the tap?

-=Jason=-

No, I heat my tap water in another pot to 170F. Then I do a dough-in and use my RIMS tube to bring my temperature back up to mash. My temperature usually ends up at about 135F after dough-in.

I think I understand what you are getting at - if I started with cold tap water every time my PID's auto tune would probably screw things up, but I don't. Instead it starts at about 135F and raises the temperature pretty quick.
 
Kenh, if it works for you, then fine. However, that is not how a PID has been designed to be used.
 
I just tested another auto-tune. My am light doesn't flash during the process. After about 40 minutes I went into the settings and my at was set to 3. I guess You have no idea when the auto-tune is finished.
 
the PID while running will learn your system will it not? Auto-Tuning just get it dialed in much quicker yes?

also how does one know when the Auto-Tune has completed? is the only way to go into setting and check if AT=3 or does that little A/M light actually stop blinking? im not sure I want to auto-tune again after doing it 2 times. but im inclined to try it just to see for myself.

-=Jason=-

I can't speak for every PID ever invented, but we did some research a while back and I actually talked to Auber and they confirmed that their PIDs do NOT have any sort of 'smart learn' mode. Once the P / I / D values are set, they're not changing until you manually change them or until auto-tune is run again.

I honestly don't recall if the A/M light blinked while running AT. The instruction manual for the Auber 2352 seems to imply that At will blink on the screen while in AT mode - but who knows. I'd say the best way to confirm is if the At parameter goes back to 3.
 
I can't speak for every PID ever invented, but we did some research a while back and I actually talked to Auber and they confirmed that their PIDs do NOT have any sort of 'smart learn' mode. Once the P / I / D values are set, they're not changing until you manually change them or until auto-tune is run again.

I honestly don't recall if the A/M light blinked while running AT. The instruction manual for the Auber 2352 seems to imply that At will blink on the screen while in AT mode - but who knows. I'd say the best way to confirm is if the At parameter goes back to 3.

I concur with checking the AT setting if at 3 auto-tune has completed. but when I let the thing run over night it was blinking AT and what ever value it was at the entire time. I had no idea if Auto-Tuning was complete or not and I hadn't thought to go into the settings to check.

maybe this thread will be useful for others new to Auber 2352 PIDs I now know Auto-Tuning DOESN"T take HOURS and HOURS to do.

-=Jason=-
 
I can't speak for every PID ever invented, but we did some research a while back and I actually talked to Auber and they confirmed that their PIDs do NOT have any sort of 'smart learn' mode. Once the P / I / D values are set, they're not changing until you manually change them or until auto-tune is run again.

I honestly don't recall if the A/M light blinked while running AT. The instruction manual for the Auber 2352 seems to imply that At will blink on the screen while in AT mode - but who knows. I'd say the best way to confirm is if the At parameter goes back to 3.

This is true, but if I understand it correctly, the "fuzzy logic" can change some behavior on the fly, while it will not change the P/I/D values.
 
Wasn't the 'fuzzy logic' just their way of describing a different method for how their PIDs analyize the signals and inputs and adjust things? Either way, it seemed clear that there was no "learning" taking place, and that if you never AT'd your PID, or you plugged random crazy numbers in for P/I/D, your PID would never "learn" the right values and adjust.

As for the blinking AT light - it's possible that you changed PID to 1, which means it'll start when you push the A-M button? Or maybe you had it in manual mode and it could never enter auto-tune mode to begin with?
 
Wasn't the 'fuzzy logic' just their way of describing a different method for how their PIDs analyize the signals and inputs and adjust things? Either way, it seemed clear that there was no "learning" taking place, and that if you never AT'd your PID, or you plugged random crazy numbers in for P/I/D, your PID would never "learn" the right values and adjust.

As for the blinking AT light - it's possible that you changed PID to 1, which means it'll start when you push the A-M button? Or maybe you had it in manual mode and it could never enter auto-tune mode to begin with?

no I changed AT to 2 and auto tuning started. I know the AM light was off so I wasn't in Manual mode. I had AT blinking on the display, but I thought Kal said when the AM light stops blinking Auto-Tune is done.

oh well. I have Auto-Tuned my system and in the future I'll check the AT settings to see if it has changed to 3.

-=Jason=-
 
Wasn't the 'fuzzy logic' just their way of describing a different method for how their PIDs analyize the signals and inputs and adjust things? Either way, it seemed clear that there was no "learning" taking place, and that if you never AT'd your PID, or you plugged random crazy numbers in for P/I/D, your PID would never "learn" the right values and adjust.

Correct. So in effect the "fuzzy logic" can make it deviate from the basic PID algorithm, in ways that Auber believes will improve its ability to rapidly reach and hold temp without overshooting. That's what I meant by changing behavior "on the fly." But that in no way constitutes "learning," as the PID values are not changed, and the "fuzzy logic" will start all over again the next time with no memory of the last session.
 
Back
Top