PID vs Thermocouple

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.

bobsstory182

Supporting Member
HBT Supporter
Joined
Jan 5, 2009
Messages
43
Reaction score
0
Location
Conifer
I have been in the process of gathering parts to get a new HERMS. I was planning on using a Johnson controller to control the temperature for the mash, and was going to get a thermocouple to be able to double check temperatures.

Lately, I've been doing research on PID controllers and I'm thinking I'm going to go this route. My question is, with a thermocouple probe, does the PID give as accurate a temp reading as a plain thermocouple? Seems to me if it is as accurate, this would be a cheaper way to go to monitor temps with a plus side of providing control.
 
A PID controller is an instrument which reads temperature from various sensor types (including various thermocouple types, RTD's and thermistors). It also provides various types of heat (or cooling) output signals for temperature regulation. In brewing most PID setups use RTDs as they are more accurate than thermocouples.
 
It will be much easier to control the temp with a PID vs the Johnson. The PID has some algorithm in it that predicts temps or something so it doesn't overshoot a set temp. Where as a Johnson will shut off when you reach a temp but it may continue to rise. Just and FYI in case that changes your mind.
 
That's what I'm thinking, thank you.

With the PID, can you adjust the temperature pretty easy or would that mess up the algorithms and what not when it's trying to learn the best way to keep the right temperature? I'm thinking in terms of the mash and then hearing it up for the mash out.

I'm pretty mechanically inclined but this electronic stuff is a whole nother story lol
 
are u doing electric or gas. a pid would be better for electric because it can prevent temp overshoot better by cycling the relay as you near the set temp. if u are using gas either one will work.
 
That's what I'm thinking, thank you.

With the PID, can you adjust the temperature pretty easy or would that mess up the algorithms and what not when it's trying to learn the best way to keep the right temperature? I'm thinking in terms of the mash and then hearing it up for the mash out.

I'm pretty mechanically inclined but this electronic stuff is a whole nother story lol


It learns how to heat and not over shoot based on your actual setup I believe. Like I said, I haven't fired mine up yet. But I believe you run a "test run" on the auto tune mode and it calibrates to your system.

So once it's calibrated to your setup you can tell it to hit any temp and it should hit it.
 
That's what I'm thinking, thank you.

With the PID, can you adjust the temperature pretty easy or would that mess up the algorithms and what not when it's trying to learn the best way to keep the right temperature? I'm thinking in terms of the mash and then hearing it up for the mash out.

I'm pretty mechanically inclined but this electronic stuff is a whole nother story lol

In a true PID controller the error (difference between the desired temperature and the temperature of the process) is differentiated and integrated and each scaled by an adjustable factor as is the error itself. These three scaled quantities are combined and the sum is the 'output'. Thus the 'algorithm' is fixed. The only thing variable are the three scaling factors. Most modern PIDs have a feature which automatically determines optimum settings for those three parameters based on throwing heat at your load and seeing what happens. PIDs work very well if the tuning matches the load. If it doesn't they droop or overshoot or take forever to reach the set point. Thus if you are tuned for a large thick mash and try to control a small thin mash you may not get good results.

Most of the 'PID' controllers you see here are not pure PID but have at least some elements of fuzzy control. The appropriate fuzzy rules still depend on the nature of the load and these controllers still need to be tuned and the popular controllers discussed here all have the autotune feature.
 
In brewing most PID setups use RTDs as they are more accurate than thermocouples.

Measure with a micrometer, mark with chalk, cut with an axe.

Your PID controller isn't going to be precise enough to matter what kind of temperature probe you get so going and buying the most expensive one doesn't make much sense.

It is actually more important that your thermometer respond more quickly to temperature changes so take a look at your whole setup and make sure that you're not using any overly bulky probes or ill-fitting thermowells.

The big thing that you win with an RTD is that they're "more repeatable" so at this level of automation the difference will be that you'll have less calibration should you need to replace a probe. Not none. Just because they use RTDs in the giant, multimillion industrial plants doesn't mean it is going to make any noticeable difference in your few-hundred setup at home.
 
I did say they are more accurate than thermocouples and they are given that systemic error is calibrated out as they are more repeatable. What I should have said is that they seem to be preferred in brewing applications for whatever the reason and this, as far as I can tell from what I see in this forum, applies to home brewing as well. How many people here use thermocouples?

If one goes to the Auberin site where most of the folks here seem to get there stuff, they have RTD's and thermocouples for comparable prices (around $15). Given that why not go with the superior technology?

Thermocouples do, of course, have their place. Try to clamp an RTD on a 1/4" liquid line in a refrigeration system, for example.

Note that the superior repeatability of RTD's refers to the same unit - not a replacement though a new RTD may well be more accurate out of the box than a new thermocouple. If your RTD says that the mash temperature is 140 °F (even though it may actually be 139 °F because of calibration error) it will continue to read close to 140 °F throughout the mash. A thermocouple cannot be expected to be so stable. Furthermore, in a thermocouple the systematic error can change over time requiring recalibration.
 
Ok cool, I think I understood that. I was looking at getting the auber 2352 which does have that fuzzy logic, so that sounds like it'll be helpful for tuning it to my system.
 
...

What I should have said is that they seem to be preferred in brewing applications for whatever the reason and this, as far as I can tell from what I see in this forum, applies to home brewing as well. How many people here use thermocouples?

If one goes to the Auberin site where most of the folks here seem to get there stuff, they have RTD's and thermocouples for comparable prices (around $15). Given that why not go with the superior technology?

...

A thermocouple cannot be expected to be so stable.

Ah-ha, this is a light-bulb moment! This explains in-part why people on these forums seem to be so reluctant to just toss on another temperature probe when they're getting mixed results.

I get my thermocouples 2 for $5 and coming from a data acquision background I don't think twice about tossing on yet another probe. At $15, I'd probably slow down throwing probes left and right.

Thermocouple temperature measurement is as good as your cold junction compensation. There has been a lot of headway over the recent years reducing that systemic error while bringing down the cost of the whole system.
 
I guess I'd be reluctant to put a $2.50 temperature sensor in my system simply on the basis of "The bitterness of poor quality remains long after the sweetness of low price is forgotten." but an even larger consideration would be the question as to where would I connect a second, third (etc.) sensor. I do know the answer to that question but would most of the membership here?
 
Back
Top