Using a PID to control fermentation temp

[AndyBWood]

Evening All,

Whilst I fully appreciate that you can't control a brew fridge (compressor) for the cooling with a PID does it seem feasible to run the heating only element with a PID ?

I've just set my fridge up for the first time this weekend and it has a 60w heater installed.

The heater is obviously designed to be pretty effecient as it stays nice and hot after the power has been switched off by the STC1000. As such, the temperature still continues to climb.......

So, fridge aside and just looking at an insulated box. Does running a simple tube heater via a PID make sense ? It learns the enviroment, it only provides enough power to keep the Status Quo as best it can. If things get too hot, however, the STC can still turn on the fridge compressor.

So two temperature probes needed but just my Sunday evening thoughts. If the PID is clever then, as long as the 'outside' ambient temp isn't too high or the brew itself is making too much exothermic heat then then the compressor never needs (perfect world) to be switched on anyway.......

Your thoughts please...

 

[evanvine]

I see where you are coming from re your heater overshoot, and yes a PID would greatly ease that condition.
Looking at your pic it would seem that you are monitoring the FV temp and not the fridge temp!
If this is so it will cause extremes of heating and chilling.
For FV monitoring you need to directly heat/chill the wort.

 

[AndyBWood]

Thanks EV,

I've read two schools of thought on this; measure the temp within the fridge or measure the wort itself.

To be honest, both sides of the story seem to be plausable with supporters on both sides of the fence.... :wha:

Once the current beer has done fermenting I'm going to trial the PID running the heater idea with 25 litres of water just to see if it makes a better job of it. Will then try some experiments with in the wort / out the wort temp control.

Thanks again.

A

 

[winelight]

Looking at your pic it would seem that you are monitoring the FV temp and not the fridge temp!
If this is so it will cause extremes of heating and chilling.

This is a compelling argument for not doing that bubble wrap thing with the temperature probe. It's also a lot easier, especially when you have to get the FV in and out like I did yesterday to check the SG.

By observing the temperature display and the cool/heat lights as shown on the STC-1000, when I make a change, eg from 21 to 12, or an increase say from 18 to 21, the temperature changes plus heating/cooling cycles are exactly correct, that is to say, if increasing, you only ever see the heater on, there is no overshoot, and if reducing, you only ever see the cooler light come on. By observing the duty cycle it's clear that it takes many hours, as you would expect, to effect a large temperature swing on 23l of wort. It takes a long time for the fridge temp to initially reach the target, and then of course the bulk of the wort quickly heats it up (or cools it down), so you get another long cool/heat cycle, and these cycles steadily reduce in time as the wort is reluctantly dragged to the correct temperature.

 

[winelight]

And a comment to the OP, personally I don't trust the plastic fridge shelf supports with the weight of a full FV and like many of the other pics of brew fridges around here I've made a wooden support frame underneath the shelf. But I suspect we're being over-cautious, I've certainly never heard of those supports breaking.

 

[AndyBWood]

Hi Winelight,

Re your comments on the bubble wrap method on the outside of the FV, in my way, attempting to measure the temperature of the wort.

It is interesting and I will conduct some proper experiments after this brew is finished. For this evenings games, however, I have removed the STC temp probe from the bubblewrap and am now just measuring the temp of the fridge air. I have, however, added a second reference thermometre below the bubble wrap.

Not that scientific but will keep my eye on it as best I can this evening and see what happens.

Both air and wort currently read 19.8

My target temp is 20.0

I'm waiting for the air to drop to 19.7 which will trigger the STC to supply power to the heater. (lowest 0.3 degree differential set) Power will then continue until it reaches 20.0. The crux though is that even though the power is cut heat will be applied to the system as the tube heater remains hot for several minutes.

It will be interesting to see what happens next.....

Re the plastic supports I think they will be fine; pretty chunky in this old fridge. To be honest I'm more concerned about the shelf and I'll put a 'just in case' piece of timber below it just in case.

Thanks for the input; appreciated :thumb:


A

 

[graysalchemy]

I must say I can see where you are coming from however it is the wort temp which you are controlling and not the air temp as the wort has its own sytem of heating all by itself, so for me it is the wort temp which needs to be measured.

 

[Belter]

John Palmer and everywhere else I've read suggests you have to measure the Wort temp. If you measure the fridge temp you aren't controlling the temp of the wort wbich is what you actually want to keep in line.

 

[evanvine]

so for me it is the wort temp which needs to be measured.

OK Alistair, how do you recommend doing that without directly heating/chilling the wort??

 

[graysalchemy]

Surely if the temp of the wort is measured it doesn't matter how hot the the air temp is as long as the heating switches off at the desired wort temp. As for chilling I do directly chill the wort with a copper coil and a maxi chiller.

Off course there will be variations in the temp of the wort throughout the FV be surely it is better to measure the temp of the wort than the air temp which may have no direct correlation to the temp of the wort as the wort is producing heat itself, which by measuring the air temp you are not directly taking into account.

 

[evanvine]

I hear all of what you say GA, and don't disagree.
The point I'm trying to make is the response time of 23 ltrs of wort in an FV to air temp.

 

[graysalchemy]

Point taken Jim :thumb:

 

[SloeBrewer]

I've been thinking about this recently, as I'm at the stage where I'm going to have to add a heater to my brew fridge, having previously only needed cooling. I think that possibly the best solution is to have the probe on the side of the FV, protected by bubblewrap, during the initial cooling/heating as required, to get the wort to whatever temp you want between 18-20C. Once it's there, then remove the bubble wrap and try and maintain a steady 19C ambient in the fridge. This should avoid the overshooting issue and therefore keep the wort at a nice stable temperature, as it's thermal mass means it changes less rapidly than the air mass inside the fridge.

For crash cooling, you'd want to insulate the probe though.

 

[winelight]

It's not difficult (although I'd never get around to it, to be sure) to put a temperature probe actually in the wort, another doing the bubble wrap thing, and another within the fridge air space, and see what happens. But eventually, assuming reasonable insulation, the fridge air temp will mirror the wort temp (and vice versa), even with the wort producing its own heat.

Not sure that any of us could taste the difference in the finished product, anyway :)

 

[graysalchemy]

Not sure that any of us could taste the difference in the finished product, anyway :)

Speak for yourself but I can sure tell the difference between a fruity estery bitter brewed at 21c and one brewed at 18-19c and I sure as hell can tell one with fusels in it.

 

[winelight]

Sure, but I don't think you're going to see a difference as big as 3C in the wort temperature depending on where you put the STC temperature sensor.

(edit) I see from another thread that someone has observed a wort temperature swing of only 0.1C when using the bubble wrap method. As pointed out in that thread, if you measure in the wort directly, you automatically have a swing of 0.2C (because that's the STC-1000 minimum).

Anyway, 0.1 or 0.2 is far better than I'd have guessed at. I was thinking 0.3-0.5C.

As I mentioned over there, I'd now like to see the figures when the probe's just in the fridge air.

 

[Fil]

in my small brewfridge with a built in blower and a pc fan i stuck in over the heater both running 24/7 so the air should be moving and transfering as much heat as possible i find that if my wort goes in a few degrees out the atc800+ i use can cycle and chase itself when the difference is set low at 1C however if i set it wide at 3C the system quickly reaches optimum wort temp without cyclig and the following day i can return and set the diff back to 1c..

my sensor is taped behind insulation against the fv side, and with this config the opperation depends on if your changing temp or maintaining temp.. and if changing temp i use a wide difference to avoid cycling..

 

[AndyBWood]

Evening All,

Had a play last night and my set up did this with air temp measured by the STC and wort temp measured by a separate thermometer.

Air Temp Wort Temp Comments
19.8 19.8
19.7 19.8 STC power to heater
19.8 19.8
19.9 19.8
20.0 19.8 STC shuts off
20.1 19.9
20.2 19.9
20.3 19.9 STC power to fridge. NB NOT CONNECTED
20.4 19.9
20.5 20.0
20.6 20.1
20.5 20.1
20.4 20.1
20.3 20.1
20.2 20.1
20.1 20.1
20.0 20.1
19.9 20.0 STC shuts off power to fridge. NB NOT CONNECTED
19.8 19.9
19.8 19.8
19.7 19.8 STC power to heater

So, the lower the better but a wort temp swing of 0.3 degrees is, I suppose, pretty acceptable.

My original point, however, is this.

If the fridge had of been plugged in to the cooling side of the STC then it would have kicked in at 20.3 degrees and applied cooling whilst the heater was still providing ‘residual’ heat. In effect, there was no actual need for the fridge to 'fire up' at all.......

Of course, I fully accept that it all depends on ambient temperature, where the sun is and which way the wind is blowing so I guess its just a case of experimentation and learning your own set ups behaviours.

Returning to my original post though…….

Once this brew is finished I shall try the PID powered heater experiment with the temp probe in the wort. I just want to see what is possible and if PID power can actually narrow the temp swing.

Thanks to all for chipping in :thumb:

 

[winelight]

Would like to see what the figures are if the fridge is plugged in.

 

[David Woods]

I know this is a very old thread but did anyone ever have any more thoughts on this? anyone tried it? I have decided to have a go at using an STC for the cooling side with the STC probe taped to the FV and a PID controlling the heater with the probe measuring the space. I am feeling that in using the STC for both I have found as the OP said that the heater over runs the temp a lot so figured this would be worth a try with the PID "learning" the space . Was wondering about how to set the temps for each of the controllers so they are not fighting each other.

Any thoughts?