better heat control / PID tuning or Thermostatic control

[Jakeintoledo]

Folks,

Pardon me, please, if this ground has already been ploughed, but I cannot find where my questions have been addressed after some searching.

I am having difficulty controlling the temperature of my mash using a RIMS system, and would like some recommendations on how to better control this system.

The particulars are this: At the mash cycle, the pump will come on and stay on, and the RTD sensor I have in the kettle (which is positioned to be in the center mass of the grain bed) will indicate the temperature to the controller. The controller senses the temp and controls a solid state relay to turn the heating element on and off.

The Problem: The temperature swing is HUGE. I can target 152, for example, and see a drop down to 147. most of my experiments have involved brews that I wanted to come out successful, so I will ignite the auxiliary propane burner under the kettle to more quickly recover the temp drop. When I do that, I'm really just adding a large error above the setpoint.

I don't know much about how to tune a PID loop, but I know that thermostatic control works very poorly. the temperature decay is faster than the system can correct for, if I set the thermostatic control to come on at .1 degrees below set point, the temperature doesn't start to come back until it's about 2 degrees below setpoint.

I think the answer is a PID loop, but help on tuning it would be helpful. The heater is an on/off control, so a PID output would have to be part of a calculation of a Pulse Width Modulator.

The first image below is of the system; the pump will pump through a semi-transparent flexible tubing to give an indication of fluid present. the enclosure you see is housing the RIMS tube (enclosed for safety). the inlet is on the far side, facing you in the image, and the outlet flows out of the top, back up toward the kettles.

 

[branman]

I have the same problem... just built a brutus style stand and the set point curb is very broad. I have experimented with different technics with no luck, I know this post doesn't help but I would also take in any suggestions from some of the vets out there

 

[KuntzBrewing]

My system is thermostatic, a herms tho. I have a 5500W element. And 13 gallons in the hlt. HLT is 4 degrees above set point, my pump kicks on when the temp drops 0.1 below the set point, turns off at .6 above set point, drifts to 1.0 degrees from turn on point. Perfect 1.1 degree window at max, sometimes only drifts to 0.8 from set point.

 

[helibrewer]

I would put your sensor on the output of the RIMs. You can use a second thermo to monitor around the grain bed and maybe tweak your controller based on an average between the two.

 

[clearwaterbrewer]

Agree with helibrewer, sensor as close to the output of the RIMS as possible.... has worked flawless, and if you mount your RIMS Vertical, you will expel air better, and if the mash sticks, the element will make some steam before burning grains, and this steam will reach the probe and keep a PID from increasing power trying to catch up... Actually, I know people that have a gas setup like yours and eliminate the RIMS tube and heater, just cycle the flame via a 2-stage lo-hi system and measure at pump outlet...

 

[branman]

Would it work just as well if I nstalled a thermowell at the ball valve with a "T" fitting. I am not using a rims tube myself just circulating from a false bottom back in through th top of my pot

The biggest issue I have is heat getting trapped under the grain bed and when I stir it can go from 150 to 168. ( But I've only done 5 brew sessions on it so I'm probably just doing it wrong)

 

[clearwaterbrewer]

My opinion is that it will.

I suggest the sensor after the pump for less chance of cavitation, position it so that any air will not be trapped, and before the ball valve. maybe the thermo-T on the pump outlet???

I would say if you overshoot, use a smaller max flame, and if wort is always running, even after you shut off the flame, you will not overshoot if you have the return wort set for your desired mash temp (do not try to increase you mash temp from 140 to 148 by having the return wort be 160, have it be 148 on the nose... it does work, even with thermal losses if you have a decent flow...)

-mike

 

[Jakeintoledo]

Thanks for all the input, guys.

I'm friends with the SABCO folks, and I know their system relies upon thermostatic control (as opposed to RIMS).

I DID take the added necessary step of installing a fitting at the end of the RIMS tube, if I decide to move the sensor there. The concern I have is that by doing this, I'm not really getting an accurate temperature reading of the grain bed, I'm getting an accurate reading of the fluid in the RIMS tube.

One of the steps I've done since creating this thread has been to install a throttling valve at the return discharge on the mashtun; by putting a valve here, at the end of the return plumbing, I've found I can better control flow to slow the whole thing down, thereby not compacting all the heat in the grain bed, and maintaining a constant flow around a loosey-goosey grain bed.

I think I may do the second sensor at the end of the RIMS tube. Version 2.0 MAY be a HERMS system for heating, though. I'd like quicker response, and my coil on my RIMS system is only about 3300W.

 

[branman]

My opinion is that it will.

I suggest the sensor after the pump for less chance of cavitation, position it so that any air will not be trapped, and before the ball valve. maybe the thermo-T on the pump outlet???

I would say if you overshoot, use a smaller max flame, and if wort is always running, even after you shut off the flame, you will not overshoot if you have the return wort set for your desired mash temp (do not try to increase you mash temp from 140 to 148 by having the return wort be 160, have it be 148 on the nose... it does work, even with thermal losses if you have a decent flow...)

-mike

This is good for me I will put the termo sensor there after the pump (never accured to me air might get stuck in the fitting)

 

[atoughram]

I've had my system both ways and putting the sensor at the outlet of the RIMS tube works the easiest in PID mode. I ran my PID controller in thermostat mode for a long time when I had the thermocouple in the grain bed. When I remodeled my system, I put the sensor at the end of the RIMS tube and ran it in PID mode.

Unfortunately though, the temperature in the middle of the grain bed is NOT the temperature at the end of the RIMS tube - the grain bed usually lags in temperature by five degrees or so, but catches up sooner or later. (I'm measuring my grain bed temperature at the outlet of the MLT). I'll have to graph the differences next time I do a step mash, that is good information to know.

 

[clearwaterbrewer]

I find that if you have a good grain bed, and recirculate 150 degree wort on top of it via a PID controlled RIMS, it only takes a few minutes to get the whole thing to that temp... the temperature gradient moves down the mash tun very quickly..... I do not have a graph showing the mash tun center temp vs the RIMS, but I can assure you it is within a tenth of a degree or two... Here is the graph I do have, I promise that all my brewing friends who watch me brew can attest that the Mash temp gets to the RIMS temp within a few minutes...

 

[crane]

I find that if you have a good grain bed, and recirculate 150 degree wort on top of it via a PID controlled RIMS, it only takes a few minutes to get the whole thing to that temp... the temperature gradient moves down the mash tun very quickly..... I do not have a graph showing the mash tun center temp vs the RIMS, but I can assure you it is within a tenth of a degree or two... Here is the graph I do have, I promise that all my brewing friends who watch me brew can attest that the Mash temp gets to the RIMS temp within a few minutes...

I agree. I have one sensor in the middle of the grain bed and one at the output of the RIMS tube. Once the sensor in the RIMS tube reaches the setpoint it's only a few minutes until the sensor in the mash tun gets there as well.

 

[atoughram]

You all are going to force me to switch to a BCS system.... Nice graphs....

 

[branman]

You all are going to force me to switch to a BCS system.... Nice graphs....

I was thinking the same thing when I saw that...

 

[branman]

Just an update, I put a thermometer in the grain bed a temp sensor at the outlet of the mlt, and a sensor at the pump outlet. Both sensors read the same and the mash went great. The temps did not spike. They never got more than 2 degrees off +/- thanx again for all of the help

 

[Tytonegro]

Maybe I'm lacking detailed knowledge of a RIMS system, but what is (in) the RIMS tube?

Circulating the wort through the grain bed IS a RIMS system or am I missing something...

Anyway, to prevent overshoot, I would place the sensor closer to the output of the pot towards the pump. That is close to the heat source. Temperature in the grain bed will then rise slower, but when circulation is fast enough it will definately get there without much overshoot.

 

[crane]

Maybe I'm lacking detailed knowledge of a RIMS system, but what is (in) the RIMS tube?

Circulating the wort through the grain bed IS a RIMS system or am I missing something...

Anyway, to prevent overshoot, I would place the sensor closer to the output of the pot towards the pump. That is close to the heat source. Temperature in the grain bed will then rise slower, but when circulation is fast enough it will definately get there without much overshoot.

The RIMS tube holds an electric heating element in it. So placing the temp sensor at the output of the RIMS tube will be closest to the heat source and it is the best location to prevent overshoot.

If this were a direct fired RIMS system (burner under mash tun) then placing the temp sensor at the output of the mash tun or underneath the false bottom would be the best choice.

 

[Tytonegro]

Thanks crane, I was confused by the picture that shows a flame under the pot,

Then I agree with you about the sensor position...

 

[Jakeintoledo]

Sorry, man. The RIMS tube is applying heat to the RIMS. You're right; my schematic for the logic calls the tube the "RIMSsystem" so i'm used to calling the heating portion of the system the 'RIMS".

In a couple of weeks, I'll have an opportunity to move the sensor position to the heating point.

I appreciate everyone's input and hope to be making better, consistent beer soon.

Maybe I'm lacking detailed knowledge of a RIMS system, but what is (in) the RIMS tube?

Circulating the wort through the grain bed IS a RIMS system or am I missing something...

Anyway, to prevent overshoot, I would place the sensor closer to the output of the pot towards the pump. That is close to the heat source. Temperature in the grain bed will then rise slower, but when circulation is fast enough it will definately get there without much overshoot.