Help with PID settings in Elsinore

[TechyDork]

Did my first brew session on my new E-HERMS system over the weekend and I have a few questions that I am hoping someone can help me with.

First I followed the steps here and plugged them in to the calculator to find the initial values for my PID in the Elsinore. My values were P=14 I=400 D=30 and a Duty Time of 10s. Using these values and a set point of 154 in the HLT the element would be at 100% until the temp was in the 140's then it would drop to 14-20% until closer to temp.

This lead to really long ramp times between temp changes so I changed P to 40 and then the element would stay at 100% until within a degree or two of the set point. This did lead to some minor overshooting of the set point, but nothing too bad. Maybe 1-2 degrees over in the HLT.

I am using 20g SS Bayou Classic kettles for the HLT and MLT with a 5500w element in the HLT with a 50' 1/2" SS coil in the HLT. The temp probe is in the out flow of the HLT and the water in the HLT was continuously recirculated during heating.

Do my PID values look way off? When i used P=40 the ramp times were faster and the temp stayed within +-2 of the set point. I would like to tighten that up to +-.5 if possible.

What do most of you use for a Duty Time when using PID control? How about when using manual control of the element?

My next question is about temp difference between the MLT and the HLT. I noticed that they never really equalized on my system during the 60 min mash. The MLT was consistently 2-4 degrees lower than the HLT set point.

The temp probe for the MLT is also in the outflow of the MLT and the mash was continuously recirculated during the mash. I ran both the water and the wort pumps full open during the mash.

Is it normal to have this big of a temp difference between the HLT set point and the mash temp?

I know we can just compensate and set the HLT to 4 degrees over mash temp, but from my reading on theelectricbreweery.com I was under the impression that the temps would equalize.

sorry for the long post, but i am trying to give as much info as i can

thanks,

 

[TechyDork]

Also since this is a software PID could I set both the HLT and the MLT probe to control the element in the HLT? Then i could use the HLT probe when heating the strike water in the HLT and then switch to the MLT probe to maintain mash temps. I have not tried this so i don't know if SBE will allow you to set two PIDs to control the same GPIO.

That would eliminate the need to manually adjust the set point in the HLT to maintain my mash temp as the temp of the wort coming out of the MLT would determine firing of the element in the HLT.

 

[TechyDork]

Anyone?....

 

[TechyDork]

@jangevaa

How do my PID settings look? Any idea how i can improve how tightly the PID will maintain temp?

 

[jangevaa]

@jangevaa

How do my PID settings look? Any idea how i can improve how tightly the PID will maintain temp?

The killer of getting a fast acting PID tuning with this method is dead time. What was your deadtime?

You could just up your P to around 30 or so...

I have stopped recommending people to use this tuning method. It takes a very long time to do and the resulting tunings I don't think are as good as they could be. I'm planning on putting together a tutorial of the Ziegler-Nichols method in the near future.

From what I understand the idea is to set I and D to 0 initially, set a target temperature, and raise the P term until the maximum oscillation is achieved. At that point, the I and D terms can be calculated from characteristics of this oscillation.

 

[bisen4]

To help with your set point temp problem, if your PID has the option, most all do, set the Pb setting to +4. This is essentially a temperature calibration offset. So for another example, if your mash temps were coming out to +3 degrees higher than what you set the temp as, you would set the Pb setting to -3 to compensate for the higher than expected reading. hope this helps. cheers!

 

[TechyDork]

The killer of getting a fast acting PID tuning with this method is dead time. What was your deadtime?



You could just up your P to around 30 or so...



I have stopped recommending people to use this tuning method. It takes a very long time to do and the resulting tunings I don't think are as good as they could be. I'm planning on putting together a tutorial of the Ziegler-Nichols method in the near future.



From what I understand the idea is to set I and D to 0 initially, set a target temperature, and raise the P term until the maximum oscillation is achieved. At that point, the I and D terms can be calculated from characteristics of this oscillation.

I don't have my notes in front of me but I think my dead time was 60 seconds before the temp began to rise.

Do you have a link to info on this other tuning method?

I did also set D to 0 and P to 40 leaving I at the value determined by your tuning method. I will test again setting I to 0 and see how that works.

To help with your set point temp problem, if your PID has the option, most all do, set the Pb setting to +4. This is essentially a temperature calibration offset. So for another example, if your mash temps were coming out to +3 degrees higher than what you set the temp as, you would set the Pb setting to -3 to compensate for the higher than expected reading. hope this helps. cheers!

I don't think the reading of the temp probe is off. I double checked it with my thermometer and it was within 1f of the reading on the PID therm.

I just observed that it took longer for the MLT temp to stabilize than I expected, 10+ minutes, and when it did it settled about 3-4 degrees below the temp in HLT.

I was under the impression that they would equalize. Allowing me to set the HLT to my mash temp. If it is normal to need the HLT set 3-4 degrees warmer than I can just bump the temp on the HLT to maintain my mash temp.

I was running both the water and wort recirculating pumps wide open. Should I slow the flow of the wort through the HERMS coil? It is a 1/2" 50' coil.

Sorry for all the questions. I am just trying to tighten up my temp control of the mash as much as possible.

Maybe I just need a few more batches to get a better feel for it. I just want to be sure I am not missing something.

 

[FuzzeWuzze]

I don't have my notes in front of me but I think my dead time was 60 seconds before the temp began to rise.

Do you have a link to info on this other tuning method?

I did also set D to 0 and P to 40 leaving I at the value determined by your tuning method. I will test again setting I to 0 and see how that works.




I don't think the reading of the temp probe is off. I double checked it with my thermometer and it was within 1f of the reading on the PID therm.

I just observed that it took longer for the MLT temp to stabilize than I expected, 10+ minutes, and when it did it settled about 3-4 degrees below the temp in HLT.

I was under the impression that they would equalize. Allowing me to set the HLT to my mash temp. If it is normal to need the HLT set 3-4 degrees warmer than I can just bump the temp on the HLT to maintain my mash temp.

I was running both the water and wort recirculating pumps wide open. Should I slow the flow of the wort through the HERMS coil? It is a 1/2" 50' coil.

Sorry for all the questions. I am just trying to tighten up my temp control of the mash as much as possible.

Maybe I just need a few more batches to get a better feel for it. I just want to be sure I am not missing something.

Everything ive read is that the temps between the HLT liquid out and MLT out will always be different, you just need to figure out what that delta is and always use it. Need it at 154? Set your HLT to 157 knowing you'll drop 3 degree's. What you probably can influence though is how consistent that 3 degree's is, and thats through the PID tuning like your doing. If its fluctuating 1-4 degree's higher then its obviously a problem.

 

[TechyDork]

Everything ive read is that the temps between the HLT liquid out and MLT out will always be different, you just need to figure out what that delta is and always use it. Need it at 154? Set your HLT to 157 knowing you'll drop 3 degree's. What you probably can influence though is how consistent that 3 degree's is, and thats through the PID tuning like your doing. If its fluctuating 1-4 degree's higher then its obviously a problem.

That makes sense. Maybe I miss read that they should equalize. I don't mind just setting the HLT to 3 degrees higher, but I would like to tighten up the PID to maintain the HLT temp as close to set point as possible.

 

[jangevaa]

Yes, there is surely to be some heat loss from the HLT to Mash in a HERMS set up. 1-3F is typical and will depend on ambient temperate and whatnot. You can also try to tune your PID with respect to your mash tun. For this to work well you need to have very consistent recirculation conditions however.

Do you have recirculation in your HLT? Decreasing the dead time would be the best way to have faster control.

 

[TechyDork]

Yes, there is surely to be some heat loss from the HLT to Mash in a HERMS set up. 1-3F is typical and will depend on ambient temperate and whatnot. You can also try to tune your PID with respect to your mash tun. For this to work well you need to have very consistent recirculation conditions however.

Do you have recirculation in your HLT? Decreasing the dead time would be the best way to have faster control.

I do recirc both the HLT water and the wort in the MLT.

Sounds like i need to focus on nailing down the temp difference between the HLT temp point and the wort temp in the MLT.