Newest Threads

**kal** · 11-15-2011, 04:08 PM

What scares me about heat sticks is the non-attached/non-permanent nature of them. There's a lot of power in there that is often just loosely resting over the edge of a kettle.

IMHO something like that needs to be properly locked down so that it doesn't move when in use. But maybe that's just me. I'm afraid that someone's going to catch an arm or a leg on the cord and pull the thing out of the kettle and hurt themselves, start a fire, who knows. Or move it and bump the element portion into something they're not supposed to, and so on. It's the things you don't expect to do that just happen sometimes that makes me want to lock anything that powerful down so that nothing can possibly happen.

Some people do have heat sticks that once installed basically don't move. They're pseudo-permanent. But then the cost is likely more than what I built so in most cases it makes more sense to just permanently install them (IMHO). No right or wrong answer here of course. I'm just giving my opinions.

Grounding or GFI has nothing to with it. That's like wearing your seat belt in your car. It's insurance or "just in case" that can help save you if a problem happens, but you should always design a safe system to begin with - not rely on the safety net.

Just like your car's seat belt, proper grounding and GFI is not a guarantee of 100% safety. It's simply added safety that *may* help if something goes wrong. You don't drive like a maniac running red lights and taking corners on two wheels just because you're wear a seat-belt. Same goes with electric brewing: You don't do proper grounding and use a GFI and then purposely do something unsafe because you think you'll be ok. There's absolutely no guarantee of any sort. It's just added (extra) protection that may help if something goes wrong.

A properly designed heat stick shouldn't leak any more than a properly designed permanently installed element. The idea that a heat stick will or may eventually leak means it's poorly designed to begin with.

I would also go gas or electric but not both on the same kettle. Even with the heat stick wire going over the top into the kettle, there's a TON of heat being produced and coming up the sides of the kettle from the massive gas burner.

This is getting off topic and doesn't really have anything to do with this thread involving building a control panel.

(You're in the design phase for your setup). Maybe start a new build thread called "Can I combine gas & electric using heat sticks?". You'll likely get more responses that way too as it's not buried on page 49 of this control panel thread.

Kal

**P-J** · 11-15-2011, 04:23 PM

Kal,

Very, very well said.
Your coments and observations are right on and important.

Thank you.

P-J

**joety** · 11-16-2011, 02:33 AM

I think my RIMS system offers a good compromise between gas and electric. The gas brings all the kettles up to temp, and you can maintain by circulating through the RIMS tube. My control panel is designed off of Kal's setup, it just has one PID vs three. The RIMS tube is grounded and sealed the exact same way Kal's elements are.

**JoshuaWhite5522** · 11-16-2011, 03:28 AM

Kal you're right, I apologise for the drift.

**Namako** · 11-21-2011, 01:06 AM

Maybe I'm missing something, but one of the REAL values of using a PID is that it "learns" your system - it "learns" how much of a time lag there is between it turning on the heat and it "seeing" the results of it turning on that heat. It then "learns" how long it takes for the temperature to stabilize after turning off the heat, and based on how far it overshot the target temperature, it adjusts it's algorithm and the next time it turns off the heat sooner.

But that algorithm only works if the system it is "learning" is stable. Any variables that you incorporate into the system will throw the algorithm being developing into a electronic tailspin.

I haven't yet built my control box (all the parts are in - almost), but my system design is different.

I will have separate PIDs to maintain my HLT, MT, and BK - I sometimes use my BK to pre-heat my sparge/HERMS water while I'm striking in from the HLT, so that when I move the pre-heated water over to the HLT, I want to be able to go immediately to recirculating and mash temp correction after striking in, if I miss my target. I'll have my RTDs at the in-flow (e.g., top) of each kettle, vice having the RTDs in the mass of the water/mash, because the volume and thermal mass changes with each batch (and during a batch) and therefore the amount of heat needed to nudge the temperature up ONE degree changes (especially between a 15 lb. mash and a 40 lb. mash), thereby messing up the algorithm. If I target 150 degrees at the in-flow, then the PID will ensure that the water/wort returning to the kettle is going to be 150 degrees, and my continuously recirculating the water/mash will result in the water/mash temperature to stabilize at 150 degrees. The PID "learns" how much heat was needed to be applied to the volume in kettle in order to give me 150 degrees at the inflow to the kettle (the volume of fluid in my hoses is constant, as is the heat loss), after it starts low (which is why I need to apply heat) at the out-flow (bottom) of the kettle, and then the water/wort looses more temperature flowing through the pumps and hosing on its way back to the kettle.

Using one PID to do everything relegates the lone PID to being a simple temperature controller - all it will know is that the measured temperature is lower than the set temperature, so power will be applied. Once it sees that it hit the set temperature, it turns off. Any subsequent overshoot is beyond its control - it takes several minutes (as much as 15 minutes in a large mash) for the temperature to stop rising. That means YOU have to decide how much lower to set your temperature (below your REAL temperature), and now YOU are performing the work of the algorithm.

I would sure like to hear if someone sees a flaw in my system logic...

shortyjacobs · 11-21-2011, 02:06 AM

Quote:

Originally Posted by Namako

Maybe I'm missing something, but one of the REAL values of using a PID is that it "learns" your system - it "learns" how much of a time lag there is between it turning on the heat and it "seeing" the results of it turning on that heat. It then "learns" how long it takes for the temperature to stabilize after turning off the heat, and based on how far it overshot the target temperature, it adjusts it's algorithm and the next time it turns off the heat sooner.

But that algorithm only works if the system it is "learning" is stable. Any variables that you incorporate into the system will throw the algorithm being developing into a electronic tailspin.

I haven't yet built my control box (all the parts are in - almost), but my system design is different.

I will have separate PIDs to maintain my HLT, MT, and BK - I sometimes use my BK to pre-heat my sparge/HERMS water while I'm striking in from the HLT, so that when I move the pre-heated water over to the HLT, I want to be able to go immediately to recirculating and mash temp correction after striking in, if I miss my target. I'll have my RTDs at the in-flow (e.g., top) of each kettle, vice having the RTDs in the mass of the water/mash, because the volume and thermal mass changes with each batch (and during a batch) and therefore the amount of heat needed to nudge the temperature up ONE degree changes (especially between a 15 lb. mash and a 40 lb. mash), thereby messing up the algorithm. If I target 150 degrees at the in-flow, then the PID will ensure that the water/wort returning to the kettle is going to be 150 degrees, and my continuously recirculating the water/mash will result in the water/mash temperature to stabilize at 150 degrees. The PID "learns" how much heat was needed to be applied to the volume in kettle in order to give me 150 degrees at the inflow to the kettle (the volume of fluid in my hoses is constant, as is the heat loss), after it starts low (which is why I need to apply heat) at the out-flow (bottom) of the kettle, and then the water/wort looses more temperature flowing through the pumps and hosing on its way back to the kettle.

Using one PID to do everything relegates the lone PID to being a simple temperature controller - all it will know is that the measured temperature is lower than the set temperature, so power will be applied. Once it sees that it hit the set temperature, it turns off. Any subsequent overshoot is beyond its control - it takes several minutes (as much as 15 minutes in a large mash) for the temperature to stop rising. That means YOU have to decide how much lower to set your temperature (below your REAL temperature), and now YOU are performing the work of the algorithm.

I would sure like to hear if someone sees a flaw in my system logic...

Close. PIDs CAN learn, if in "autotune" mode...but they aren't continually learning.

PIDs have three main variables, P - proportional, basically, the farther away I am from my set point, the more power I'll give the element. I - integral, basically, the longer I am away from my set point, (the more "area under the curve"), the more power I'll give the element. D - derivative, basically, I watch how the current temperature is changing. If it's changing FAST, and approaching the setpoint, I'll back off on power.

You can set the coefficient variables for these three terms yourself, or you can tell the PID to learn the best settings. When you do this, you set up your system as close as you can to how it will be running while brewing, set a set point temp near where your normal temps will be, and tell it to "Autotune". It'll goose the power, and intentionally overshoot the temp. Once it passes the setpoint, it kills power, and observes how long it takes for temp to drop back down. From how fast the temp rises and then drops, it can calculate how it should set the P, I, and D variables.

But then it's done. No more learning, unless you tell it to autotune again. You can't "confuse" a PID.

SO, if you are using one PID to control multiple systems, you should probably autotune it on the system that requires the MOST temp control, (however you're controlling mash recirc temp, usually), and then not worry if it's a bit sloppy about the other systems' control.

**kal** · 11-21-2011, 01:12 PM

Quote:

Originally Posted by shortyjacobs

SO, if you are using one PID to control multiple systems, you should probably autotune it on the system that requires the MOST temp control, (however you're controlling mash recirc temp, usually), and then not worry if it's a bit sloppy about the other systems' control.

Correct. In my control panel I use separate PIDs. The boil PID is only used for boiling in manual so it's really not used like a PID at all.

The mash PID actually only displays the mash temp. That PID does nothing.

The HLT PID on the output of the HLT is what actually controls the mash temp. This is the critical PID that needs to be running right. It'll fire the HLT element to keep the temp consistent so that heat is exchanged into the wort from the MLT through the HERMS coil in the HLT.

Kal

michaeltrego · 11-21-2011, 01:31 PM

Quote:

Originally Posted by kal

The boil PID is only used for boiling in manual so it's really not used like a PID at all.

The mash PID actually only displays the mash temp. That PID does nothing.
Kal

Since you aren't using the PID functionality, is there a cheaper solution for these two temp displays?

**kal** · 11-21-2011, 01:38 PM

Quote:

Originally Posted by michaeltrego

Since you aren't using the PID functionality, is there a cheaper solution for these two temp displays?

Possibly. I don't know. You'd need something that can read an RTD temp, probe display the temp, allow for temp offset (calibration),, and support alarming. The PID's only about $40 so I didn't look around at all because the cost savings would have been minimal if any especially when you consider the total cost of the control panel.

Even if I had seen something that would have been usable, I like having 3 identical PIDs for aesthetic reasons and to also have a backup.

Kal

**joety** · 11-22-2011, 03:02 AM

Quote:

Originally Posted by michaeltrego

Since you aren't using the PID functionality, is there a cheaper solution for these two temp displays?

They are dirt cheap as is. I do need to autotune mine to my mash, however, this has been a good reminder. It is overshooting by three degrees now and I've been manually correcting.