Q:Why put temp probe in RIMS tube?

[Wizard_of_Frobozz]

But why use two when one hold exact temps fine?

If your only goal is to maintain temps then I agree with you. I'm a big believer in KISS. However, the OP stated their goal was to maximize ramp times.

How you measure performance depends on what your objective is: a Ferrari is the "best" car if you want to hit zero to 60 in 4 seconds. A Prius is "best" if you want maximum fuel efficiency and range (just an example - not trying to turn this into a car debate!)

The industrial control systems I've worked with do exactly what the OP is looking for by using a PID loop and one or more probes but they can get very complicated. I would think you would need some sort of mixer in your Mash tun to get the higher temperature from the RIMS tube to mix quicker than just recirculating in order to avoid overshooting your rest temps in the mash tun. Might not be worth the extra complexity but that depends on how important the ramp times are to the OP.

 

[augiedoggy]

fair enough,
But I would just use a larger rims with a faster flow rate if I wanted faster temp ramps my self because I do think putting the temp probe elsewhere will cause temp overshoots. while it would be better at ramping it wouldnt be as good at maintaining and it wouldnt be as stable while ramping I believe. a person could make a dual probe temp controlller like the kraken and figure out some way to make it work im sure but then I think we are off topic from the OP's original question.

 

[Wizard_of_Frobozz]

Yeah, I think we've answered the OP's question pretty thoroughly now: Precise temp control and avoiding overshoots is why the probe is usually placed at the RIMS tube outlet.

The engineering geek in me likes trying to figure out how to meet the goal of rapid ramp times within the limitations that the OP gave (specifically using 120v, which in turn, limits heater size). I'm with you - I'd probably go for two heaters on different circuits and a larger pump before trying to add the complexity of rapid mixing of hotter wort with the mash tun.

I would say this kind of debate is what leads to new innovations within our community, and it's one of the main reasons I stopped lurking and started contributing here.

 

[Jwin]

For all the reasons we've discussed on this thread. Optimized ramping and the ability to control the temp of the grain bed directly.

More so even, just optimization of the ramping. If monitoring the grain bed helps with that, then so be it. I'm not worried about denaturing. The flow rate required will avoid overheating and the mass of wort should avoid significant denaturing. Again, I have experimented with recirculation at full bore pump speed. Works fine. Remember also, I am gyle brewing, so we're talking pretty thin mashes here.

Now, all this could be in vain. With significant flow, and proper reintroduction of the wort in the tun, the controller may work fine.
But adding 3 wires and a single stc-PID as a loop between the existing PID and SSR would be pretty easy.

Now, as far as the debate on overshooting, BIABers basically do what I any to do with a larger element, no? Monitor the mash directly while recirculating and heating...

Now as far as the debate of temp probe inlet v outlet, I have seen RIMS tubes mounted both horizontal and vertical. Perhaps that is the origin of different opinions?

To further p*ss on the fire, what would be the downside of pulling wort through the rims over pushing it through. seems that could be safer, with proper implemention. May need a bleeder valve.
Now, one could argue that biab has better though flow than a cooler tun.

Thank you to everyone for their replies. Who knows, maybe we'll figure out a better way
Advancement comes though debate.

 

[Wizard_of_Frobozz]

To further p*ss on the fire, what would be the downside of pulling wort through the rims over pushing it through. seems that could be safer, with proper implemention.

You want to minimize any flow restrictions on the inlet side of your pump. A pump needs a minimum "Net Positive Suction Head", which is a fancy way of saying it needs a minimum pressure at it's inlet. Any flow restrictions (such as a RIMS tube) on the suction side of the pump reduces the pressure into the pump suction. This in turn, leads to cavitation, which causes flow reduction and eventually damage to your pump. Cavitation is basically boiling inside the pump because the pressure in the pump impeller drops below the boiling point of the liquid. Hotter liquids are closer to their boiling points, so it's a very real concern with our setups.

 

[Jwin]

So, my original question has been answered. I was in assumption that the most prominent reason for the probe in the tube is mostly safety. Now, it is utilized also to regulate temperature. I was just reading a white page on PID loops. I don't think that the RIMS probe is the best placement for temperature regulation.
There are PIDs out there that accept multiple probes, but hey are quite costly. RPi would probably be the most cost effective way to do this, but I have no experience with them. So, I think I'll look into designing a system with 3 controllers in a loop.
The probe in the RIMS acting as a safety to one controller. Two probes in the mash. Essentially running the two mash pids in parallel, followed by a PID in series to them(rims safety), then the SSR. Probe placement would be pretty critical. I think one in the mash and one on the MT output would be sound. That way, if channeling occurs, it will trip the MT output and let the mass of the mash catch up.
The rims PID could be set to ~10° below boiling, and then lowered once the target step temperature is close.
Now, all this is based on the assumption that PIDs have a resistance on their output, and seeing live DC voltage on the output wouldn't damage them. Otherwise, I guess I could wire a resistor into the picture.
Any opinions on this? I'm sure a diagram would be better, but I'm about to head to work and don't have the time...

Thanks all.

 

[augiedoggy]

theoretically having the probe at the inlet of the mt would just allow for very little overshooting and you may very likely get closer MT temps to the setpoint... The other thing to consider here is with a 120v element the reality is you would get better performance and more utilization out of it if you circulate SLOWER to give it more contact time to heat the wort per pass... This is a reason I used a long narrow rims tube with a long element... the wort exiting my rims IS fully and eaqually heated to the desired setpoint I set my temp co0ntroller at... the only practical realized variable here is the time delay between the wort exiting the rims and the time it takes to make it through the grainbed long enough to ramp it up to the desired temp..

I would argue that a 5500w rims flowing at 5gpm still wont work as well as my 42" long rims with a 36" long 1800w element and 1/4" space all the way around the element between it and the rims tube wall.. Just my opinion but I feel I have taken the possibility of scorching out of the equation as well as denaturing enzymes (if it is a real concern) as well as eliminated any safety concerns.. I get the advantages of a herms with the advantages of a rims. I can still increase my mash temps quite quickly for step mashing when brewing an 11 gallon batch

 

[Wizard_of_Frobozz]

I would argue that a 5500w rims flowing at 5gpm still wont work as well as my 42" long rims with a 36" long 1800w element and 1/4" space all the way around the element between it and the rims tube wall

Augie, where did you find that 1800W element? That's a pretty unusual element. I completely agree with you on contact time - that's something a lot of people miss when designing systems. Just because you have 5500W on your element doesn't mean you are actually transferring all of that heat into the the wort. Heat exchange takes time. Dwell time is a very important factor in heat exchanger performance but it gets missed because the normal heat transfer equation (Q=mCp∆T) doesn't involve time; it assumes 100% transfer. If you put 5500 watts into the heater, and there isn't sufficient flow to remove 100% of that energy, the element will heat up to push up that ∆T, and potentially scorch your wort.

 

[thekraken]

Now as far as the debate of temp probe inlet v outlet, I have seen RIMS tubes mounted both horizontal and vertical. Perhaps that is the origin of different opinions?

I was more under the impression that horizontal vs vertical is about either convenience of mounting or mounting such that you don't get air bubbles in the tube. Either way the probe is typically on the output.

So, my original question has been answered. I was in assumption that the most prominent reason for the probe in the tube is mostly safety. Now, it is utilized also to regulate temperature.

Ehhh I still feel if you are going to use an off the shelf PID controller it is the best place to put the probe at the output. I think it is more than an afterthought.

So, I think I'll look into designing a system with 3 controllers in a loop.
The probe in the RIMS acting as a safety to one controller. Two probes in the mash. Essentially running the two mash pids in parallel, followed by a PID in series to them(rims safety), then the SSR. Probe placement would be pretty critical. I think one in the mash and one on the MT output would be sound. That way, if channeling occurs, it will trip the MT output and let the mass of the mash catch up.
The rims PID could be set to ~10° below boiling, and then lowered once the target step temperature is close.
Now, all this is based on the assumption that PIDs have a resistance on their output, and seeing live DC voltage on the output wouldn't damage them. Otherwise, I guess I could wire a resistor into the picture.
Any opinions on this? I'm sure a diagram would be better, but I'm about to head to work and don't have the time...

Thanks all.

Woof.. that makes my head hurt! I know I'm being a complete hypocrite here but... man that sounds complicated and expensive. Just for the record I was talking about an imaginary controller that does these great things in a single neat little purpose built package. As you mentioned a rpi or microcontroller would be the cleanest DIY solution. But I understand that the software side of things isn't for everyone.

Good luck with your build I'll be curious to see how it works.

 

[mabrungard]

The only temperature that matters in a RIMS is the peak temperature of the wort downstream of the heating element. Remember, all the enzymes are in the wort...not only the grain bed. If you overheat the wort, you will prematurely denature those enzymes. Monitoring that wort temp at the RIMS tube exit is the ONLY thing that the PID needs to see.

I agree that reaching the desired temperature in the grain bed is also desirable, but there is no way around the physics and thermodynamics of the system. The only way to add heat to the system is via the wort flow and it takes time to get that fully heated wort through the bed and bring the bed to that desired equilibrium temperature. I monitor temperatures via regular electronic thermometers mounted in the mash tun inlet and outlet pipes along with the PID sensor in the RIMS tube. I can tell you that it takes several minutes for that heat wave to travel through the tun after a temperature step. I don't care that the temps in the bed aren't up to temp instantly, but my RIMS element does have enough power to bring that discharged wort to my intended temp step immediately. That's what really counts. So don't worry so much about incorporating a multi-sensor control system into a RIMS, one sensor will do.

 

[Jwin]

Augie, where did you find that 1800W element? That's a pretty unusual element. I completely agree with you on contact time - that's something a lot of people miss when designing systems. Just because you have 5500W on your element doesn't mean you are actually transferring all of that heat into the the wort. Heat exchange takes time. Dwell time is a very important factor in heat exchanger performance but it gets missed because the normal heat transfer equation (Q=mCp∆T) doesn't involve time; it assumes 100% transfer. If you put 5500 watts into the heater, and there isn't sufficient flow to remove 100% of that energy, the element will heat up to push up that ∆T, and potentially scorch your wort.

But, energy can neither be created or destroyed, correct. Short of boiling/steam, that energy has to be transferred, does it not?

 

[Jwin]

I was more under the impression that horizontal vs vertical is about either convenience of mounting or mounting such that you don't get air bubbles in the tube. Either way the probe is typically on the output.



Ehhh I still feel if you are going to use an off the shelf PID controller it is the best place to put the probe at the output. I think it is more than an afterthought.
.

I agree, but the probe would be there as a safety then a regulator.

.

Woof.. that makes my head hurt! I know I'm being a complete hypocrite here but... man that sounds complicated and expensive. Just for the record I was talking about an imaginary controller that does these great things in a single neat little purpose built package. As you mentioned a rpi or microcontroller would be the cleanest DIY solution. But I understand that the software side of things isn't for everyone.

Good luck with your build I'll be curious to see how it works.

Not that complicated. The two tun PIDs outputs (12v I'm guessing)combine then go to the through the rims PID. If rims overheats, it kills the flow. Essentially cascading them, where any one could break the flow to the SSR, but the rims PID couldn't turn on the element without getting signal from the others.
By setting the tun a couple of degrees shy of the target temp, it would hopefully prevent overshooting. Now, I have a silicon hose loop that will redistribute the post rims wort, so, in theory, it would transfer through the bed relatively evenly.
Not really expensive, just adding a couple $12 controller s and wires.
At this point, I may just go with a ezboil and see where that gets me.
If I can get what is needed out of it, I won't overengineer it. But I'll go with a bigger than needed box Incase I change my mind.

 

[Wizard_of_Frobozz]

But, energy can neither be created or destroyed, correct. Short of boiling/steam, that energy has to be transferred, does it not?

You are correct in that the energy can't be destroyed but it can be stored vs. transferred. The wort has a certain ability to transfer heat (called "heat transfer coefficient" in these applications). This is dependent on many things like surface area, flow rate, fluid temperature, materials used, etc. Overall, it's a measure of how well a heat exchanger can move heat through it. There's a whole lot of other stuff going on that affects it as well, but it's probably beyond the scope of this discussion to get into film boiling, turbulent vs. laminar flow, nucleation sites, etc. If anyone is really interested in this, here's a good resource: http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html

If you try to push more heat through it than it can handle over a time period, the heat can't transfer into the wort fast enough and the energy is stored in the heating element. So the element heats up since the heat can't leave and the sugars in your wort burn when they contact the now extremely hot surface of your heating element. The burned sugars stick to the element surface, creating a layer of carbon on the element, which slows down the heat removal even more, leading to even higher temperatures and eventually damage to your element.

This is what happens if you dry fire a heating element but it also happens if your fluid flow is too low for your heater.

 

[augiedoggy]

Augie, where did you find that 1800W element? That's a pretty unusual element. I completely agree with you on contact time - that's something a lot of people miss when designing systems. Just because you have 5500W on your element doesn't mean you are actually transferring all of that heat into the the wort. Heat exchange takes time. Dwell time is a very important factor in heat exchanger performance but it gets missed because the normal heat transfer equation (Q=mCp∆T) doesn't involve time; it assumes 100% transfer. If you put 5500 watts into the heater, and there isn't sufficient flow to remove 100% of that energy, the element will heat up to push up that ∆T, and potentially scorch your wort.

http://www.ebay.com/itm/390993852374?_trksid=p2055119.m1438.l2649&ssPageName=STRK:MEBIDX:IT

 

[augiedoggy]

nevermind I snatched that one up as a spare...

heres a similiar one..
http://www.ebay.com/itm/391266720240?_trksid=p2060353.m1438.l2649&ssPageName=STRK:MEBIDX:IT

 

[Jwin]

You are correct in that the energy can't be destroyed but it can be stored vs. transferred. The wort has a certain ability to transfer heat (called "heat transfer coefficient" in these applications). This is dependent on many things like surface area, flow rate, fluid temperature, materials used, etc. Overall, it's a measure of how well a heat exchanger can move heat through it. There's a whole lot of other stuff going on that affects it as well, but it's probably beyond the scope of this discussion to get into film boiling, turbulent vs. laminar flow, nucleation sites, etc. If anyone is really interested in this, here's a good resource: http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html

If you try to push more heat through it than it can handle over a time period, the heat can't transfer into the wort fast enough and the energy is stored in the heating element. So the element heats up since the heat can't leave and the sugars in your wort burn when they contact the now extremely hot surface of your heating element. The burned sugars stick to the element surface, creating a layer of carbon on the element, which slows down the heat removal even more, leading to even higher temperatures and eventually damage to your element.

This is what happens if you dry fire a heating element but it also happens if your fluid flow is too low for your heater.

I dont think lack of flow rate is an issue with a chugger nearly full blast. Now, I'm reading this as you are talking about to slow of a flow rate. I'm I incorrect in this?

 

[Wizard_of_Frobozz]

I dont think lack of flow rate is an issue with a chugger nearly full blast. Now, I'm reading this as you are talking about to slow of a flow rate. I'm I incorrect in this?

Low flow rate is indeed the issue with effective heat transfer. Just because your chugger is running full out doesn't mean you are getting enough flow through your RIMS. There are a whole lot of other factors that can affect it: pipe size, fluid temp, watt density of the heating element, even height of your mash tun above your pump all can affect flow rate.

If you have enough flow to ensure 100% of the heat goes into your wort, then the old Q=mCp∆T equation works and you can calculate the fastest ramp time possible if you know the total amount of mass in the mash tun + the piping system (grains and water combined). In practice, you'll never hit that ramp time, since you lose heat through the piping and mash tun (no insulation is perfect).

From a process control standpoint, you are correct that the process variable should be the thing you want to control, in this case mash temperature, not RIMS outlet temperature. A PID loop sensing actual mash temperature, in theory, will, if tuned properly, allow you to hit your target mash temperature with minimal "off" time of the heater and minimal temperature overshoot.

The real problem with it is tuning the PID loop. Let's say you tune the PID to work perfectly for your favorite 5 gallon recipe that uses 12 lbs of grain, a grist-to-water ratio of 1.25 qt/lb, and a single rest at 152°F. You set your PID parameters such that you hit your target temperatures as fast as possible and it all works great. Next brew session, you do a Dopplebock, with 20 lbs of grain, 1 qt/lb in order to fit it all in your mash tun, a protein rest at 121°F, and saccarification rest at 154°F. Your carefully selected PID parameters don't work anymore, since your flow rate, mass of grain and water, and temperatures are all different.

Using the RIMS outlet as your process control variable minimizes the affect of all those factors by keeping the delay time between output (heater) and input (temperature) as short as possible. This saves you from having to change your PID tuning for different brew sessions. Your system is simpler (One heater, one sensor), and works well for a wide variety of recipes. It's an engineering compromise - simplicity and flexibility vs. complexity and maximum ramp times.

 

[Wizard_of_Frobozz]

nevermind I snatched that one up as a spare...

heres a similiar one..
http://www.ebay.com/itm/391266720240?_trksid=p2060353.m1438.l2649&ssPageName=STRK:MEBIDX:IT

Dang it Augie! How are we supposed to build an epic system like yours if you keep buying up all the good stuff!

How did you seal the end of that heater at the end of your RIMS tube. I haven't made it through all 31 pages of your build thread...

 

[augiedoggy]

Dang it Augie! How are we supposed to build an epic system like yours if you keep buying up all the good stuff!

How did you seal the end of that heater at the end of your RIMS tube. I haven't made it through all 31 pages of your build thread...

I drilled out a compression fitting so the element passed through it and tightened it down..
Sorry I saw it was the last one and dont want to be scrambling for a replacement if it fails being I designed my rims based on its dimensions..

 

[brewman !]

I used elements just like that to heat my first boil kettle back in 1996 ! 2 of them, each plugged into a different circuit.

Back to the regularly schedule program...

 

[PlexVector]

Edit: Deleted...sorry, wrong thread...