Thermowell Lag Time . . . Heatsink?

[AnOldUR]

I use a 16" thermowell in an direct heat electric MLT that comes down vertically from the top. The problem is the lag time between the actual mash temperature and what the probe inside the thermowell reads. My guess is that this is caused by the large amount of surface area on the thermowell that is exposed to the head space and not the actual mash.

There are two things that I'm thinking of trying to help. First is to cover the upper part of the thermowell with tubing to insulate it, and second, to put a heatsink over the bottom of the thermowell to increase its surface area within the mash.

Has anyone tried either of these or have other suggestions for decreasing the lag time? Is there a commericially available heatsink made for this purpose that will fit over the 3/8" OD of the thermowell?

 

[TipsyDragon]

how far down inside the thermowell do you shove the probe?

if its all the way down at the bottom i would think the lag time would be minimal and the amount of thermowell sticking out of the wort wouldn't be a factor.

 

[AnOldUR]

how far down inside the thermowell do you shove the probe?

It's all the way down to the bottom, but the lag is there. If I do an infusion to get a quick jump in step temperatures the lag is most noticeable. 5 minutes or more to equalize out after a 30 degree infusion step.

It's a 15 gallon MLT, so with a 5 gallon batch there's a lot of head space.

 

[TipsyDragon]

it makes sense that it would take time for the water to heat the metal of the thermowell. so i believe there is a lag. just curious how are you verifying that there is indeed a lag time?

personally i would put a calibrated thermometer in the mash right next to where the probe is in the thermowell and see if the temperatures match up or not and how long it takes them to match up if they are different.

 

[AnOldUR]

just curious how are you verifying that there is indeed a lag time?

Pretty much testing it the way you said. I've got a couple of calibrated digital thermometer that read one thing while the thermowell reads a lower number until it catches up. A little lag is expected, but this is enough that overshooting temperatures becomes a problem.

Not having any luck finding a heatsink. May have to make something.

 

[Catt22]

What material is the thermowell made from? SS doesn't conduct heat very well. Copper would be much better for a fast response time. Five minutes does seem excessive even for SS though. I use a copper thermowell and there is almost zero lag time. What type of probe are you using in the thermowell? With either an RTD thermistor type or a thermocouple type probe, it's my understanding that the temperature is measured right at the tip, so the head space in the thermowell shouldn't be an issue. The tip of the probe should be in direct contact with the bottom of the thermowell. IIRC, there's a type of grease or gel that you can put into the thermowell that will improve the contact. I've never used the grease (or whatever it is) so I can't help with more info on it. I think it's the same stuff used between computer chips and heat sinks. I suspect the probe is not making good contact with the thermowell bottom and that is contributing to the lag. It would take a few minutes for the air to warm up to match the mash temp I would think. I could be completely wrong on this guess. Just something to think about.

 

[AnOldUR]

Yeah, it is a stainless thermowell. I didn’t even think about this making a difference when I bought it, but copper would have been the way to go. Not sure of the exact type of probe. It’s what ever comes standard on a Ranco. Thermal grease sounds promising. I’ll have to look into something that will be compatible with the probe.

Still like the heatsink idea. I have a piece of 1-1/4" diameter copper. I want to cut deep grooves in it and put a 3/8" hole through to slip over the stainless. What I don’t know is how well the copper will transmit the heat into the stainless.

 

[TipsyDragon]

What I don’t know is how well the copper will transmit the heat into the stainless.[/SIZE]

thats just what i was thinking. your going to have to maximize the amount of direct metal to metal contact between the copper and SS. you may have to solder it together some how. also your going to want to maximize the amount of surface area of the copper that is in contact with the wort.

you might be able to make your own thermowell out of copper. you just need to get a pipe thats the same ID and OD as the one you got and solder one end shut with plumbers solder (i think). the only tricky part would be that flare at the open end.

 

[SpanishCastleAle]

That does seem way excessive for one of those thermowells. I use one of those exact same ones in a carboy and it equalizes plenty fast but I don't have near the temp swings you do in a mash.

I'm not sure adding more thermal mass is the way to go. Then there would be even more mass that must change temp in order for the probe to read the correct temp. But I haven't actually seen the exact set-up you have. Is there any way you can get better contact between the well and the wort?

 

[Photomask]

You could try using a thermally conductive paste between the thermowell and the probe. It's usually a white stuff that looks much like toothpaste. It's normally used between the heat-sink and SSRs, power transistors, SCRs etc.... Most electrical suppliers carry it, including Radio Shack.

 

[AnOldUR]

Is there any way you can get better contact between the well and the wort?

Other than stuffing the probe down the tube the only thing I can think of is the thermal grease mentioned earlier.

Looked up the thermal conductivity of material this morning (stainless 12 to 45, copper 401.) This tells me that copper is the way to go, but the transfer of heat from putting the copper over the stainless might not be ideal. Here's the new plan is:

Cut the thermowell so that it’s open at the end
Run a 3/8-24 die over it
Turn fins in the copper like on one of those things you stick in honey
Drill and tap the copper (not through)
Use Teflon tape to seal the copper to the stainless

With the base of the thermowell being copper and the upper being stainless, I should have the best of both worlds. Quick reaction time at the probe, but minimal loss of heat through the stainless that's above the mash.

 

[Sawdustguy]

I don't have a solution over what has been mentioned already but I am curious why a 16" thermowell? Are you placing the thermowell in the vessel from the top or through a bulkhead?

 

[AnOldUR]

Are you placing the thermowell in the vessel from the top or through a bulkhead?

It's down from the top. The idea was to prevent the stirring problems you can get with a horizontal probe.

 

[SpanishCastleAle]

Other than stuffing the probe down the tube the only thing I can think of is the thermal grease mentioned earlier.

I assumed the probe was inserted all the way. Is there a reason you aren't inserting it all the way so it's down at the bottom?

Once again, I haven't seen your set-up. I use the display and thermocouple from one of those heat-resistant meat probes you can get for cheap. I just pulled the wire out of the stock probe and stick it in the thermowell, to the bottom though.

 

[TipsyDragon]

I'm not sure adding more thermal mass is the way to go. Then there would be even more mass that must change temp in order for the probe to read the correct temp.

its not about mass its about surface area. the added copper could be paper thin and add minimal mass to the probe. besides with coper being about 9 times more heat conductive than SS it wouldn't take that long at all to heat up.

 

[AnOldUR]

I assumed the probe was inserted all the way. Is there a reason you aren't inserting it all the way so it's down at the bottom?

Bad choice of words on my part. I the probe is at the base of the thermowell.

 

[AiredAle]

My guess would be you have poor thermal contact between the theromcouple and the thermowell. Adding a copper heatsink in this case would do nothing to help.

Here's a half-a$$ed thought for you: If your thermocouple is liquid-proof, you could test this without spending money on the thermal paste. add some olive oil to the thermowell to improve the thermal contact between the t-couple and the t-well. If poor thermal contact is the problem, this will reduce your lag time, because the oil has a higher thermal conductivity than the air that is now in the t-well (0.17 for olive oil vs. 0.024 for air). If this helps, then try a thermal paste (they have conductivities in the 2 to 7 range); this will improve the lag time even more than the oil. If the t-couple isn't liquid proof, then the paste may be a problem anyway, because they consist of silicone oil and some powdered conductor.

If this doesn't have any effect, then change to a copper thermowell, and save yourself the effort of making a copper heatsink.

 

[AnOldUR]

Finished making a copper tip for the stainless thermowell.
Hope to get a chance to try it out this weekend.

 

[SpanishCastleAle]

Looks great. How's it perform?

 

[AnOldUR]

Looks great. How's it perform?

Brewed with it yesterday (American Wheat.) Big improvement. Still a little bit of lag, but now it's a workable situation where I don't have to worry about overshooting step temperatures. The only thing I would do different is make the copper tip a little longer to better accommodate the length of the probe portion of the Ranco's external wire. But I'm not sure if the temperature is read over the entire length of the probe or just the very tip?

 

[chucklebee]

If you want to find out the characteristics of your heating loop then make your thermowell with your rtd/thermocouple setup and put in ice bath with water. Wait until it reaches 0 deg celcius. Have a pot of water boiling and put it in.

Immediately move the setup in to the hot boiling water and measure the amount of time it takes to reach 63.2 deg celcius. That is one Tau. Accepted in industry for final temperature is 5tau. 5*the amount of time it took to get to 63.2 deg celcius which results to 99.8% of your actual temperature.

If you want to speed up the response time of your temperature loop, use either a thinner thermowell. In my experience I have used 1/4" SS tubing or Ti. Whatever metal that is compatible with your liquid. Best result is used with spring loaded rtd/thermocouple element that fit in to your thermowell. Both items can be found ~100-150 dollars on omega.com

Thermal paste when put liberally as a glob is worse than metal to metal contact. In a pinch (when you don't have spring loaded thermocouple/rtd elements) what I have used when I used thermocouple was using type K thermocouple wire. twisted it together. Put a small piece of silver solder down the thermowell. Then I put the thermocouple wire inside and then heat the thermowell on the bottom with a butane torch. (2 man operation btw) One person heats, the other person pushes in the wire. When the silver solder melts, the thermocouple wire slips right in.

I've done that at home where I made my own meat thermometer with detachable thermocouple junctions.

I speak only from my own humble 3 year experience in the instrumentation field in the mining industry.

Brewed with it yesterday (American Wheat.) Big improvement. Still a little bit of lag, but now it's a workable situation where I don't have to worry about overshooting step temperatures. The only thing I would do different is make the copper tip a little longer to better accommodate the length of the probe portion of the Ranco's external wire. But I'm not sure if the temperature is read over the entire length of the probe or just the very tip?

 

[Swilldog]

Sorry I'm trying to get a mental picture of what you doing. Are you using the ranco to control a heating element in your mash tun?

 

[Swilldog]

Make copper shim out sheet copper, wrap around the probe and stuff down the thermowell. It will help against the intermittent contact between probe and thermowell.