GreenBean wrote:2. I would strongly recommend that you insulate the thermocouples, both electrically and thermally, from the roasters frame this can save much heartache later and will ensure accurate results. This can be done by:

• 2.2 Using grounded sheathed thermocouples (the thermocouple hot junction is welded to the sheath) but insulate the sheath from the roasters frame. I have been using silicone tubing to do this for well over two years and hundreds of roasts without problems.

Hi GreenBean.

First, thank you for all the efforts that you put in Roast Logger. Integration with TC4 + HTC, ingenious roast profiling, single file for log and actions, templates, among all other features, make your program by far the best roaster automation choice currently available.

I would like to thermally insulate my thermocouples. Have you already posted somewhere pictures of your installation with silicone tubing? Do you have some good source for the silicone tubing?

Thanks,

Miroslav

Hi Miroslav, thank you for the kind words and for beta testing the new hardware and software for us. I do not have any photos and do not think they would help clarify what is required. There are many ways to achieve the same result and different approaches may be best for B-2 and P-2 roasters than for the rest as the motor support bracket was moved in these roasters making it impossible to fit a long sheathed thermocouple.

Perhaps this rather crude diagram will give an idea of one way to do it:


A metal support cylinder is fixed to the rear wall of the roasting chamber with J-B Weld. Silicone tube is used between the thermocouple and cylinder to provide both electrical and thermal insulation. The cylinder internal diameter is sized to compress the silicone tube sufficiently to grip the thermocouple reasonably tightly. Note that the rear wall is drilled with a clearance hole so that there is no possibility of the thermocouple touching the wall.

This general approach can be used with either bare wire thermocouples or sheathed thermocouples and allows for disassembly and replacement of thermocouples.

As for sourcing silicone tubing, it is readily available on the internet in many sizes and wall thicknesses and I believe that Randy found some from an aquarium supplier (just be careful that it really is silicone tube).

GreenBean wrote: ... just be careful that it really is silicone tube.

And make sure it's rated at high enough temperature. Many are rated under 400F and shrink to fit at about 390F.

Here's one that's rated to 500F.

Diagram is very informative, thank you very much.
I bought food grade silicone for the insulation, but tubing is much better idea for this particular application.
Regards,
Miroslav

drgary wrote:And make sure it's rated at high enough temperature. Many are rated under 400F and shrink to fit at about 390F...

Thank you drgary, I should have mentioned that. If you can find tubing rated to 500 F then that is ideal. The tube is not inside the roasting chamber, however, and I doubt it ever sees more than about 390 F, even on the face nearest the chamber, and that only for very short periods of time. I have been using tubing rated to 220 C (428 F) for well over two years and many hundreds of roasts and it shows no signs of problems so far.

iginfect wrote:Jim Shulman in DIY tc recommended Kapton® Insulated Thermocouple Wire type K 30 gauge available from Omega http://www.omega.com/pptst/KK_WIRE.html. Sorry I can't give the URL of Jim's post as my home computer is down where all my info is stored.

Just to clarify, I should order 25 feet of the 30 gauge type K wire and cut, twist, and feed a couple of of thermocouple wires into the chamber as shown, with silicone tubing for insulation? The bare wires are okay with the beans tumbling around them? I don't think I need any connectors, since the board has sockets for bare wires, right?

It looks like 1/16" ID silicone tubing should be okay (nominal size of the KK-K-30 wire is 1.0 mm x 1.4 mm), but correct me if I'm wrong and should get larger tubing. I have to order a lot of these items from the States, so it would be annoying to get the wrong thing.

I want to clarify that we do not have a recommended type of thermocouple or method of installation for thermocouples. Those of us associated with this project have all used different types of thermocouple and different approaches and have all got them working. I do not think that any of us are particularly proud of our approach or want to recommend it as the best way to go. Early adopters are going to have to review what others have done and decide for themselves what will work best for them.

I have tried to give some general pointers in the post above on 5 September at 5:05 am. In reply to the question by Miroslav I have provided a simple diagram of one possible way to go if using sheathed thermocouples. I use sheathed thermocouples as they are more robust especially, as you say, with beans tumbling around them. Others use bare wire apparently without problems.

I stated in the notes after the diagram that the same approach can be used for bare wire thermocouples but it is not necessary to go to this much trouble if using bare wire thermocouples. As stated in the general pointers, point 2.1, "the necessary insulation for bare wire thermocouples can be provided by their own insulation". In it's simplest and least robust form this would simply be a hole drilled in the back wall with the bare wire thermocouple pushed through the hole so it's insulation forms a tight fit in the wall. The addition of a dab of silicone sealant on the cool side of the wall may improve stability a little. I believe that some people have used J B Weld to fix the thermocouple but this could make it difficult and messy to remove.

I have never used kapton insulated thermocouples but have no reason to doubt that they would be good for this service. The wires are already insulated with kapton so do not need silicone tube. Purchasing 25 feet seems more than a little extreme to me. Twisting the wires together to form the hot junction will work, at least in the short term, but I would not recommend it as a robust solution. I have not looked but would be surprised if you could not obtain kapton insulated thermocouples made up ready to go.

Glass fibre insulated bare wire thermocouples are cheap, readily available and are also suitable for this service.

You are correct that you do not need any connectors, the TC4/C uses screw terminals.

I have received emails from people thinking of using stainless steel or inconel braided thermocouples. If using these it is important to insulate them and fix them in place so that there is no possibility of any part of the braid coming into contact with mains voltages inside the roaster. Suitable insulation would be glass fibre or silicone sleeving/tubing. If using grounded thermocouples the insulation is also needed to insulate the thermocouple from the roaster frame.

I am sorry that we can not give more specific advice at the moment. Hopefully others who have already installed thermocouples and the early adopters of this system will provide information on their approach.

If you really can not decide what solution to adopt but are keen to try the HT Roaster Interface then one possible way forward is to start with a very cheap simple solution that is not robust but will allow you to try the system and to change to a more robust solution in the future. That cheap solution might be to purchase two thermocouples similar to these:

These are readily available all over the world. Drill holes (carefully positioned) in the rear wall of the roast chamber to form a tight fit on the black insulation near the tip of the thermocouple. Push the thermocouple into place so that about 4 to 6 mm protrude into the roast chamber. Cut off the connector, connect to the TC4/C and enjoy.

The reason for the 25' length is because it's the minimum order (about $40).

I am kind of conflicted about the bare wire TCs now. I like the idea of not having to worry about insulation and metal conduit (more parts to order from more places and wait for delivery), but I'm also worried about running a bare twisted TC inside rolling beans. It sounds like Jim has no problem with this setup, though.

Omega sells pre-fabricated Kapton insulated bead TCs without connectors here, but you can only order in quantities of five. It's $89 for 30 gauge and $58 for 24 gauge with 36" leads. I'm prepared to shell out for the 24 gauge pre-fabricated if that would be appropriate for this application. I've looked at all the probe configurations Omega sells, and I am concerned about the 6" minimum length of most of them, as well as the insulation and mounting requirements.

Edit: I went ahead and ordered a five-pack of the 5TC-KK-K-24-36 (24 gauge Kapton-insulated type K beaded, 36" length).

That should provide a good solution Chris. Perhaps someone will purchase the spare TC's off you. I look forward to hearing how you get on with installation and testing of the TC's and the HT Roaster Interface system.

I have posted on TMC about another way to go, chosen by a TMC member, which we hope will provide a simple and robust solution but is as yet untested.

Using spare thermocouple leads I made mine by carefully and tightly twisting the pair of wires at the end. I used some small, smooth-jaw pliers. It is the junction of the two dissimilar metals that create the sensing point. How long they will last is another question, but in such an environment I would think that the life of any exposed thermocouple is in jeopardy to some extent. If one fails I can just pull it out of the tube, cut and re-twist, and go again. Are they accurate? The ET shows a differential of about 1 degree through most of the preheating of the roaster compared to the HT's display. I also tested them against a commercial fine-wire thermocouple before installing and they were very close- fractions of a degree which is probably ten times more accurate than needed.

If you want to get fancy, try this solution to weld your own:
http://kingpincooling.com/forum/showthread.php?t=961