Thermocouples: Can You Place A Switch Between Two

[Justin]

Hi guys.

Just a quick question in case someone knows. I have a basic understanding of thermocouples and how they work but I just have a quick question that I haven't seemed to be able to find an answer to on google.

Say I want to measure/control two vessels with one Temp controller which uses a single thermocouple.

Can you have two thermocouples, one mounted in each vessel, and then just wire in a switch to flick between the two thermocouples? Are there any calibration issues, variation, errors in accuracy that you would see by installing a switch in this instance. I'm guessing it would be best to keep the same lengths of copper wire, and of course use identical thermocouples. But are there any problems with mounting a switch to flick between the two.

Really just wondering if the switch might cause any interference with the voltage output generated across the TC.

Obviously there are issues with using the single output from the controller and which vessel it controls, but I can deal wit that. That's easy.

Cheers, Justin

 

[coolum brewer]

Hi guys.

Just a quick question in case someone knows. I have a basic understanding of thermocouples and how they work but I just have a quick question that I haven't seemed to be able to find an answer to on google.

Say I want to measure/control two vessels with one Temp controller which uses a single thermocouple.

Can you have two thermocouples, one mounted in each vessel, and then just wire in a switch to flick between the two thermocouples? Are there any calibration issues, variation, errors in accuracy that you would see by installing a switch in this instance. I'm guessing it would be best to keep the same lengths of copper wire, and of course use identical thermocouples. But are there any problems with mounting a switch to flick between the two.

Really just wondering if the switch might cause any interference with the voltage output generated across the TC.

Obviously there are issues with using the single output from the controller and which vessel it controls, but I can deal wit that. That's easy.

Cheers, Justin

[post="89334"][/post]​

Well here goes - my first post (be gentle with me). I feel I'm well qualified to answer this as an instrument engineer.

Technically you can switch between two thermocouples, but you need to use a special thermocouple switch which has contacts made from the same alloys as the thermocouple wires. If you use any old switch you create a new thermocouple junction because of the dissimilar metals and will introduce unpredictable errors.

Thermocouple switches are expensive, and you'd be better off using a second controller anyway. Hope this helps.

Cheers
Peter

 

[Justin]

Wicked answer. Thanks mate, that is exactly the answer I was looking for.

I assumed that it would complicate matters along those lines-hence my question. I suspected the switch would cause some errors with the voltage output from the thermocouple so it appears my first throughts were correct.

I appreciate the response, i'll go with a different plan. It wasn't essential to run two vessels off one controller, it just struck me that perhaps I might be able to do it seeing as I didn't need temp control on both vessels at the same time.

Cheers and many thanks. Justin

 

[KillerRx4]

Is there a problem then with using banana plugs on the ends of thermocouple wires being dissimilar metas & all?

 

[Justin]

Not sure. Wondered that as well. I was thinking I could by a couple of the $10 K type thermocouples from Jaycar. If I mounted the same type of plug that you have in multimeters I could potentially just swap them over rather than switch them.

Still keen for anyone elses thoughts.

Cheers, Justin

 

[coolum brewer]

Is there a problem then with using banana plugs on the ends of thermocouple wires being dissimilar metas & all?

[post="89347"][/post]​

If the banana plugs are at at the terminals of the controller it's not such an issue. If you're extending the thermocouple wires with the banana plugs, it is. Once again there are special plugs/sockets available with the contacts made of the same alloys as the thermocouple wires - this is what you should use. Also, when you extend thermocouple wires you need to use special thermocouple extension cable i.e. not copper wire.

This link might explain things a bit better Omega

By the way, no self-respecting brewery would use thermocouples for temperature measurement. they rely on RTD's which is a different technology based on a changing resistance of a coil of platinum wire. More accurate, linear and reliable (and usually more expensive) than thermocouples.

Cheers
Peter

 

[Beer bum]

Killer,

Same problem as with the switch. You have the a different metal on the plug than the t/c wire
Different thermocouples use different metal in the wires and if you mix and match, no matter how small, you will be surprised how much inaccuracy can be introduced into the controller.

Cheers

 

[Wortgames]

My understanding (I'm NOT an instrument engineer!) is that thermocouple controllers have an internal temperature sensor, to negate the effects of the join at the controller itself.

So if you extended the rear terminal block with a bit of copper wire and a switch, providing it is all in the same enclosure and at the same ambient temp, any effects would be minimal.

I'm happy to be corrected though...

 

[coolum brewer]

My understanding (I'm NOT an instrument engineer!) is that thermocouple controllers have an internal temperature sensor, to negate the effects of the join at the controller itself.

So if you extended the rear terminal block with a bit of copper wire and a switch, providing it is all in the same enclosure and at the same ambient temp, any effects would be minimal.

I'm happy to be corrected though...

[post="89359"][/post]​

You're basically correct with what you say.

The controller has an internal temperature sensor to measure the temperature at the terminals. This is because the thermocouple actually measures the difference in temperature between the measuring point (known as the 'hot' junction -even if it's in a fridge) and the controller (the 'cold' junction). So if the temperature being measured is 100C and where the controller is located is at 25C, the output of the thermocouple is equivalent to 75C. The controller's internal sensor then 'adds' the 25C so that the controller 'sees' 100C.

So, in short, if the point that you join the wires is at the same temperature as the terminals of the controller, the error will be zero. However, in practice, because the thermocouple generates a tiny voltage (millivolts) any poor joins, or dirty switch contacts, will contribute unpredictable errors.

Hope I haven't sent you all to sleep!

 

[Wortgames]

Hope I haven't sent you all to sleep!

[post="89368"][/post]​

You'd be surprised how many of us are nerds, CB.

Bring it on, we love it...

:super:

 

[Wortgames]

Is this also true for NTC and PTC probes, or is it just K & J type TCs?

 

[coolum brewer]

Is this also true for NTC and PTC probes, or is it just K & J type TCs?

[post="89377"][/post]​

Different beasts.

Thermocouples provide a millivolt output proportional to (but not linear) the difference in temperature between the hot and cold junction. Widely used throughout industry, very robust and can measure high temperatures and over a wide range. There are a handful of standard types (K, J as you mention also T, B, W, and a few others the code relates to the material the thermocouple wires are made from and their applicable measurement ranges).

PTC/NTC devices are passive electronic components (similar to a resistor). NTC (Negative Temperature Co-Efficient) sensors change their resistance opposite to the change in temperature i.e resistance goes down as temperature goes up. PTC (Positive Temperature co-Efficient) are the opposite. These are used in scientific or electronic applications and have very few applications in industry due to the limited temperature range they operate over and they are relatively fragile. There are literally hundreds of types of NTC/PTC devices, so it is nearly impossible to match a random sensor with a random instrument. Leads can be extended using copper wire, although this may cause an offset in the output (due to the resistance of the copper wire) which can be corrected for in the instrument.

Sleepy yet??

 

[Wortgames]

Sleepy yet??

[post="89382"][/post]​

Not remotely! Because iocane comes from Australia, as everyone knows, and Australia is entirely peopled with criminals (name that movie...)

 

[coolum brewer]

[/quote]

Not remotely! Because iocane comes from Australia, as everyone knows, and Australia is entirely peopled with criminals (name that movie...)

[post="89394"][/post]​

[/quote]

This suddenly got a bit weird.



The Princess Bride??

 

[Wortgames]

Top marks!

Sorry, my normally dizzying intellect has turned to mush this afternoon, easily distracted.

We now resume our normal programming...

 

[wessmith]

Hi Coolum Brewer and welcome. We picked up a bunch of PT100 (RTD) sensors in China a couple of weeks ago from a specialist electrical shop in Ningbo. The sensors were Y10 each (about A$1.60) and we also found an IR Thermometer gun with laser sighting for A$110.

We actually prefer the PT100 sensor for the lower temps in brewing and specify them on all our brewing kit. I have never liked the thermocouple approach where you rely on the installer to have clean hands when twisting the wires! So many calibration issues.

Great to have you on board.

Wes

 

[Justin]

Would you like to sell some of those $1.60 PT100's Wes. I know a bloke who'd like one or two h34r:



Me! Nice work Wes, that's a great price. Thermocouples seem to be easier and cheaper to come by but I agree PT100's would be nice.

Cheers guys and thanks for the great response. I've learnt a lot today in this thread.

JD

 

[BrissyBrew]

WesSmith

We picked up a bunch of PT100 (RTD) sensors in China a couple of weeks ago from a specialist electrical shop in Ningbo.

Hi there Wes how many did you pick up do you have any spare or have an email address for the chinese supplier.

It appears that I can use an RTD on my PIDs I have, from what I gather due to my dumby experience an RTD is easier for me to install (without stuffing things up) and should provide better accuracy for brewing then a thermocouple.

Can anybody provide clarification of this, is there different types of RTDs like theremocouples?

 

[coolum brewer]

WesSmith

We picked up a bunch of PT100 (RTD) sensors in China a couple of weeks ago from a specialist electrical shop in Ningbo.

Hi there Wes how many did you pick up do you have any spare or have an email address for the chinese supplier.

It appears that I can use an RTD on my PIDs I have, from what I gather due to my dumby experience an RTD is easier for me to install (without stuffing things up) and should provide better accuracy for brewing then a thermocouple.

Can anybody provide clarification of this, is there different types of RTDs like theremocouples?

[post="89524"][/post]​

Firstly Wes, thanks for your welcome.

BrissyBrew, there are several types of RTD's, but by far the most common is the Pt100. The Pt indicates it is a platinum winding, the 100 means 100 ohms at 0C. There are also Pt1000 and some rare ones that use copper. It is likely that if you have a microprocessor based temperature controller that it will accept Pt100's. You will have to select the input type in the set-up mode of the controller - this will be shown in the manual; it is also possible you will have to make a change to the circuit board of the controller (usually break a solder-link on the board) to disable the cold-junction temperature sensor inside the controller (refer to previous post). This should also be covered in the manual.

I said earlier that all commercial breweries use Pt100 sensors, but for the home brewer thermocouples should be fine.

cheers
Peter