Source an RTD

I need to find a 2-wire RTD for a controller I am putting together. I can find them all over the interwebz but I'd like to get one from a reputable place. Any ideas or recommendations?

This is for a keezer/fermentation chamber build if that matters.

Digi-Key

That being said, a 4 wire RTD is much more accurate. A 2 wire is prone to two errors; compartment temperature affecting the resistance of your leads, and self-heating due to increased current causing the RTD itself to heat up. 3 wire eliminates compartment temperature, but still has the self heating effect.

Mind you, it's probably not going to have a huge effect overall on your accuracy, but it would bug me. 3 wire only needs an extra line run; 4 wire needs a constant current source (not hard to source or make).

Thank you I'll have a look.

As a person with OCD tendencies I can appreciate your comments about the accuracy of 2-wire compared to 3 or 4. With short leads and low current, applied in an environment with a relatively high heat sink capability, these are just not likely going to be an issue for me.

Just to clarify, you need a current source with any RTD. Ohm's law pretty much requires it.

Also, with 2 wire RTDs, you need to account for line resistance, this is why a 3 or 4 wire configuration is more accurate: it negates the line resistance of the wire.


You could do a 2-wire setup, you would just have to calibrate the rtd measurement system. RTDs are not linear temperature devices, meaning that of 0 degrees is 0 volts, and 100 degrees is 100 volts, 50 degrees won't be 50 degrees. It might be close, but curve fitting is required. I generally use 4th degree polynomials when curve fitting a calibration curve. Best way to calibrate is to use a dekade box and inject the resistances that correspond to a given temperature. Without this, a small added lead resistance can lead to huge errors in temperature. A 4 ohm change corresponds to roughly 50 degrees. 10 feet of 20 AWG wire would have a resistance of around 0.1 Ohms, that works out to be about 1/2 a degree.


Granted, I'm used to dealing with tighter accuracy requirements than the average home brewer probably cares about, but you'll have to decide for yourself if you want to give up accuracy.


Generally get thermocouples and rtds from Omega

DougK said:

Just to clarify, you need a current source with any RTD. Ohm's law pretty much requires it.

Click to expand...

I hope someone putting together a system involving RTDs understands that.

DougK said:

Without this, a small added lead resistance can lead to huge errors in temperature.

Click to expand...

Big time on this. You'll want to figure out the length of your leads first, make all your connections (crimps, solder, what have you), and measure that resistance to add into your final calibration. Or, use those leads during the cal.

Stephonovich said:

I hope someone putting together a system involving RTDs understands that.



Big time on this. You'll want to figure out the length of your leads first, make all your connections (crimps, solder, what have you), and measure that resistance to add into your final calibration. Or, use those leads during the cal.

Click to expand...

Just making sure no one misunderstood and thought that a 2 wire RTD could be run like a thermocouple, only to be confused when they get no change in voltage with temperature.

It is MUCH easier to calibrate the lead resistance out. Some components won't give you their impedance, or they give you a nominal range. The highest accuracy will be obtained by minimizing this as much as possible.


I calibrate at the point at which an RTD is calibrated in its CAL lab. Generally at the terminal block.

I've used RTD's in non-brewing systems for years ... I think some of you are over-studying for the test. For what we are doing a simple calibration/check is fine, self heating is not going to be an issue. You can run these out with a couple hundred feet of wire and at that point you start seeing some shift but for a 24" lead and what we do, I'd really doubt you'd see too much.

When I used them in a professional setting they were so easy to find. Now, trying to find ONE at a decent price makes me wish I had accumulated more "Stuff" when doing that work.

Wouldn't mind knowing the answer to this too. I have a rock bottom ebay PT100 in my smoker and it is rubbish. But the alternatives seem way more expensive.

I got 5 DS18B20's in waterproof configuration for 5 bucks from an ebay seller out of China. I am using the +5 volt dedicated power and reading the data line into a 16f877A. I have 3 PID's that also feed the alarm and relay signals to the PIC to allow me to automate the sequence and hit all of my temps and times. They are definately available and much cheaper now too.

Wheelchair Bob

DS18B20s are great for microcontroller builds but you can't use them with off the shelf PIDs.

alien said:

Wouldn't mind knowing the answer to this too. I have a rock bottom ebay PT100 in my smoker and it is rubbish. But the alternatives seem way more expensive.

Click to expand...

What about them was rubbish? Mine all worked exactly the way I'd expect. Am I lucky or are we expecting different things?

The calibration was off by 1 at 0C and 7 at 100C, compared to a thermapen.

alien said:

DS18B20s are great for microcontroller builds but you can't use them with off the shelf PIDs.

Click to expand...

Uh, why not? Unless you mean just buying the device doesn't work off the shelf - they are just a three-wire RTD. All the eBay sellers I see put them in a sheath with some conductivity paste and some pigtails. Should work fine, at least I'm willing to invest $5 to find out.

LBussy said:

Uh, why not? Unless you mean just buying the device doesn't work off the shelf - they are just a three-wire RTD. All the eBay sellers I see put them in a sheath with some conductivity paste and some pigtails. Should work fine, at least I'm willing to invest $5 to find out.

Click to expand...

The DS18B20s aren't RTDs. They're ICs that use the 1 wire protocol. Each chip has a separate address and they're all connected to the same data pin on your microcontroller. They don't provide a change in voltage or resistance or anything with temperature. I use them and highly recommend them, they're easy to work with and mine were very accurate.

bobotwf said:

The DS18B20s aren't RTDs. They're ICs that use the 1 wire protocol. Each chip has a separate address and they're all connected to the same data pin on your microcontroller. They don't provide a change in voltage or resistance or anything with temperature. I use them and highly recommend them, they're easy to work with and mine were very accurate.

Click to expand...

Huh, well that would definitely prevent them from working as I intended. I did purchase a few different ones from eBay for under what I would have spent on one from one of the automation companies so hopefully I end up with one that will work for me.

All I really want to do is use this Pilsener yeast I have in the fridge!

bobotwf said:

The DS18B20s aren't RTDs. They're ICs that use the 1 wire protocol. Each chip has a separate address and they're all connected to the same data pin on your microcontroller. They don't provide a change in voltage or resistance or anything with temperature. I use them and highly recommend them, they're easy to work with and mine were very accurate.

Click to expand...

Yep, you wuz right. I know you know you were right, but you know how men are.

The other one I bought was the proper one and all is well now. I'm in $10 for two sensors, one of them worked. so not too bad. Here's what I ended up getting:

Probes Unlimited Pt100 Temperature Sensor RTD Automation Controller Teflon Probe





This shows it as a two-wire but it is a three wire so .. not sure but whatever it works and is within a couple tenths of matching a calibrated sensor setup so all is well in the world. Excellent service from this company, very fast shipping if anyone else has the need.

I used to do building controls but have been out for a long time having decided to stay with just the IT side of the house. My brew buddy is still in controls thankfully so he was able to get me straightened out.

LBussy said:

Yep, you wuz right. I know you know you were right, but you know how men are.

Click to expand...

Heh, hopefully it'll come in handy later, it's not wasted money it a future partially funded project.

This shows it as a two-wire but it is a three wire so .. not sure but whatever it works and is within a couple tenths of matching a calibrated sensor setup so all is well in the world.

Click to expand...

The picture shows two red wires that are equivalent. That's so you can detect the inherent resistance in the wires in your circuit and subtract it from the observed resistance the RTD circuits reports. If you don't do that you'll be off by a tiny little bit.

Oh! Okay that makes sense. I'll let my buddy know and he can go do. Thanks.