pid accuracy and sensor type??

Hi,

I posted the question below in the electric brewing forum but it seems no one answers, so I thought I'd try here. My question is:

My pid manual says "Indication and control accuracy 0.1℃, ±0.2%FS " What does this mean and how accurate can I expect the temperature measurements to be???

Thanks for any help!

Previous post in Electric Brewing:

Hi All,

I've recently bought a sestos PID with the aim of upgrading my brewing system to a herms type system.

When reading the PID instructions I see that the accuracy of the PID is quoted as "Indication and control accuracy 0.1℃, ±0.2%FS ".

My question is what does this mean???

My understanding is that it says the PID reports back in 0.1C increments and 0.2%FS is 0.2% of the full scale accuracy.

If I use a k-type thermocouple the scale is -50 to 1350C i.e. 1400C full scale. So the accuracy is 0.2% x 1400C = +/-2.8C - this seems bad??? Is this an offset that doesn't change is is it really saying that my reading is within 6C ?

If I used a CU50 the range is -50 to 150C and the accuracy is +/- 0.4C

Or a PT100 -200 to 600 so an accuracy of +/- 1.6C

So should I use CU50 ?? or have I misunderstood ????

Cheers!

The PID accuracy does not include the accuracy of your thermocouple. If you are going to use a thermocouple I would order a GE ultra premium high accuracy T type with a seal end and of course you must calibrate both as a system to determine the accuracy. If you can get a 2 wire thermistor probe or a platinum resistance probe you will be able to calibrate it to tight tolerance as well. Here's a link for the GE thermocouples http://www.ge-mcs.com/download/validation/Validation_TC_Datasheet_final.pdf

Either way you must calibrate it to check the accuracy. I hope this helps.

I have used K-type thermocouples for almost a decade in my old system. But I recently upgraded the system and it includes longer leads (~6') on the thermocouple wiring. That has significantly degraded the accuracy. I plan on moving to a 3-wire Pt100 RTD for this system to restore the accuracy. Either 3 or 4 wire cabling to a RTD helps take out the issue of the long leads.

If your system will have short leads, I think that the K-type thermocouple is still a decent choice. Less expensive too.

RDWHAHB
You are understanding the spec correctly.
Try to use a controller and probe where your operating arena is close to the middle of the 'full scale'.

An RTD is more accurate by design.

It is best to setup your controller with a pot of boiling water and enter the offset that matches 212degF. (sea level)
Mash temp of ~165degF may then only be 1/2degF off.

If you have a trusted thermometer you've been using, set that offset into the controller. It will help with consistency until you get a grasp of the controllers and their accuracy.

For fun, find the accuracy of your current temp measurement system.

'da Kid

oxonbrew said:

If I use a k-type thermocouple the scale is -50 to 1350C i.e. 1400C full scale. So the accuracy is 0.2% x 1400C = +/-2.8C - this seems bad??? Is this an offset that doesn't change is is it really saying that my reading is within 6C ?

If I used a CU50 the range is -50 to 150C and the accuracy is +/- 0.4C

Or a PT100 -200 to 600 so an accuracy of +/- 1.6C

So should I use CU50 ?? or have I misunderstood ????

Cheers!

Click to expand...

Any of those options will work for you. Don't get hung up on full scale accuracy. You will be using the temperature probe in a very limited range. The PT100 is your best option but a TC will work just fine. I don't have any experience with a CU50. Your sensors will ony be as accurate the calibration you perform so make sure you focus on that and get the sensor dialed in at the setpoints/range you will be using it.