Created Date: April 17,2014

Created by: Richardv

I have used 2 RTD's on 2 channels scaling 1 channel for the full range and the second for the working range. I switched to the working range when I reached that temp for control and monitoring. Just extend the working range by a few degrees. The only added cost is the RTD as the RTD input module has 4 inputs. I did this on a DL06 used to test Temp senders used in automotive applications. You may be able to do it with 1 input and change the scaling when needed. I had a spare RTD so I used that.

Created Date: April 17,2014

Created by: nickh

Rich/Bob,

Thanks for your replies, the system currently uses Eurotherm Mini8 loop controllers in conjunction with a 260 plc, my thought is bring all the loop control into the plc, the settings could be done on the HMI.

So, in brief are you saying that the 260 analogues would give me the resolution I need if the system is scaled correctly ?

Thanks,

Nick

Created Date: April 17,2014

Created by: Bob S BN

So, in brief are you saying that the 260 analogues would give me the resolution I need if the system is scaled correctly ?Nick

I don't think so. You can scale whatever you want after the number is in the PLC, but that RTD has a range of "A " - "B ", and the analog card will divide that range into the number of pieces the card is capable of resolving to.

I will do a bit more studying to see if the RTD input card has some different setting that the TC cards do not.

EDIT: I don't see any settings on the RTD input card that would allow you to reduce the scaling at that point.

Are you sure your existing hardware is capable of this spec, I couldn't find the info on their RTD modules. Do you have the possibility of using serial or ethernet to allow the PLC to write SV and read PV from your existing MINI8?

Created Date: April 17,2014

Created by: Tinker

Have an application that will use some RTD's with a range of -200 to +850 degrees C, the working temperature of the proposed system is 16 to 22 degrees C, the intended temperature control needs to be to .001 degree.

How do I work out the resultion of the required analogue input to give me the correct temperature resolution?

The range of the RTD is 1050 degrees, dividing this by 32767 for 16 bit card gives .032 degrees, does this mean I would have to go to a 24 or even 32 bit analogue to give me the required .001 ?

Thanks in advance,

Nick

"the control needs to be to 0.001 degree "? is that a literal accuracy figure, or do you just need that resolution to detect changes? While the resolution is probably manageable (though probably not with a conventional PLC input module by itself), that level of accuracy is asking a lot. Though, since your span is so small, if you can calibrate at your working temperature then repeatability is the main issue, full rage accuracy might be less critical.

For that span and resolution requirement you might consider thermistors rather than platinum RTD.

over the 16 to 22 degree range a Pt RTD will change resistance about 2.33 Ohms, for your 0.001 deg. resolution you'd need to measure with a resolution on the order of 100 ?? (a fairly tall order). With a 10k thermistor the change is more like 3210 Ohms and a 1/10 Ohm resolution would much easier to manage

I suspect that you will likely be better off looking for a laboratory instrument that sends data to your PLC via a communications channel. Sounds like that is what you are doing now?, I think I'd stay with the current setup

Odd than in a couple of weeks there have been two people asking about 0.001degree resolution, while I have never seen that kind of resolution discussed before.

http://forum.automationdirect.com/showthread.php?t=13469 Oh, that was you too, never mind.

EDIT: I did a little calculating (no guaranties I got it right) and if you can use a thermistor, it looks to me like you just might almost pull this off using a 10K thermistor, a F204THM in millivolt voltage measuring mode and proper values of resistors in the bridge circuit. Well, at least you could manage the resolution, to get an overall 0.02% accuracy will still be a challenge. Self heating in the thermistor (hint, consider pulsed measurement) and stability of the bridge will be an issue.

Of course thermistors are not linear, however that can be dealt with a bit of math. But I have no idea how to do it in Direct soft, or if a 32bit real is even up to the task, possibly a lookup table generated in a program using an extended precision number format would be better. Though that could take significant chunk of the memory in a 260, so you might consider a DoMore

Created Date: April 17,2014

Created by: nickh

Have an application that will use some RTD's with a range of -200 to +850 degrees C, the working temperature of the proposed system is 16 to 22 degrees C, the intended temperature control needs to be to .001 degree.

How do I work out the resultion of the required analogue input to give me the correct temperature resolution?

The range of the RTD is 1050 degrees, dividing this by 32767 for 16 bit card gives .032 degrees, does this mean I would have to go to a 24 or even 32 bit analogue to give me the required .001 ?

Thanks in advance,

Nick