Hi,

We are developing a new process using three reactors. Two reactors are each being heated/cooled by circulating oil temperature control systems. These units have their own controller as well as remote. Each reactor has two RTDs (one main, one backup) that report to the remote control cabinet which can switch between RTD inputs. Switching between inputs does not help. The controller in each of these cabinets displays temperature and functions correctly. The displays for the backup RTDs (Jenco 765 displays) are all over the map. I can use the zero function to dial it in and it remains stable for awhile then goes screwy. All RTDs are connected to their cabinets using 18AWG 3 conductor shielded cable. The heating units to have contactors which seem to affect the Jenco display for awhile. After a time they don't seem to bother it. Any ideas?

Thanks,
Joshua

 

It isn't clear to me exactly what the problem, "all over the map" means (no examples provided).

If the temperature drifts on the backup indicators but not on the controller and you believe the controller, then the indicators are not stable.

If the indicator shows an offset from the controller's temperature, and you believe the controller, then the indicator is not stable.

If the indicator's values are jumping around, but the controller's temperature values are stable, then the indicators do not have the ability to reject electrical noise like the controller has.

All this is premised on your trust in the controller's readings. Maybe the controller's indicated temperature value is damped down (filtered/averaged) so much that you can't see the changes the indicators are showing.

 

To be honest, you have not really told us much. There are so many things we don't know from your description that it is hard to even start.

Perhaps you could tell us more about how you switch to the backup RTD.

is the controller somehow connected to any of these RTDs?

your description of what you have is so vague that it is going to be really hard to help you.

Incidentally, "screwy" is not a technical term that has any meaning.

--
Bob
http://ilbob.blogspot.com/

 

> If the indicator's values are jumping around, but the controller's temperature values are stable, then the indicators
> do not have the ability to reject electrical noise like the controller has.

> All this is premised on your trust in the controller's readings. Maybe the controller's indicated temperature value
> is damped down (filtered/averaged) so much that you can't see the changes the indicators are showing.

The controllers reading agree with temperature taken using a IR thermometer so they seem to be correct. Once the indicator is zeroed in to meet the displayed temperature on the controller it agree with the controller for most of the day. Sometime later the indicators will get off track. For instance, right now I have one indicator bouncing between -2 and -10 F while the corresponding controller shows 104F. The second vessels backup indicator is bouncing between 290 and 320F while its controller show 68F. Neither vessel is actively heating nor cooling at this time. I'm thinking its an electrical noise issue, but I'm not sure how to get rid of it.

Thanks, (I hope I was more clear this time)

Joshua

 

> To be honest, you have not really told us much. There are so many things we
> don't know from your description that it is hard to even start.

> Perhaps you could tell us more about how you switch to the backup RTD.

> is the controller somehow connected to any of these RTDs?

Hopefully, my reply above will help clarify the issue.

I will try to answer your switching question. We had another company build us a small control cabinet that includes the controller, selector switch and backup display. Inside the box they have terminations for the RTDs and run through a couple of ice cube relays to accomplish the switching between inputs to the display and controller.

I hope this helps.

Thanks,
Joshua

 

You should look up the part number on the relays that switch the RTD signal and make sure that the relays have gold plated contacts. The excitation current for an RTD is generally < 1mA, and such low currents require gold plated contacts. Any resistance added in an RTD circuit appears as a false 'higher temperature'.

I assume you're using 100ohm Platinum RTD's. (Are you sure that they're RTD's?) What's interesting is that your indicators drift in different directions, one is lower than controller temperature and the other higher than the controller temp and both errors are significant error.

Since the readings originally agree, then one assumes the wiring is correct, but assumptions are the devil's tool.

Any chance the extension wiring from the RTD to the indicator accidentally used thermocouple wire, instead of copper wire? A temp gradient across thermocouple wire could account for the changes over time and if wire were one polarity in one circuit and swapped polarity in the other circuit it could account for the change in the direction of the error.

Any chance that the original thermocouple sensors were replaced with RTDs and the wiring wasn't changed?

RTD's are normally isolated from their sheath, not grounded like a grounded thermocouple, which if that's the case, rules out common mode from the RTD sheath to the indicator. These aren't home-brew RTD's are they, something a grad student fabbed to save some money? They are commercial RTDs from a reliable source?

The relay are electromechanical relays, right? contacts isolated from the coil? Dry contacts? "Ice cube" infers that they are, but I have to ask because the use of a solid state relay would reek havoc on an RTD circuit.

If they're commercial RTDs with copper wire through an electromechanical relay, then take is something is wrong with the indicators - either a cheap front end with little or no noise rejection or you got a couple indicators from the same production batch where something is really amiss; the excitation source drifts, whatever.

Apparently, the noise is not overwhelming because the controllers, using the same wiring and same RTD's are functioning properly. It could be the indicator just can't handle whatever electrical noise is present. You can mess around with shielded wiring, but sometimes it's easier to get an indicator that has good noise rejection.

I'd suggest that you beg or borrow another controller and temporarily hook it up in place of an indicator. If it's stable, your indicators can't handle the task.

 

Thank you for the suggestions and a place to start.

Joshua

 

I guess the indicators are just not up to the task. I temporarily installed a different indicator/controller that I had lying around and the temperature indication is rock solid.

There shouldn't be a lot of electrical noise in this area (VFDs not running, etc.) so I'm kind of wondering in what applications the original indicators would be adequate but all's well that ends well. I've another indicator on order to try as a permanent replacement.

Thanks again,
Joshua

 

Hi Joshua,

It sounds like you have a ground loop problem on the cables for the back up indicators. Is the RTD cable shield grounded at both the RTD and at the indicator? I've seen this before - An instrument will appear to work fine for a while but then go screwy. Grounding the shield at both locations will sometimes cause current flow because of different ground potential. I would ground the shield at the indicator only and tape it off in the RTD head.

 

Thanks for the suggestion. The grounding wire is only landed at the indicator. That was a problem I learned about a long time ago the hard way so I remember to only land one side now.

Thanks, again.