General Input stability - suggestions?

Hi folks.

I'm having an issue gaining stable sensor readings using a General Input Module, and hope someone may have a suggested fix. The issue is a wandering reading for each sensor, which equates to an unscaled fluctuation of between 1-7 out of 255 level, or up to 3%, with the reading changing in this range at every "Show Current Level" sample (in Toolkit). When scaled, this makes the temperature readings bounce around by +/- 1 degree Celcius.

The sensors and associated gear are high quality, and of differing types, which indicates the sensors are likely not the issue.

I have four sensors currently connected:
- Pt100 via Delta Ohm HD988TR1 temperature transmitter (4-20ma), calibrated 0-50C
- Pt100 via Delta Ohm HD988TR1 temperature transmitter (4-20ma), calibrated 0-50C
- Viasala Humitter 50Y temperature transmitter (0-1V)
- Viasala Humitter 50Y humidity transmitter (0-1V)

Input power is 24VAC. Tried both DIN rail PSU and the Clipsal supplied PSU to isolate that, and no change.

A digital volt meter reads the 0-1V sensors stable at 1/100 of one volt, so no fluctuation there. I have not put a meter on the 4-20ma sensor transmitters as yet, but am not expecting fluctuation given the newness and quality of the kit (and there are two of them that give the same result).

The fluctuations have also been consistent over a one week period so far, and occur whether shortly after the unit is powered or after many days of operation.

I have not tried a regulated input voltage to the unit or sensors as yet, as it didn't seem worth it given my DVM readings of 0-1V sensors not moving and accuracy rating of the 4-20ma transmitters with respect to voltage fluctuation (0.4uA/V, with transmitter resolution of 4uA).

Any ideas?

Steve.

 

Are the sensors and the general input unit running off the same PSU? Having different supplies can cause this sort of thing.

Wiring can have an effect here. Long cable runs, shared common connections, cable type, cable loops, etc can all introduce noise. Try with only a single sensor connected to the general input unit on as short a cable run as practical and re-test.

Multimeters and the like can be pretty slow in their response. Whilst the average reading might be stable it's quite possible that there's higher frequency noise that is causing the readings to be variable. The input impedance of the general input unit is pretty high so the additional load of the multimeter may be affecting the reading anyway. I'd try powering the sensors and the general input unit from a good bench DC supply that you know has a clean output (check with a CRO if you can). If the issue goes away then it's the quality of the power coming from the PSU that is the issue.

If all else fails, you can always average the readings in logic, if you have a logic-capable unit on your network.

 

Newman, great thoughts as always. And thanks for the reply.

Been doing some testing on and off to try and get a stable sample from the General Input, following your suggestions, with no luck whatsoever.

Tried the following power combinations:
- Sensors and GIM are powered from the same source (as it was before)
- Two different 24V AC adapters - same result
- Clean bench DC power supply at 15V - same result
- Battery supply at 18V - same result
- Polarity of DC power source to GIM reversed - same result

I don't think input power is the issue.

Also tried (all while on battery power):
- Single sensor to channel 2 only, 4..20mA, cable to mA temp transmitter 20cm, cable from transmitter to 3 wire Pt100 2m - same result
- All channels on adjacent dimmers/relays/other loads off to minimise adjacent noise - same result

By 'same result' I mean the sample wanders between level 102 and 107.

And also:
- Disconnect channel 1 Pt100 sensor lead to send transmitter output to 22mA (force error condition) - similar result, with sample wandering between 248 and 254, with most samples at 250/251/252.
- Disconnect everything from channel 4 (no connection), and sample wanders occasionally to 1, then back to zero.

And then... weird stuff happened...

I reconfigured for 0..20v input, channel 4, connected up the battery source to channel four... and got a rock solid sample from that channel. Then I checked channel 1, 2 & 3 ... all channels were rock solid - for several minutes that I stared in wonder. Then I continued playing around...

I disconnected 18V source from channel 4, reconfigured for 0..1V sensor, hooked the sensor up & all channels started wandering around again. Damn. Reconfigured channel 4 for 0..20V, hooked up 18V DC to it (just like the test before), and all channels remained wandering around!

The only differences between the first 0..20V reconfiguration and the second were two things: I accidentally did a full save in toolkit on the first reconfig, not changes only. And secondly, I had to disconnect CBus to hook up the wire link for 18V DC. So I tried a full save with configuration for channel 4 at 0..20V, and it made no difference. Samples still wandering. Then I disconnected CBus and reconnected. Samples still wandering.

As an aside, all channel sample levels, regardless of whether 4..20mA, 0..1V, 0..20V wander by exactly the same order of magnitude - around 5 to 7, which is enough to be an annoying margin of error, without making the readings unusable.

This this is driving me nuts!

Have absolutely no idea why the GIM started producing stable samples at one stage during testing, then went back to wandering. A couple of minutes of stability is unheard of for this thing, so it was definitely anolmalous.

As a final test (just to prove that the definition of insanity is to try the same thing twice and expect a different result) I did another complete config save after hooking up the 18V to channel 4. Samples still wandering.

Any more thoughts gratefully received.

Steve.

 

Does anyone have General Input experiences they can share?

Wondering if I'm the only one with a General Input that will not read a stable sample...

 

I suggest that you call Tech Support. If you're having stability issues that can't be traced back to anything they will probably assist you in organising a replacement.