SLC Analog In Deadspot

[Aabeck]

I just had to replace a SLC 1746-NIO4I analog card that was doing something I have never seen before.

Only 1 channel was used, and I tried the other and it acted the same. Set to 4-20mA input. It is on a Banner LTF laser distance sensor for a hoist position, and I replaced that first, then checked for blockage or stray reflection on the reflector.

The reading of the signal from 4 to 18mA was perfect, and from 18.7 to 20mA was perfect. However, from 18.1 to 18.6mA the reading on the PLC dropped from 14,800 to 9,000 - stayed about 9,00 for 8 inches of movement, then jumped to 15,600 and climbed steadily. This happened moving both directions, and was very repeatable on both channels.

Has anyone else ever had an analog input develop a "deadspot"?

 

[parky]

Never assume new equipment is ok, I once had a similar experience. I had an analogue signal that sometimes varied but not all the time, trying to find the problem was difficult as this would only happen on odd occasions. First call was to change the sensor, no luck. Same problem. even re-wired it directly to sensor bypassing the panel terminals, tried grounding only one end, then both.
I replaced the analogue card with a brand new one and guess what, still the same. At this point I was looking at all manner of other options nothing worked I even thought about replacing the complete PLC. When the new maintenance shift came on at night (these will only attempt PLC repairs if in emergency as their knowledge is limited) one engineer did not read the log and on a call from production looked at the problem, replaced the card (original was still in my office). hey presto it worked. My name was mud for a few days, I had replaced it but it seems the brand new one I fitted had the same fault. so, sometimes never assume because it's new and packaged its ok.
Worth a thought if you have more spares.

 

[cncsparky]

I've replaced at least a dozen SLC analog cards due to bad power supplies. These cards were typically >20+ years old. Look closely under the on board power supply electrolytic capacitors. You may see where they are leaking.

 

[Aabeck]

cncsparky,

I have replaced quite a few SLC analog card over the years, but the majority shorted and shut down the power supply or 1 channel failed.

This is the first one i've ever seen cut out in a certain range.

 

[BruceJW]

Did it work before?

Maybe need to add a load resistor?

this is where I run away, not my expertise.

 

[JesperMP]

How do you know that it is between 18.1 and 18.6 mA that it drops in this way ?
By a display on the laser sensor ?
Or by a mA-meter you have placed in series with the laser sensor and PLC input ?
I would check with a meter, and also check from sensor to 0V instead of from sensor to PLC.
Some laser sensors have programmable outputs that can set to various functions. Also, some sensors have inputs that can be set to manipulate various functions.

 

[Aabeck]

BruceJW,

This hoist, with this input card, has been running for 6-7 years, this problem just started yesterday.

Jesper,

I tested the laser output with a meter in the loop. The control input on this laser is set to disabled, the digital output is not used. The laser puts out 4-20mA (actually 4.3 to 19.6 for the travel limit of the hoist) very consistent through its travel.

I have bench tested the card with a mA supply and watching online in RSL the value drops in the same mA range on both channels.

 

[James Mcquade]

Yes,
this is common with early laser sensors.
at the low end of the measurement you have a curved hill - 100 ft let's say
you then have a dip to 50-75 ft
another hill - 200 ft
a second dip - 150-175,
a 3rd big hill that goes to 1200 ft.
a dip.
another hill
then nothing or another hill and then nothing.

what you need to do in this instance is to stay within the 3rd large hill area
since there is a dead band, it has not been linearized for the full range OR you are going outside its designed range.

james

 

[willxfmr]

BruceJW,

I have bench tested the card with a mA supply and watching online in RSL the value drops in the same mA range on both channels.

I just want to make sure I'm reading this correctly. Are you saying that with the analog card from the hoist installed in a different rack, and powered with a known good mA supply, you still get the same dead spot as you did when the card was in the rack for the hoist, and the laser on the hoist was providing the signal?



Bubba.

 

[Aabeck]

Bubba,

Yes. The analog card in a benchtop PLC, with a calibrated 4-20 supply, on both channels drops the value in that mA range.

When I metered the signal from the laser (before swapping the IO card) the output was steady and consistent, the mA did not drop when the hoist was in the 'questionable' area when the input showed it dropped considerably.

Before I metered the signal to the PLC I was checking for a dark spot on the reflector, reflections of shop lights that might have blinded the laser, made sure the laser spot did not move off the reflector, and tried the second, unused input channel.

 

[Jeff23spl]

interesting problems... the ones i have seen blown was either short or not working at all...It is even hard to imagine how the electronic can behave this way...

give that broken card to a school. Teachers would be happy to gives some headaches to their students...

 

[OkiePC]

New one to me too. I have had a couple of occasions where an Allen Bradley analog input was "off", but not in the middle of the range like that. I recently had one where the signal ranged from 4mA = 3200 to 20mA = 17500 on a 1746-NI16I where is should have been 4000 to 20000. This was the case on the last three channels of the card (all the free spares), so I ended up just changing the scaling logic in the PLC for that particular signal... I have also seen a 4 channel SLC input card where I got full range at 12mA (12ma = 32767)... that one had to be replaced. I think I have seen that symptom more than once in A/B analog input cards and also once or twice on a VFD analog input.

 

[willxfmr]

Well, you have me stumped. Maybe it's an un-announced "feature" of those cards.


Bubba.