LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Noise on digital data

I have a cDAQ-9185 chassis with two modules:

  • NI 9205 Voltage
  • NI 9425 Digital.

My question has to do with the digital data. I am measuring 0-24V outputs from a PLC to valves, solenoid and other industrial devices. I am collecting 1,000 samples/second. I am seeing little spikes where a signal goes from zero to one when it shouldn't be. Here are a few examples (where the red arrow is pointed):

flycast_0-1626900136085.png

 

Every time I see a spike it goes from zero to one on a single data point. The spike is always happening at the same time that some other digital signal is legitimately going from zero to one. We are wiring with a wire from the output of the PLC to the positive terminal on the DI and the ground from the PLC to the common on the DI.

flycast_1-1626900466048.png

 

I don't seem to see any spikes in any of the data coming from the NI 9205 Voltage.

 

I have tried to speed up the data collection to 25,000/sec (the max of this system) and the result is the same - the spike happens on a single data point rather than being spread out over many data points.

 

I would like some suggestions regarding determining if these spikes are coming from the PLC or are they some kind of noise. Ultimately I want to get rid of them so I can completely trust my data.

0 Kudos
Message 1 of 5
(489 Views)

@flycast wrote:

I have a cDAQ-9185 chassis with two modules:

  • NI 9205 Voltage
  • NI 9425 Digital.

My question has to do with the digital data. I am measuring 0-24V outputs from a PLC to valves, solenoid and other industrial devices. I am collecting 1,000 samples/second. I am seeing little spikes where a signal goes from zero to one when it shouldn't be. Here are a few examples (where the red arrow is pointed):

flycast_0-1626900136085.png

 

Every time I see a spike it goes from zero to one on a single data point. The spike is always happening at the same time that some other digital signal is legitimately going from zero to one. We are wiring with a wire from the output of the PLC to the positive terminal on the DI and the ground from the PLC to the common on the DI.

flycast_1-1626900466048.png

 

I don't seem to see any spikes in any of the data coming from the NI 9205 Voltage.

 

I have tried to speed up the data collection to 25,000/sec (the max of this system) and the result is the same - the spike happens on a single data point rather than being spread out over many data points.

 

I would like some suggestions regarding determining if these spikes are coming from the PLC or are they some kind of noise. Ultimately I want to get rid of them so I can completely trust my data.


It's been a long time, but I think I remember that there was a situation where I had unshielded wires running right next to each other that caused this.

Bill
CLD
(Mid-Level minion.)
My support system ensures that I don't look totally incompetent.
Proud to say that I've progressed beyond knowing just enough to be dangerous. I now know enough to know that I have no clue about anything at all.
Humble author of the CLAD Nugget.
0 Kudos
Message 2 of 5
(456 Views)

I'm no expert, but solenoids (all coils, incl. relays), can give very high induction peaks when a step (positive or negative) is applied.

 

IIRC, you can use 'bleeding' diodes so the over current can safely flow back.

 

It seems online they're called Avalanche diode - Wikipedia or Transient-voltage-suppression diode - Wikipedia.

 

I'd expect industrial solenoids to have them build in, but there might be reasons that they are not.

 

Note that once you start measuring the induction, you're changing the system and you might not see anything. These peaks are very short, so a multimeter or scope might not show them because of that.

Message 3 of 5
(423 Views)

I have been using the median filter with good results to filter out the spikes in already collected data. We will explore the avalanche diode though. Seems like a better option.

0 Kudos
Message 4 of 5
(364 Views)

@flycast wrote:

I have been using the median filter with good results to filter out the spikes in already collected data. We will explore the avalanche diode though. Seems like a better option.


If such a diode is needed, not having it will damage your hardware, sooner or later.

 

Induction spikes can be in the kVs range. If you're input is TTL-ish (~5V) it won't last long.

Message 5 of 5
(361 Views)