Running problems with HSC or Pulse Measurement Inputs

[jcperis]

I have running problems when the PLC gets pulse data from high speed counters inputs or from pulse measurement inputs. The PLC slows or even stops the ladder-logic program. When input pulses stop, then execution continues normally. And this problem is only at pulse frequencies between 450 Hz and 1250 Hz, aproximately. Outside this range PLC works fine, even with frequencies greater than 1250 Hz.

I am using a FMD88-10 PLC and I detected the problem when I configured two high speed counters inputs to get pulses from two rotatory encoders with 2000 pulses per revolution. Rotatory speeds are very low, between 15-100rpm. Then I tested to read pulses using pulse measurement inputs, and I got the same problem.

We have looked for noise problems, but is all ok. We have also made tests with ladder-logic programs wiht only the minimum lines to configure the high speed counters and/or pulse measurement.

Thanks,
Juan C Peris

[support]

We tried to replicate your reported issue and we indeed observed the same issue that right after the program is transferred to the PLC the CPU seems to slow to a crawl when both HSC channels are activated and subject to pulses of your reported input range.

However, we then power reset the PLC and it seems to run properly. Re-transfer the program to the PLC did not cause the same issue. However, when the PLC is first powered on and the program is transferred to the PLC it will behave as reported. Power on reset of the PLC seems to resolve it.

There could be a bug in the system firmware that led to this strange behavior. We will bring this to the attention of the firmware development team. For now you may like to try to power on reset the PLC after transferring the program and see if the issue will be resolved.

[jcperis]

I tried to power reset the PLC after transferring the programa and I got the same problem. If I re-transfer the program the PLC also works wrong.

If I do not activate the HSC channels, the PLC also fails when is subject to pulses between the range of frequencies indicated.

Thanks,
Juan C Peris

[support]

After conducting further investigations we found out the cause of the slow down. This is what happens:

During the I/O scan cycle the PLC performs software filtering by checking the logic state of all the digital inputs and compare the result with previous scan. If there is a difference, it performs a re-scan immediately to ascertain if the difference is due to a noise glitch. If the re-scan shows the same logic state then it is accepted and the I/O scan is completed.

In this case it appears that when the multiple inputs are exposed to pulses of certain frequency range, they cause the I/O scan not be able to get a stable reading and it causes the CPU to spend a lot of time rescanning the I/O, hence the ladder logic scan is adversely impacted. The CPU will be able get off the I/O scan when the frequency of input pulses changes increase or decrease because it will not resonate with the I/O scan time. Note that the HSCs are not losing counts and the PM measurement  are still accurate. All interrupt based custom function will still be executed in time. It only affects the ladder logic scan time.

In typical motion control scenario the pulse inputs typically varies from low to a stead state velocity and then slow down to a stop. So the impact on the ladder logic scan time will not be too significant. But if your application requires the two channels HSCs to be exposed to certain frequency range for a long time then the interference can be a problem.

To fully address this issue we will need to work on the I/O scanning firmware to limit the number of rescan that the CPU will perform if it cannot get a stable readings from the digital Input scan. This will prevent the ladder scan slow down issue you have reported.

You can send the PLC back to us for a free firmware upgrade when new updated firmware that fixes this problem is ready. Please email support@tri-plc.com directly and we will correspond with you on the new firmware release schedule and RMA for returning the PLC for a firmware upgrade.

[jcperis]

Ok, I understand your explanation. Our application requieres the two HSC channels to be exposed to that frequency range. By now we can reduce the pulse frequency, and we can work out of the problematic frequency range.

Thanks,
Juan C Peris