AVW200 Series devices with Serial Numbers less than 5768 need to be tested for a fault. The method perscribed by CS is to load the new OS (Std.04) on the device then test for a frequency reading of -555,555.
For this test can the Troubleshooting panel of the Device Configuration Utility be used, i.e. Poll each of the two channels for a reasonable response? I did this for one channel with an attached VW sensor and got a classic spike in the spectrum window, along with a logrithmicaly declining amplitude plot. However, for the other channel without an attached VW I got no spectrum plot and in the Time Series amplitude plot I got a band of noise that drifted monitonically upward during the 0.3 second duration.
It is helpful to know if the Device Configuration Utility Troubleshooting method is valid because it is faster, avoiding the need to hook the AVW200 up to a datalogger to collect vibration data. Is this method OK?
Yes, you may use the Troubleshoot tab/panel for the AVW200 series in the Device Configuration Utility to look for the -555,555 frequency after loading OS Std4. Watch the result box called "Peak Frequency" (located between the Spectrum and Time Series output frames) for -555,555. The behavior you are describing for your second channel (with no VW sensor attached) is a typical response for a non-connected state. With Std4 OS loaded, the "Peak Frequency" box should show NAN when no VW sensor is connected.
Yes, feel free to check your devices with DevConfig, bypassing the need to connect them to a datalogger, etc.
If you are checking a large number of devices, you have the option of not plotting the Time Series data, and then the polling process will go faster. Just uncheck the "Poll Time Series Data" box in the Measurement Options screen which is brought up by the Options button.
I've about 35 AVW200/206's with serial numbers in the 1K to 2K range. I'll update them to OS4 and testing them out. Just to be sure I understand, the units should be tested with a vibrating wire sensor attached to each channel?
Thanks
If you are only testing for the -555,555 problem, you don't need to have any VW sensors connected to the AVW200 series device. The check is performed internally on the device by the OS and is independent of whether sensors are connected or not. Please go ahead and use the Troubleshoot tab in DevConfig to do this.
If Std4 OS is loaded, the device will give NAN when no sensors are connected and the problem is not present, or will give -555,555 when no sensors are connected and the problem is present. When operating sensors are connected you will either see a a frequency if the problem is not present, or you will see -555,555 if the problem is present.
If you see -555,555 for the frequency on Channel 1, there is no need to check channel 2 (as it will always have -555,555 for the frequency also).
If Std3 OS or earlier is loaded, you will see a frequency regardless of whether a sensor is connected and regardless of whether the problem is present or not.
Thanks,I've updated a dozen or so units to OS4 so far, no sign of the -555,555 yet. Is there any reason to believe that this possible hardware issue would effect the accuracy of the vw measurements? I've not seen anything weird, though my database people have spent a considerable amount of time trying to reduce the temperature effects on our vw pressure transducers readings. The sensor manufacturer provides a temperature compensation formula for the sensors, but our people are attempting to improve on it.
Regards
Everything we have seen shows that the AVW200 devices measure accurately up until the point that the component issue comes into play (for those few instances where it does occur).
Once the issue occurs, the measurement process can (in some cases) immediately isolate a different frequency than that of the VW sensor and it appears to remain very close to that frequency after that time. With the Std4 OS installed, you will see -555,555 for the frequency when that happens. With an earlier OS, you will suddenly see a frequency that is not correct (frequency step or jump), which you would have to say "affects the accuracy".
However, if I understand your question properly, I think the answer is No. We don't have any observations that the device gives readings that are "slightly" or "moderately" inaccurate during the time before the component issue occurs. They should be accurate up until such time that the issue occurs.
This all underscores the need to run the Std4 OS in
all devices as soon as is reasonable.
So when I update to OS4 and run the poll the -555,555 will show up for certain if the unit has hardware issues?
I'd like to be sure on this, my current projects are ending after 2 1/2 years and I won't be able to test them in the field until the next deployment.
Thanks
The answer to your question is Yes.
Once the component problem has occurred on the device and OS Std4 is loaded, the frequency will show as -555,555 regardless of which channel is polled, regardless of what method is used to poll the device, and regardless of whether a VW sensor is connected or not.