I’ve encountered a case where a protection vibration signal was lost just before a trip took place ( the trip was caused by an adjacent point). Part of the analysis of the data revealed that, in Bently Nevada system ,when some problem happens to the field loop (like open circuit) and the “Not OK” light is on, the recorder output will maintain a value of 4 mAmp. The reading (for the receiver) will be a low value but not abnormal. Also, during this period of “zero vibration”, the Bently protection from the rack be will be by passed

I wonder if it is acceptable for major machines, protected by recorder output, not to be connected to shut down relay if the vibration signal is lost from the field. What about machines connected to racks from other vendors- do they have same problem?

Any experience in this regard is appreciated.

Thx-Ali M. Al-Shurafa

Yes this is pretty standard. You can trip on Not OK, but that is not ususally done. It is possible for a sudden event to take out the probes putting them into not OK and then the trip never happens since they drop into bypass. But that is pretty rare. On some machines I have put thrust probes into two different racks and then voted the OK circuits, so if both thrust probes go not OK the unit trips, but I would only do this on a very very critical machine, one where there is a high chance of collateral damage if the machine wrecks.

There is also a time ok channel defeat function that will hold the probes in bypass until ok has been re-established for a certain amount of time.

Oh, and if you talk to Bently you will find that the recorder outputs were never designed for safety shutdown functionality. They are for data information only. If you are hooking the 4-20s off the Bently rack to a shutdown PLC you are taking a risk because they were not designed for that. Shutdowns should only be done off the relays.

Hello Steven. Thank you for your usual help.

FYI, this machine was an 53 krpm Elliott air compressor whose panel contains all protection elements (including the vibration). Well, yes the main vibration protection is through recorder output signal- a fact I had to accept. Maybe as you described: unusual.

I’ll check with our DCS people if it possible to connect the Not OK signal to a DCS alarm and put it as a permissive to start the machine.

A related point here is the latching of Not OK relay. Dose this also latch the vibration protection pay pass or if the signal return to normal the protection is activated automatically?

Thank you.

I don't know about 3500, but on the old 3300 if you had a latching ok relay the way I remember is that you get a green flashing ok light, and you have to punch the reset on the front panel for it to come back into ok.

I know I have hit that reset button many times over the years getting ready for startups after having probes out of the machine.

On the Bently site one can find the following regarding the 3500 series monitor relay.

quote:

Notes:
1. Dozens of possible event conditions within the rack can drive the NOT
OK relay. For this reason, it is not intended for use as part of a machinery
auto-shutdown circuit and should be used for general annunciation
purposes only.

I would investigate setting alarms on the gap reading and use that as a parameter. 'Not OK' functions off the gap reading for the proximity probes. There are probably several factors to investigate; the filter for the gap has an associated time constant, which implicitly indroduces a delay into the measurement - alarms also have delays which can be set. The product manual and or vendor may provide further technical information.