Hot loop check of proximity sensor on site

Hello all
I have a question related to problem we face in calibration of proximity sensor in SGT100 gas turbine vibration system.
The proximities are Bently Nevada and we have a monitoring system for protection. We are not sure about displacement values which are monitored in monitoring system and that’s why we are searching for a method to calibrate the measurement chain(cards) in site. At the first step we measured the output voltage of proximeter in stop mode and convert it to gap between shaft and probe. But we do not have any sense about this numbers and relation of these numbers with monitoring data.
My question is that: is the gap voltage in stop mode of gas turbine equal to zero journal and thrust bearing vibration monitoring?
I just want to know which is reference for gap voltage setting (DC offset): the regarding voltage distance between shaft and probe in stop mode? Or the center of linear range of proximity?
And do anyone have any idea of how could we make sure about our monitoring data? What is the hot loop check of proximity sensor in site?
thanks
 
The monitoring system is the area of focus not each sensor. I’m assuming you have a 3500 rack.
Ultimately for proper calibration you need to pull in a Bently Nevada engineer or agent.

Each proxy sensor needs to be mounted with a physical gap between sensor head and shaft according to specifications. The gap voltage doesn’t relate at all to the physical gap which is why the numbers don’t make sense. Generally (as I recall) they have preset gap voltages around 7.5 to 9.5vDC.

I’m guessing hot loop check is connection of oscilloscope to each rack amplifer - not sure if input side or output side. The gap voltage is DC, once vibration is detected this is modulated on top of the gap voltage. So hot loop check you need not only to create vibration, but to control it’s frequency as the rack amplifier is frequency sensitive.

I worked on the calibration of Rockwell monitoring system - very much easier to understand, but even so the frequency sensitive amplifiers were quite complex.
 
If one wants to observe to understand how the B-N proximity sensor output varies with proximity, you need a wobulator. They can be rented or purchased, or you can have someone with a wobulator visit site and help with the checkout. You could use an oscilloscope to look at the waveform, and a storage oscilloscope to record the waveform for later purposes. But, I also gotta believe there are probably LOTS of YouTube videos about this...!

Gap is usually physically checked/adjusted after the proximity sensor is installed in the bearing housing, and set using lock-nuts/jam nuts.

For the proximity reading (output) it's necessary for the monitor to know the speed of rotation (the keyphasor input).

If you have serious problems with the data, you can have someone come with an ADRE system and they can do some very amazing analysis of the proximity readings (and the keyphasor and any other B-N equipment).

Finally, it's my understanding there's no "calibration" to do--adjust the gap voltage (a NEGATIVE DC voltage) to the proper range, and as long as you have the correct cable for the proximity probe and the proximitor in use, that's it. No calibration required. If you are talking about confirming the output of a particular proximitor probe and proximitor, that can be done by removing the suspect probe and prox and using them with a wobulaltor.

I'm NO B-N expert; but I spent man-weeks on the first couple of start-ups I was assigned to "calibrating" prox probes, using a wobulator. A friggin' waste of time. Of the seventy-odd probes I documented not a single one was bad. And, if a probe is bad, you will know it.

If you want to confirm the readings being provided by a B-N prox/probe combo, you don't do it by calibrating anything. At least that's been my experience. There's no pot on the prox or probe to adjust. And, as long as the cards in the monitor rack are not alarming because of issues, and are configured properly (it used be jumpers and one or two pots decades ago; I think it's all digital/software now) then the reading is what it is (presuming no wiring issues ...).

I have seen probes/proxes removed from a machine and verified to be bad using a wobulator, but the monitor rack cards they were connected were already alarming a problem with devices.

I have to imagine there is some kind of "static" test which the monitor rack cards can be run through to check them for errors, but sometimes my imagination is pretty far off from reality. (Not usually; but occasionally.)
 
The monitoring system is the area of focus not each sensor. I’m assuming you have a 3500 rack.
Ultimately for proper calibration you need to pull in a Bently Nevada engineer or agent.

Each proxy sensor needs to be mounted with a physical gap between sensor head and shaft according to specifications. The gap voltage doesn’t relate at all to the physical gap which is why the numbers don’t make sense. Generally (as I recall) they have preset gap voltages around 7.5 to 9.5vDC.

I’m guessing hot loop check is connection of oscilloscope to each rack amplifer - not sure if input side or output side. The gap voltage is DC, once vibration is detected this is modulated on top of the gap voltage. So hot loop check you need not only to create vibration, but to control it’s frequency as the rack amplifier is frequency sensitive.

I worked on the calibration of Rockwell monitoring system - very much easier to understand, but even so the frequency sensitive amplifiers were quite complex.
Thank you for your answer.
Our monitoring system is rockwell ,IRD entek 4500. There are potentiometer on it to set offset and span
But I do'nt know how can I find the gap voltage.
Also I do not know if having information about gap voltage will help us for calibration or not?
As I understand for hot loop check, we can simulate the signal with an oscillator, so we should know the amount of DC voltage for simulating. Am I correct?
 
If one wants to observe to understand how the B-N proximity sensor output varies with proximity, you need a wobulator. They can be rented or purchased, or you can have someone with a wobulator visit site and help with the checkout. You could use an oscilloscope to look at the waveform, and a storage oscilloscope to record the waveform for later purposes. But, I also gotta believe there are probably LOTS of YouTube videos about this...!

Gap is usually physically checked/adjusted after the proximity sensor is installed in the bearing housing, and set using lock-nuts/jam nuts.

For the proximity reading (output) it's necessary for the monitor to know the speed of rotation (the keyphasor input).

If you have serious problems with the data, you can have someone come with an ADRE system and they can do some very amazing analysis of the proximity readings (and the keyphasor and any other B-N equipment).

Finally, it's my understanding there's no "calibration" to do--adjust the gap voltage (a NEGATIVE DC voltage) to the proper range, and as long as you have the correct cable for the proximity probe and the proximitor in use, that's it. No calibration required. If you are talking about confirming the output of a particular proximitor probe and proximitor, that can be done by removing the suspect probe and prox and using them with a wobulaltor.

I'm NO B-N expert; but I spent man-weeks on the first couple of start-ups I was assigned to "calibrating" prox probes, using a wobulator. A friggin' waste of time. Of the seventy-odd probes I documented not a single one was bad. And, if a probe is bad, you will know it.

If you want to confirm the readings being provided by a B-N prox/probe combo, you don't do it by calibrating anything. At least that's been my experience. There's no pot on the prox or probe to adjust. And, as long as the cards in the monitor rack are not alarming because of issues, and are configured properly (it used be jumpers and one or two pots decades ago; I think it's all digital/software now) then the reading is what it is (presuming no wiring issues ...).

I have seen probes/proxes removed from a machine and verified to be bad using a wobulator, but the monitor rack cards they were connected were already alarming a problem with devices.

I have to imagine there is some kind of "static" test which the monitor rack cards can be run through to check them for errors, but sometimes my imagination is pretty far off from reality. (Not usually; but occasionally.)
Hello
I really appreciate your complete answer. That was very helpful.
I have some questions regarding your answer:
1- I am completely agree with you that probe and sensors do not require any calibration, but we are not sure about configuration of our cards and the values which are monitored do not make sense for us. So I am searching for a method to confirm the monitored values. As you mentined we can use oscilloscope for recording the output of proximeter, but after that we need to know the gap voltage.
Knowing gap voltage will help us to obtain the offset value and eliminate it to gain AC voltage amplitude and convert it to displacements.
2- does the output voltage of proximity in stop mode provide us any specific information about gap voltage ?
Thank you in advanced
 
According to my information, CSI 4500 is an Emerson product. Non of the Entek range I came across had potentiometers - parameters are all pre-programmed. Offset on Entek was non-adjustable - approx. 11.5vDC for Supply 24vDC.

Sounds to me you need a Manual - I’ve never found Emerson all that helpful; plus do what I did - have the system on a bench with oscillator, scope and recently calibrated voltmeters.
 
RTFM

Read The F-F-F-F-F-Fine Manual
Hello
Thank you for your suggestion.
Our problem in field is that one of the measuring point of proximity suddenly disconnected during turbine runup. I've checked peak to peak voltage of all journal bearing measurement point and all of them are in the range of: 9.2 to 10 except one which is in the range of 8.2 to 8.4 , it sometimes works correctly but occasionally cause turbine to stop. Also the DC voltage for this point in stop mode is 7.9 whereas for others is 9.2. can this difference cause to stopping it? I ' ve attached the trend of this point for you to give me direction. Thanks
 

Attachments

Hello
Thank you for your suggestion.
Our problem in field is that one of the measuring point of proximity suddenly disconnected during turbine runup. I've checked peak to peak voltage of all journal bearing measurement point and all of them are in the range of: 9.2 to 10 except one which is in the range of 8.2 to 8.4 , it sometimes works correctly but occasionally cause turbine to stop. Also the DC voltage for this point in stop mode is 7.9 whereas for others is 9.2. can this difference cause to stopping it? I ' ve attached the trend of this point for you to give me direction. Thanks
Hello

Do you have any OEM instrumentation list /setting sheet..
Also what is vibration level during these phases that you are referring...
 
The story keeps changing.

Please work with the suppliers to get their help with [whatever] the issue(s) [is/are].

It seems to be some kind of cobbled together system that may or may not work well together. And with this newest partial piece of information (... one of the measuring point of proximity suddenly disconnected during turbine runup. ...) we really don’t know what happened. Why did the “disconnection” occur? If the voltage isn’t correct is the wiring good? Did the electronics experience some damage resulting in failure?

Best of luck with the “problem.”
 
Hello

Do you have any OEM instrumentation list /setting sheet..
Also what is vibration level during these phases that you are referring...
Vibration level is near 20 microns and the shutdown level is 150 microns. I 've just attached a setting sheet of card.
It is mentioned that offset is 12 volts. But our measured voltage of proximity is around 8.2-8.4
For others points that we do not have any problem the voltage is around 9.7-10
It could be worth mentioning that sometimes the turbine take shutdown with x probe and sometimes with y probe of PT turbine. We do not have problem with probes of GG shaft.
 

Attachments

The story keeps changing.

Please work with the suppliers to get their help with [whatever] the issue(s) [is/are].

It seems to be some kind of cobbled together system that may or may not work well together. And with this newest partial piece of information (... one of the measuring point of proximity suddenly disconnected during turbine runup. ...) we really don’t know what happened. Why did the “disconnection” occur? If the voltage isn’t correct is the wiring good? Did the electronics experience some damage resulting in failure?

Best of luck with the “problem.”
Dear sir
I do'nt think we have problem with wiring. We have checked all the wiring and connection. We also changed the probe measurement chain to another way which measured correctly, and the measuring voltage remained as before in range of 8.2-8.4
Our proximeters are 7200,9 meter with extension cable of p/n 21747-065-00
I do not know if this type of extension cable is suitable for this type of accelerometer. I 've searched but did not find any information
 
Do you mean Offset is 12 mV not 12V can you reread your post ...
I meant 12V, it was mentioned in document that I 've attached in previous post.
Maybe the problem is incompatibility of proximety and cable.
Our proximeter are 7200/ 9 meter
And our extention cable for x and y are as below:
1- P/N 21747-065-00. S/n SEP Y244311
2-P/N 21747-065-00 S/N SEP Y244303
Are these two extension cable differenet?
I could not find any information in internet about these cables
 
I have some answers to your questions:

Have a read and carreful read ! on this document as all is well described and can respond to your questions about cable Extension (not extention)
Looks like there is complete intercompatibility betweens probes /sensors & cables

There is also "How to order Extension cable with part& serial number..

Any time !
Always gald to support here!

James
 

Attachments

Top