I have some low speed readings that I would like to turn off the data overlapping for using my 2120's and RBMWare.

Does the 2120 always take that instruction from the firmware or is there somewhere in the AP set that I can change it?

I think the only place to change this is in the 2120 itself. I don't think it can be changed from the AP set.

Ian,

I don't think so either.

Danny,
What you are trying to gain by having no overlapping? What is the RPM of your machine?

David

David,

.5 rpm probably explains everything.

Danny,

You mean you can't set overlap up in the softwae? it's a hardware function?

Dave

Danny,

You are probably going to use PeakVue, where just 1 batch is taken anyway. In this case you don't have to worry about overlapping. I also have low RPM applications ( 8 RPM) and will be interested in seeing your data if possible.

David

Dave,

I guess so, but it has never presented a problem because I have never needed to change it. I just have to remember to change it back for regular data acquisition.

David,

I am trying to convince myself of the idea of just one average so I'm getting some comparative readings.

I know Peakvue is a "given" for many of you on slow speed machines, but I'm not sure it is enough. Just because you only have 1 rpm, doesn't mean you don't have many events in that revolution. I think it is worthwhile to take a regular, averaged reading along with your peakvue reading.

Danny, what will you gain by not using overlapping? Why do you want to do this?

50% overlapping is generally the standard for most but I think CSI uses 67%. Is there a loss - yes to some degree but it speeds things up using overlapping. Using RTA's that have zooming is good. a minimum of 16-32 avgs if averging is used. LOR is also a problem probably - why average at all. Are you doing this on a route, or, is it a trouble-shooting problem for every forty fortnights.

quote:

Originally posted by Danny Harvey:

.5 rpm probably explains everything.

Danny,
I don't think low rpm by itself necessarily warrants NOT to use overlapping.

In general, if you have 2 avg with 0% overlapping vs. 3 avg with 50% overlapping you still have same amount of data in the time domain with only difference that in the latter case you have more averages, hence you are closer to true avg.

Rusty,
Will normal signal processing provide you with any additional to enveloped methods vibration information for low speeds? There is no much vibration going on in the range of first 8-10 orders and even 60 orders of RPM ( for bearing defects detection) at 0.5 RPM

This is on some big slow reactor bearings that have shown indications of significant wear through grease analysis, but I don't see anything with PeakVue or conventional vibration analysis. The integrity of the grease samples are questionable, though. The bearings have heat medium pumped through the pipe/shaft that is way over the maximum temperature for the bearings.

I plan to collect PeakVue with 1 average, 3 overlapped averages and 3 non-overlapped averages. Any suggestions as to filters? I have been using a 500 hz hp filter but might try some of the band-pass filters. It's very hot and high work, so I don't want to spend too long doing my experiment, but I need to get the right result. It may turn out that the grease samples were crossed or contaminated with the grease samples from the big, worn, open gearing, too.

Let me clear something up about the speed.

I was considering 2 different machines one at .5 rpm and one at 5-7 rpm. The one at 5-7 rpm became more important when I got the results from the latest grease analysis so I sort of switched horses in mid stream on you. Sorry for the confusion.

Maybe try some non-conventional things; structure wobble - spot through a scope to view visually deflection from ground level. Does this job have anything to do with the #2?

For PeakVue which recommends just 1 avg:
for a 6 rpm you may try
Fmax=10 HZ and NL=1600. This will give you T=160 sec worth of data and 16 revs.
As far as filters.. HP=500hz is one of them, but you have to try BP as well. So, drink a lot of water.

Sam,

There are 2 reactors with 2 bearings each and I use 2 acceleromters to collect 2 readings at a time and have 2 data collectors and a few other things that have to do with the #2.

Was there something specific you were referring to?

David,

Is the 10 hz fmax for PeakVue new? Maybe for the 2130?

I did not get your question, Danny? You can set all that in 2120 as well.

In the 2120 and the Consultant there is an indication during the boot up. It will say HV1, HV2 or HV3. These refer to sets of discrete fmax's that can be used for the various types of data acquisition. As far as I know 1200 cpm is the lowest available fmax for all three sets. If you set it lower, it will default to the next highest. No real problem though, I'll just adjust my lor accordingly for a 64 sec. sample

Thanks,

In 2120, if you go to:
Off Route Program/Analyze/Acquire Spectrum, you can setup any Fmax and the rest of parameters.
I think 64 sec sample is not sufficient for a 6 rpm shaft.