When I collect data I use high res readings. On 120K Fmax=1600 lines on 300K fmax 3200 lines.
The reason i can separate frequencies better and this can keep me from going back for another reading. My question is on magnitude readings does resolution make a difference.
Will I get different amplitudes if I use 400 lines compared to 1600 lines.
I use a Datapac 1500 and it takes a different TW for every magnitude and spectrum.
Also, can anybody tell me what data they collect at each point
example #1 hor IPS 1600 lines 120K Fmax
acceleration 3200 lines 300K Fmax
TW 10 orders 800 lines.
I inherited this database from last anlyst and would like to streamline it.
Thanks Dennis

The bin bandwidth won't make any difference on pure tones. However, if the speed varies slightly during a measurement, the energy will be spread across several or many bins, resulting in lower apparent levels. When considering what the effect might be, think of high frequency peaks like rotor bar. A 1% change in rotation rate of a 3600 RPM machine would result in a 6 Hz change at 100 orders. 300k cpm at 3200 lines results in about 1.5 Hz resolution, so this tone would be spread over 4 or more bins, which would result in the appearance of low levels.

Broadband noise, such as might be caused by bearing deterioration or cavitation, will vary with the bin bandwidth. The measured level will vary as a function of the square root of bandwidth. For example, doubling the bandwidth will result in a 1.414 increase in apparent level. This may lead to some major errors in analysis.

Jon
Spintelligent Labs

If you have a pure non-varying tone that is in the middle of two bins doubling the resolution will put the tone exactly at a bin. Thus, changing the resolution can affect the amplitudes at each bin location(rms should remain the same if computing this from the spectra with corrections for window function).

Good point Bill. I hadn't thought of that.

Jon
Spintelligent Labs

Great point Bill. I've been using windowing as a tool for 20 years. Some thought in PdM. Why do you need 32 wheels when 4 will do the job? I try to separate PdM and vibration analyses. A tape recorder and RTA are still good tools.

Thanks for your input. I did a test with 400 lines of resolution and 1600 lines. Took three readings each at same place and averaged.
Difference was .0018 IPS. The higher resolution was the lower amplitude.