Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment   Page 2 of 2      Prev   1   2
OLi

Sr. Member
Registered:
Posts: 1,916
Reply with quote  #16 
Maybe it was using the demod circuitry in the 2120 thus being semi analog hybrid solution where you may have a diode voltage drop on the negative side depending on the design and later went for full digital solution or a signal rectifier that would not leave anything negative. If it was like that, just a suggestion.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #17 
The CSI 2120 can measure a Demodulated (Demod) waveform and spectrum and the PeakVue waveform and spectrum. I have a measurement location on a speed increasing gearbox on a small Hydro turbine that I've been monitoring since 1995. I measure wideband acceleration, Demodulation, and PeakVue. The demodulation and PeakVue measurements are made with the same high pass filter setting, but there is no indication they share the same circuitry or calculations. There are interesting visual differences between the two types of measurements. In both waveforms there are negative peaks. I would still like to hear from somebody about the current incarnation of the PeakVue measurements.

Walt
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #18 
Quote:"My bottom line question: Do you guys agree I should just ignore the references to peak-to-peak PV TWF since those references are either outdated or inaccurate?"

I see, in the link you provided, the paper was written in 2001 and some of the actual waveform examples were dated as far back as January 1998..

I talked to a couple of guys who still use these older version, including MTWin and 4.61 RBM in order to get a comparison of the same data being opened in different versions of software.

I attached 2 old Peakvue waveforms, same data, dated December 04, 1998.

One that is displayed with version 4.61 and the other with Version 4.81. One can see that something has been modified since 4.61 was built.

If you would notice the positive and negative values in 4.61 have been added together, or so it seems, to show a total positive value equal to the sum of the “old” positive and negative values. IMO.

It appears as though the negative display has been eliminated from data after the paper was written.

To me this means some things written in that paper from 2001 may now be null and void, after some 16 years. IMO.

I personally would not even try to trend a Pk-Pk parameter now.

Attached Images
jpeg Peakvue Same Date as 12-04-1998 but with Version 481 Bu...eptember 2003.JPG (100.21 KB, 13 views)
jpeg Peakvue dated 1998.JPG (69.86 KB, 13 views)


__________________
Ralph Stewart
AlertAnalytical.com

MichaelS

Member
Registered:
Posts: 18
Reply with quote  #19 
Ralph stole a bit of my thunder, but I noticed that PLOTDATA does a better job of putting the most negative point at 0, where the Vibration Analysis module allows the bottom of the PeakVue waveform to float. The PLOTDATA plot has a value for DC offset, which VA tab apparently does not take into account.

Attached is the same plot, with the cursor at the same point, with a difference = DC offset. So it makes a bit of sense to me that pk-pk would provide a more consistent value across various software versions.

Attached Images
jpeg 5556pd.JPG (158.01 KB, 13 views)
jpeg 5556va.JPG (164.58 KB, 13 views)

Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #20 
"I personally would not even try to trend a Pk-Pk parameter now." -- Ralph

Actually Ralph your data shows no difference between the peak to peak value and the peak value with DC offset added. The only difference between the two plots is the application of DC offset for visual effect.

I agree with Michael that continuing the use the peak to peak value is still consistent with older data (and possibly Alert/Danger levels) whether or not a DC offset is applied.

I am still curious what the PeakVue waveform looks like in the 2140 and the 2130 analyzers; do negative values appear on the analyzer display? The PeakVue analysis process is carried out in the analyzer and not on a PC software. Is there any evidence that the PeakVue process was actually changed since the 1990s?

Let the User be aware that any published Alert/Danger PeakVue levels should be considered guidelines only!

Walt
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #21 
Walt's Quote: "whether or not a DC offset is applied."


Thanks Walt.

Can you explain to me where the DC offset difference is coming from in these 2 plots? [smile]

Thanks and Have a Great Day,
Ralph

__________________
Ralph Stewart
AlertAnalytical.com
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #22 
"Can you explain to me where the DC offset difference is coming from in these 2 plots? [smile]"

First of all, it is your data and not mine. I have no clue where the DC offset value was either entered or automatically part of the amplitude scale. I am assuming that the data is the same for both plots. The plot on the left shows a plus peak of 2.35 with a DC offset of 0.255. The sum of those two values is 2.605. The plot on the right has a plus peak value of 2.608 with no DC offset specified. The values of 2.605 and 2.608 are quite close. I assume that this is not a coincidence. The key to me is what the analyzer display shows.

Walt
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #23 
You are right, the data I am asking about is my data.

The data is from a 2120-2 meter, taken only one time, December 1998. The only difference in the plots is that they were opened and plotted with 2 different versions of software and therefore I had no control over the peak value numbers or the DC value being on one and not on the other since I changed nothing but versions of software that opened and plotted the data.

That leaves only one conclusion I can accept and that being, the differences in the 2 version of software.

What I do not understand now is where you are getting these numbers from?

Walt's Quote:
"First of all, it is your data and not mine. I have no clue where the DC offset value was either entered or automatically part of the amplitude scale. I am assuming that the data is the same for both plots. The plot on the left shows a plus peak of 2.35 with a DC offset of 0.255. The sum of those two values is 2.605. The plot on the right has a plus peak value of 2.608 with no DC offset specified. The values of 2.605 and 2.608 are quite close. I assume that this is not a coincidence. The key to me is what the analyzer display shows." [confused][confused][confused]

__________________
Ralph Stewart
AlertAnalytical.com
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #24 
Ralph,

Left Plot = Older Version
Right Plot = New Version

Left Plot on right side:
Pk(+) = 2.35
DCoff = 0.2550
RMS = 0.829
CRESTF = 2.84
[2.35 + 0.255 = 2.605]

Right Plot on right side:
Pk(+) = 2.608
RMS = 1.066
CRESTF = 2.446

The only numbers that do not match are the Crest Factor = Peak(+)/RMS and the RMS values differ slightly. Where is Ken Piety or Dr. Robinson when you need him?

Walt
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.