Wondering if anyone has experienced or knows about a glitch we have been experiencing.

We collect route data, dump the data, and upon looking at the data on the computer we find stretches where the spectra look identical except for the low end. See attached document.

This has happened in 3 different databases, with different AP sets.
Notice page 4 - The Fmax is different on A&B compared to C&D, and the spectra seems to expand/contract.

We're probably going to send the 2130 in under maintenance agreement, but thought I would see if anyone has seen anything like this?

Thanks for your replies.

Duplicates307.doc (56 Kb, 70 downloads) Duplicate Spectra

Mike,

I would guess that the problem is in the ap sets if you have fmax set to orders and it is variable speed equipment.

If not, then I don't know about the 2130. Never used one.

Danny Harvey

I have noticed in the field while taking data with a 2130, that when I clear data if it was bad (ski slope for example), the second data collection is much faster. Itmakes me think that something from the first set was used in the second set. Its as if when I choose "Clear Data", it doesn't clear all of the data... just an observation....

Looks like ski slope on the low end and order based Fmaxs on the other stuff. The 2130 sort of defaults to its own Fmax depending on the order that is set or assigned to a point through an AP Set. If this is not a variable speed piece of equipment the solution to the different length Fmaxs would be to assign a fixed fmax. Even if this is variable speed the same holds true, fixed Fmax that is within a reasonable area of an average Fmax based on history.

I totally dislike an order based Fmax. Of course this is only my opinion.

quote:

I have noticed in the field while taking data with a 2130, that when I clear data if it was bad (ski slope for example), the second data collection is much faster.

As I understand it, the reason that the 2130 is faster at data collection than the 2120 is to do with the autoranging feature; the 2130 looks at the previous reading, and uses that to decide what amplitude range to start off with. This cuts down the time to find the right range.
This probably holds true if you re-take the reading, and thus the time to find the right range will be much less.

Ralph, I'm interested to know why you don't like an order based Fmax.

Ian

I agree with Ralph. I just don't like order based either. With my variable speed equipment I just strobe it and input the speed and use the same Fmax all the time. I like to look at consistent binwidths.

Mike, I guess I don't "see" the problem... I think it is much easier to make good visual comparisons using a "strip" format with no overlap and no slant. It's just much easier to see the differences and similarities, IMO.

I also do not take order-based spectra because if you do "visual" analysis it just complicates things.

I have noticed the 2130 is much more susceptible to ski slope than the 2120 was. I'm not sure why and have not really solved that problem yet... I just retake the data if I see ski slope (not very efficient, but it takes "time" to figure stuff out and who has enough of that?).

Mike,
I am trying to follow what you were doing. The data on page 4 was taken within one minute and the data from all 4 pages was taken within 6 minutes. Are you sure the 2130 made the difference? Did you do anything different.....fmax, speed change.....remember, data taken in "analyze" with a different fmax will cause this too.

Rusty, I have been using a 2130 for a year and the "more common" ski slope is the first thing I noticed too. I have switched a lot of my pesky points that always give a ski slope to "analog-analog" and do a "STW" to get my acceleration waveform back.

As far as the order based fmax, I didn't like it with the 2120 bacause it messed up the simultaneous 2-channel data collection. I NEVER take an order based fmax with peakvue because it can bump your filter without you knowing it. An order based fmax has an advantage for regular data with the 2130 if you set your alarms the way T-A teaches.

David Eason

The problem is that the same frequencies and amplitudes are showing up on All of the points. The spectra have identical amplitudes at all the frequencies. Seems like the analyzer is holding the spectral amplitudes "in limbo" and saving it under a series of points.
I don't believe this has anything to do with particular AP sets or order based Fmax.
The point I was making on page 4 was that the analyzer seems to be holding an amplitude level in each "line" of the spectra, regardless of the fmax!

"I have noticed the 2130 is much more susceptible to ski slope than the 2120 was. I'm not sure why and have not really solved that problem yet... "

Rusty: If the 2130 is not auto ranging, but using the previous overall data for the range, then the data collection starts sooner, which means more ski slope.

IanS:
I realize the 2130 uses the previous data instead of autoranging, but I mean that it seems even faster than normal for the 2130 if I have deleted data.

quote:

IanS Posted Apr 03, 2007 7:27 PM Hide Post
quote:
I have noticed in the field while taking data with a 2130, that when I clear data if it was bad (ski slope for example), the second data collection is much faster.


As I understand it, the reason that the 2130 is faster at data collection than the 2120 is to do with the autoranging feature; the 2130 looks at the previous reading, and uses that to decide what amplitude range to start off with. This cuts down the time to find the right range.
This probably holds true if you re-take the reading, and thus the time to find the right range will be much less.

Ralph, I'm interested to know why you don't like an order based Fmax.

Ian
Posts: 81 | Location: Tasmania, Australia | Registered: 14 March 2005

Ian,

I am referring to the AP Set where the data is collected using an order based Fmax. This has a tendency to leave data with long and short Fmaxs which causes a few distortions when analyzing the data, especially in multiplot form. Same way with the waveforms using orderbased.