Weird PEC result ...

[Demonperformer]

I got a PE recording last night - NEQ6, 2m scope, 0.66"/px, using Altair (after it had passed the meridian) - which has given a weird result in PECprep.

Firstly, in the analysis screen, I tried applying the autofilter:

Although 2 of the cycles lined up nicely, the other two (red & green) are totally out of sync with them. I found that removing the high pass filter, produced a result where all four traces were at least of a similar pattern and phase, if not exactly lined-up:

Moving onto the PEC screen, when applying all four traces, I get a peak PE+ of 8.56" and a peak PE- of -6.59", which looks about right for the graph:

However, looking at the lower graph on the page, the red line is clearly the "odd one out", so I tried removing it:

This shows the three remaining traces much more aligned, and a correction curve pretty similar to the previous one. It states that my peak PE- has now reduced to -3.61" (which looks about right), but that my peak PE+ has INCREASED to 11.60" - which I can't see from the graph at all, it looks about +3/+4 to me.

But let's move on to the frequency spectrum:

This states that the biggest component is something with a period of 675.5 - which does not seem to match anything in the system! Furthermore, there is a small peak at 134.9, which is very close to the 4th harmonic of 675.5 (which would actually be 135.1). In addition, the high 470's, where we would expect a massive peak as this is the frequency of the worm gear, there is a massive trough, although the 2nd, 3rd & 4th harmonics do line-up with smaller peaks:

All-in-all, I am at a total loss to know what to make of it all. My best guess would be to ignore the high-pass filter and remove the red trace (ignoring what it says about the peak PE+ having increased). I have saved both the "EQMOD PEC file" and the "Smoothed PE file" (would guess the former is the one I should be using). If anyone has any ideas what is going-on, I would be most interested.

Thanks.

pe_raw_data_2016_10_3_20_43.txt