Analysing PE with PECPrep

[MG1]

This is really the continuation of a question I've asked as part of another thread on a slightly different subject, but thought I would post it in a new thread so it doesn't hijack the other...!

I'm trying to establish the periodic error of my setup and have recorded data in PHD to bring into PECPrep v1.78. The odd thing I'm having is that it is only giving me a small sample size, when in fact I thought I had collected data for around 12min.

Anyway, I have attached the original file if anyone would be able to assist by looking at it and seeing if I'm doing something wrong?

Data was captured using a lodestar with pixel size 8.2x8.4um on a William optics ZS66 (388mm f5.9)

There was no binning and the guide star was Capella (Dec 45.982778)

Thanks in advance and maybe see some you in the next few days.

michael

PHD_log_27Mar11no output.txt

[chrisshillito]

There are some oddities in your PHD log.

For some reason PHD hasn't written DECDistance and StarMass data for many of the log records. This was making PECPrep think those records were invalid and so it only used the ones with DECDistance present and interpolated between the gaps in the data (using a 1s interval).

The log file has three separate "guiding" sessions contained within it - two being identical (have you been in there with the scissors and gluestick ). Probably best to always start with a clean log rather than append to existing ones if you can.

Anyhow I've just released PECPrep version V1.79 which now manages to cope better if some of the PHD log record fields are missing.

Attached are two PECPrep screen shots of your data (with declination angle applied). As you can see the frequency analysis is now much better and the stepper gear peak is easily identifiable.

Parameters used: FL=388, PixelSizeX = 8.2, PixelSizeY= 8.4, declination=46, DecCompensation = On

I note that the parameters stored in the PECPrep output file you provided in the other thread (PHD_log_27Mar11no output - Copy2_RA_EQMOD.txt) were quite different:

fl=995, PixelsizeX=6.4, PixelSizeY=6.4 Declination=0

Look forward to seeing some longer duration data.

Chris.

[MG1]

Cheers Chris....Yes I messed up the parameters. I had been playing with APT and framing images and had my imaging scope in my head, not the guidescope.

So my PE isn't very good afterall!

Still I will capture more data and try to plot the curves using the correct pixel size and focl length.

Is the new software up on the Yahoo group? I'm a member over there too so will check it out.

Thanks again...this is a super bit of software for fiddlers like me....I just need to be a little more methodical in my setup and not rush!

Cheers

Michael

[chrisshillito]

For a Synta mount your PE is very good! - just not as unbelievably good as you first thought

PECPrep V1.79 should be available via Yahoo EQMOD group and Sourceforge.

The thing to remember with PECPrep is that it works best when it has lots of data to play with. The longer the capture duration and the shorter the sampling time the better. Choose a bright star (like you did) and use the shortest exposure that the recording program can track with, ideally not too high in declination (PECPrep will try to compensate but can only do so on pixel movements the data the guider can detect - at dec=0 you get a greater pixel displacement for a given angular movement). You can then record your PE data during dusk when there isn't much else you can do as PECPrep will cut through the poor seeing conditions.

For those seeking to hypertune PECPrep can help in identifying areas of likely mechanical vibration/eccentricity and in measuring changes. That said PECPrep is limited to dedicated tinkerers, those who have no intention of touching the mechanics will still benefit from an understanding of what the current limits of performance are.

Chris.

[MG1]

For a Synta mount your PE is very good! - just not as unbelievably good as you first thought

PECPrep V1.79 should be available via Yahoo EQMOD group and Sourceforge.

Chris.

Yes...that would have been amazing to get paramount performance, just by changing bearings...

I've just had a look but could only see v1.78 on yahoo and v1.77 on sourceforge. Do you have a link?

Thanks

[chrisshillito]

They are there now. Not sure what happened there (on both sites), perhaps I managed to delete the new ones rather than the old!- Guess I'm doing to much multi-tasking.

Chris.

[MG1]

Got it thanks...

You're doing a grand job with this software....and supporting it too!

Many thanks

Michael

[MG1]

I finally managed to squeeze in some more data on this mount and here is the raw phd log file...over two hours to crunch...why wasn't I imaging you ask? Well you'll see by the amount of noise that it was pretty windy and also later on fairly cloudy.

I picked vega to guide on though and the mighty V shon through the cloud to enable me to keep going on a single fresh log file.

Again this was with the 66 so Fl=388, no barlow, 8.2x8.4px and zero binning

I think Vega Dec is 38.78369 (J2000)

Would be good to see what you make of this. Oh and I should note it was a pretty quick Polar alignment using the EQmod align tool.

Off to bed now and hope SGL6 goers don't wake too early in the morning.

Cheers

PHD_log_02Apr11.txt