PHD - my final success may be useful to others

[Sparrow]

Since I started guiding last Autumn, I have been plagued with guiding problems. When watching the graph in PHD, the RA would behave nicely but the DEC line would behave very erratically. Sometimes it would over correct, sometimes it would sruggle to correct at all and gadually drift away from the centre line. It rarely achieved that lovely sight of "hovering" about the centre line. I was losing about 2/3 of my subs due to poor guiding at every session and had tried everything I could think of - just about every variation of DEC settings possible. Eliminating Flex in the setup. Choosing different sized and brightness of stars.

To put this in context, I am pulse guiding through a USB to serial adaptor plugged in to the handset of my Celestron CG5 mount.

I know a lot of imagers prefer to guide direct in to the ST4 port of the mount, but from what I've read of Craig Stark's info on PHD, there's no reason why it should be able to guide just as well by pulse guiding through the handset even though it's competing signal wise with the handset's own signals.

So what was my final fix - out of sheer frustration last week after yet another disastrous guiding session, I decided to tweak the RA settings - about the only thing I hadn't tried - no logic behind it - just frustration. Why should RA adjustments make a difference when the RA guiding seems fine!! ANyway, I ended up reducing the settings for RA Aggressiveness to 60, and RA hysterisis to 15 - all of a sudden the guiding started to work perfectly and has done so on every session since.

Why? I really have no idea but do have one theory which might be absolute rubbish - I reckon that the RA guiding was "hogging" the limited and prescious available signal so that the DEC guiding signal wasn't getting a chance to get through to the mount. By turning down the aggressivenesss of the RA settings, the DEC is now getting the chance it needs to get through.

Hopefully this might prove usefull to someone. Also, someone more knowledgable than me might be able to explain why it had the desired effect.

Regards

John

[George]

DEC will either drift N or S, determine which way your guiding is correcting for and just set your DEC corrections for that direction, this will then stop any bouncing between N or S.

Should smooth your graph out.

[barkis]

DEC will either drift N or S, determine which way your guiding is correcting for and just set your DEC corrections for that direction, this will then stop any bouncing between N or S.

Should smooth your graph out.

Would a pier flip cause any problem then George, or will the software adjust for it.?

Ron.

[George]

Would a pier flip cause any problem then George, or will the software adjust for it.?

Ron.

Good question

I only ever image east side of the pier as I have a limited view westside so a definitive answer would be welcome......thinking about it you would have to recalibrate PHD after a pier flip in any case so it would be simple to change the DEC bias.

[barkis]

Good question

I only ever image east side of the pier as I have a limited view westside so a definitive answer would be welcome......thinking about it you would have to recalibrate PHD after a pier flip in any case so it would be simple to change the DEC bias.

Cheers mate.

[RobH]

I image both sides of the meridian, and after a pier flip I recalibrate. (I always recalibrate on my target if I change targets anyway, even if the previous target has been on the same side of the meridian)

For me, on the east side my DEC guising is set to south, and on the west, north.

I actually ended up deliberately slightly misaligning my E/W axis in order to get reliable DEC guiding.

When it was too accurately aligned, DEC would tend to oscillate a bit, and you needed DEC corrections in both directions, which meant backlash hadd to be taken up and the guiding wasn't as good.

Cheers

Rob

[Sparrow]

Thanks for the replies

George - I've smoothed my graph out very successfully - the point is that DEC guiding was previously very poor - not just oscillating but sometimes just drifting in one direction without any serious correction being made by PHD.

The relevant point is that with pulse guiding, by turning down the aggressiveness of the RA guiding, it allowed the DEC guiding to work properly - Guiding is now most definately sub-pixel