PHD2 very spikey graph
Andyb90
bw_msgboard
Hi Andy. There seems to be a lot going on here, much of it not very good. L If you haven’t already done so, you should download and study the tutorial on analyzing guiding performance:
http://openphdguiding.org/news/
Your most serious problems are very large, spontaneous excursions in both Dec and RA, events I refer to in the tutorial as “gremlins.” Just to pick a few examples:
- A 40 arc-sec displacement in RA at 20:42
- A 100 arc-sec RA displacement at 20:52 and a 20+ arc-sec Dec displacement at the same time
- a 7.5 arc-sec slow displacement in Dec at 21:30
These are essentially fatal guiding problems, they can’t be corrected by PHD2 and they will ruin any frames taken with the main camera. They are mechanical issues with the mount or the gear riding on top of it, not something PHD2 has done to you. There is no easy solution here, you simply have to find them and fix them. My guess is that something is loose somewhere but it’s only a guess. Keep in mind, a 40 arc-sec shift is equivalent to 35 microns with your set-up, so it’s not like something is about to fall off the telescope. Since you were guiding far to the north with the telescope presumably pointed way off vertical, the likelihood of slipping, sagging, or shifting is increased. Focuser assemblies are definitely possible sources for these kinds of problems.
Once you eliminate these problems, there is another layer of performance you may want to look at. The Guiding Assistant results point to a couple of things:
- You have a very large amount of Dec backlash, not a big surprise for these mounts I think. This means it probably isn’t practical to try guiding in both Dec directions, and I see you went with north-only guiding during the last run. You may be able to improve the backlash by working on the gear mesh in the mount but I don’t know how likely that is. There is a large south excursion in Dec during the last guiding sequence, and even that is very slow to respond. With only a short interval to look at, I don’t know what to make of that.
- It also looks like you may have a large amount of uncorrected periodic error in RA – that also shows up in the GA run. Even with just 145 seconds of measurement, the unguided RA excursion was 13.5 arc-sec. This might translate to a 25-30 arc-sec periodic error during the full worm period, or even worse. If you haven’t applied a periodic error correction to the mount, you probably should. Your results may improve if there is less work for the guiding system to do. Even with these issues, you were still getting periods of guide performance with 1.3 a-s RMS in RA. That’s not too bad – it’s just a question of whether you can do better by improving the mount’s native performance.
When running tests, you don’t want to be working at a high declination like Ursa Major. It’s much better to work close to Dec=0 so you can get a better view of the way the mount is responding to guide pulses. This usually reduces the mechanical stress on components that are attached to the mount. Your polar alignment is excellent, by the way, but this may perversely work against you if you need to guide in only one Dec direction. You may end up intentionally mis-aligning a bit more to establish a consistent pattern of Dec drift. If you keep the mis-alignment under 5 arc-min or so, you are unlikely to see field rotation problems.
In any case, these are second-order problems, things to be dealt with after the gremlins are fixed.
Good luck,
Bruce
--
You received this message because you are subscribed to the Google Groups
"Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.