Not another PHD graph

[chrisshillito]

Chris, could you answer my question..

So if using a QHY5 which has a 1280 x 1024 pixel CCD with 5.2um pixels and the trace is zig-zaging a pixel either side of that line - is this going to have much effect in real terms of movement on the mount and resulting images on a typical f5 telescope ?

Or are we all getting hung up on trying to get the perfect graph ?

I'll could try ,

But first I would need to know:

1. the declination are you guiding/imaging at

2. the focal length are you guiding at

3. the pixel size and dimensions of your guide camera (ok you've given that)

4. the focal length are you imaging at

5. the pixel size and dimensions of your imaging camera

6. the scale of the PHD graph (I assume 1 pixel corresponds to 1 pixel of movement?)

With this info you can calculate how much movement 1 pixel on the graph actually represents and then you can determine whether this movement corresponds to a shift of greater than 1 pixel in your final image. Of course depending of you imaging resolution you may still find that seeing fluctuations will swamp and residual error form the guiding loop.

Ultimately the only thing that matters is what the imaging camera sees not the guide camera. If there is a problem with movement in the image then the PHD graph is one tool that may help with the diagnosis but as you say it isn't something to get hung up on if there isn't a problem to solve.

Chris.

[r3i]

Hope this isn't a stupid question but if I'm guiding at say 3 seconds should I also be doing the calibration at 3 second intervals?

The reason for asking is if I started guiding at a patricular rate then decided to alter this to see if this improved things or not, should I first recalibrate at the new rate?

[chrisshillito]

The purpose of calibration is for PHD to learn how responsive the mount is and in what direction. Provided PHD includes some settle time in its calibration routine (so that it waits for a few exposure to complete prior to calculating the movement scale) I wouldn't see the need to recalibrate. If in doubt calibrate on a nearby bright star and use a short exposure. This way it won't matter if you subsequently lengthen the guide exposure.

If you change guide rate then you should in theory re-calibrate as PHD won't be applying the "correct" corrections any more - however if PHD isn't guiding well then changing parameters like guide rate on the fly simply acts as a kind of gain control - to boost or dampen the response.

If you ever significantly change target position you should always re-calibrate as the direction and scale of movement is subject to change in different areas of the sky.

Chris.

[themos]

I never change guide rate but I sometimes reduce the guiding "gain" to 40% after PHD calibration. That gain is the sliders (one for RA, one for DEC) you get on the main EQMOD window, below its own guiding graph.

Maybe I should not be doing that...

[martin_h]

I always calibrate PHD set at 1 sec loops, but guide at 3 sec loops, no reason for this its just the way I do it, and I always re-calibrate after a significant position move.

[chrisshillito]

I never change guide rate but I sometimes reduce the guiding "gain" to 40% after PHD calibration. That gain is the sliders (one for RA, one for DEC) you get on the main EQMOD window, below its own guiding graph.

Maybe I should not be doing that...

PHD calibration only measures the movement of the mount in order to dtermine the characteristic of your mount - which is only one factor it has to consider. At no point does it run an active guiding session to automatically tune the various controls and parameters that control how/when the guiding algorithm is applied i.e. aggressiveness, minimum motion, hysteresis etc. What is more, being a closed loop system it can be very hard for a user themselves to determine optimum setting for these as changing one may well affect the way the others work.

With early version of PHD it wasn't possible to change these parameters whilst PHD actively guiding and so the gain controls in EQMOD were added to provide a means to adjust the overall strength of guiding on fly.

If the EQMOD controls allow you to reach the guiding performance you need then I can't see any reason to stop using them.

Chris.

[themos]

I think, Chris, that the EQMOD slider set to 50% would turn a PHD request for a 400ms guiding pulse into a 200ms one, is that right?

[chrisshillito]

I think, Chris, that the EQMOD slider set to 50% would turn a PHD request for a 400ms guiding pulse into a 200ms one, is that right?

Yes, that's correct. The only thing that is affected by the gain slider is the pulse duration. Direction and guide rate remain the same.

Chris.

[lukebl]

There's been some really interesting discussion since I started this thread. I've learnt a lot. Many thanks for the input.

It's all prompted me to have a major review of my guiding system since starting the thread. For the past 18 months I've persevered with an unmodded SPC900 webcam to guide, from which my original PHD graph was derived. It has been a heck of a fiddle finding a guide star, but I've always managed it by adjusting the guidescope in its guide rings. It's always concerned me that this was a potential source of flexure but, to be fair, the results have generally been OK. I say generally OK, as it's always been a bit hit-and-miss, with some nights perfect, other nights when the stars always trailed.

However, a couple of days ago I invested in a QHY5 cam, and what a difference that makes. Much bigger field of view, and loads of stars to guide on without having to move the guide scope. So I ditched the guide rings and tightly clamped the guidescope to the main scope with its original rings, so its now rock solid. Did a quick test last night, and managed 10 minutes with perfectly round stars. Looks like the clouds have rolled in for the next week or so, but I'm optimistic about the next session.

Incidentally, I started guiding via the ST4 cable directly to the mount (guiding 'on-camera'), and it guided OK for a couple of minutes but then the RA graph just gradually slipped down off the scale. I then reverted back to guiding via pulseguiding/EQMOD and it was fine. Any idea what was going on?

[George]

Incidentally, I started guiding via the ST4 cable directly to the mount (guiding 'on-camera'), and it guided OK for a couple of minutes but then the RA graph just gradually slipped down off the scale. I then reverted back to guiding via pulseguiding/EQMOD and it was fine. Any idea what was going on?

Don't know what that's about, I remember reading a discussion somewhere on ST4 vs EQmod guiding with a QHY5 and the end result was identical, graphs were the same etc tho it was noted that there was miniscule delays on guiding commands from EQmod.....or was it ST4?

Anyway the consensus was to use EQmod as it does away with the extra cable needed for ST4.

[Ant]

Yes, that's correct. The only thing that is affected by the gain slider is the pulse duration. Direction and guide rate remain the same.

Chris.

What I also do once every thing is connected up and talking to each other (and before calibrating in PHD) is move the two sliders in EQMOD back up to the full 90%.

Then run the calibration in PHD, once complete slide them back down to around 50%.

If you don't (my thinking is) that PHD adjusts during the calibration routine and doubles the pulses sent to the mount because EQMOD only sends 50%

So if you do not reset the sliders in EQMOD back up to 90% before calibration, you end up back at full 100% going to the mount irrespective of the % setting in EQMOD. Because PHD doubles the signal only for EQMOD to half it before sending to the mount.

That's the conclusion that I've come to anyway.

Ant

[Geno]

Incidentally, I started guiding via the ST4 cable directly to the mount (guiding 'on-camera'), and it guided OK for a couple of minutes but then the RA graph just gradually slipped down off the scale. I then reverted back to guiding via pulseguiding/EQMOD and it was fine. Any idea what was going on?

Coincidentally I have much the same problem but the RA seems to slip of the scale straight away. I wasted so much time trying different settings in PHD, different ST4 cable but the fault persists. I was however, guiding all last night through EQMOD with no major problems at all.

All I can think of is there is a problem with the mount, camera or possibly USB cable. May also be a software issue but it was nice to see it guiding after months trying to sort this RA issue out.

[chrisshillito]

What I also do once every thing is connected up and talking to each other (and before calibrating in PHD) is move the two sliders in EQMOD back up to the full 90%.

Then run the calibration in PHD, once complete slide them back down to around 50%.

Yes our guiding documentation has a comment (http://eq-mod.sourceforge.net/docs/EQASCOM_Guiding.pdf page 9, in bold) specifically about this. You should always perform calibration with both gain sliders at 100%. Only adjust the gain whilst actively guiding if you suspect the guiding itself is too strong (or weak).

Chris.