Guiding and mount control??

[Merlin66]

OK, I've got to ask.....
Our "usual" guiding software (PHD etc.) can provide guide correction signals (+/-RA and +/- Dec) to within 0.1 arc sec (depending on the guide set-up) but how accurate are the actual MOUNT corrections??
Even if you pulse the corrections do you ACTUALLY get this level of drive control from your mount???
I'm thinking of the generic HEQ5/NEQ6 SW mounts.
Open to discussion/ comment.

[Merlin66]

Hmmm - no comments?
The NEQ/HEQ mounts have 1.8 degree stepper motors and a gear ratio of 1:705 this gives 9.19 arc sec/ step.
The guide rates in Synscan go down to x0.125 Sidereal rate (=15 arc sec/ sec / 8 = 1.8 arc sec/ sec
By my reckoning too achieve a guide correction of 0.1 arc sec, the motor would have to be micro-stepped at 9.19/0.1 = 1/92 steps to give the required correction.
Based on the lowest Synscan guide rate it would require a pulse of 1/18 sec....

[Merlin66]

Andy from the PHD2 team has replied:

PHD2 will attempt to send small pulses in the 1-10 ms range to the mount. Some mounts are able to honor these small guide pulse commands; but, if the guide pulse duration is too short for the mount, the mount ignores it. The tracking error accumulates, and PHD sends a larger pulse. In this way PHD and the mount can reach a kind of equilibrium to achieve tracking performance close to the error that phd2 can measure on the guider chip and the smallest pulses that the mount can react to.

[Bassnut]

That makes sense, the actual mount movement doesnt matter really, given setting agressiveness is a suck it and see excercise, and as above, if a correction is too much/little, the next one compensates.

[Merlin66]

I understand....but what if your mount doesn't/ won't react to the short pulse commands......

[lazjen]

Then I think you're in for a world of pain. I can't see how you'd be able to get decent results with such a setup, since the corrections would either be large or ignored?

[SteveInNZ]

It's a closed loop system like you, the gas pedal and the speedo when you're driving. You give it a little bit and nothing noticeable happens so you give it a bit more until something does happen. That will probably take you too far so you'll back off a little and then a little more. Eventually you'll settle at a point where the on again/off again matches the minimum amount of controllable movement the car/mount (system) can work over.
Control loop tuning is a form of wizadry that plays off optimum response against stability. You really want to be at the stable end.

Steve.

[Geoff45]

The question is why would you want to make a correction of 0.1" when the seeing is probably never going to be better than 1.5"? Your mount will be trying to make unnecessary corrections that will never be visible. The fewer corrections you have to make, the better will be your guiding. See here http://www.ccdware.com/resources/autoguidercalcv4.cfm for a good discussion.
Geoff

[Merlin66]

Geof,
I agree.
The issue was if the guider has a significantly different plate scale from the imaging system, say 10:1, a 0.1 arc sec guide correction may result in a 1 arc sec "correction" in the imaging...
I was just interested to see how the mount could interact with the guide signals to achieve such an outcome.