Bad PA or bad PE?

[Demonperformer]

This is not exactly scientific, but I found the visual representation to be quite instructive.

I took a series of (unguided) images and then registered the entire series in DSS. I then took and plotted the dx (blue line), dy (pink line) and rotation (yellow line) corrections calculate by DSS for each image against the time in minutes from the time of the first image (x-axis). I also did a regression analysis on the three data sets to put values on the 'trend' on each line.

Firstly, the rotation is very small (I think this would be zero with perfectly aligned PA). The regression shows that it would rotate roughly -1° in 60 hours (a slope of -0.00029 per minute, although as the input of that data was only to 0.1° accuracy, that result is not reliable).

The dx values seem to be a bit all-over-the-place. But the regression line reveals a trend of about -1 unit (pixel?) in 38 minutes (a slope of -0.02616 per minute).

The dy curve, however, is more obvious. There is a clear upward trend (presumably primarily due to the error in PA), which regresses to about 1 unit every 15 minutes (a slope of +0.06682 per minute). However, there are also 5 clear cycles of roughly 8 minutes, produced by the PE of the mount. Each of these travels a distance of between 2-3 units up and down in each 8-minute cycle; which equates to a total travel of about 5 units in 8 minutes, or about 10x the effect of the regression line.

All of which suggests to me that, having aligned the mount with Polemaster, I am getting significantly greater errors from bad PE than from bad PA.

But, as I said at the start, this is not exactly hard scientific data ...

Thanks.

[dph1nm]

Yes , the units in DSS are pixels. So you need to know the pixel scale of your setup (and declination the images were taken at) to work out what the PE is in arcseconds. One assumes from the image that RA is in the y direction.

NIgelM

[Demonperformer]

Thanks, Nigel.

The image scale is 11.76"/px and the declination of the image-centre was 8° 55' 05.5". So, as the peak-to-trough distance is over 2½ units, that would put it in the region of 30" of travel in each direction in 8 minutes, compared with a PA error of somewhat under 1" per minute?

[alexbb]

There's definitely an error and it seems periodic. So...

Also, I remember that for many of the Synta mounts, a full gear revolution takes 8 minutes. Found it here as well:
http://eq-mod.sourceforge.net/eqspeed.htm

 

Before posting I saw that you have another thread regarding PHD. At least for short-ish focal lengths, guiding is enough.

[Demonperformer]

18 minutes ago, moise212 said:

 guiding is enough.

Thanks, Alex.

With all the problems I have encountered with my (admittedly limited) experiments on guiding, anything that reduces the need for it is welcome. Added to which, my system is not well suited to adding guiding to photographs that are just taken with camera and lens.

If I am getting travel of 30" in both directions due to PE, that is a total of 60" travel in 8 minutes, or about 7½” per minute. If I am getting 1" of travel due to PA per minute, that gives me a total of 8½” per minute, of which 88% is due to PE, which in itself would suggest that, get rid of PE (which I can apply equally to the mount whether I have a guided scope, or an unguided camera) will enable me increased exposure times with more sharpness.

Thanks.

[alexbb]

2 minutes ago, Demonperformer said:

1. With all the problems I have encountered with my (admittedly limited) experiments on guiding, anything that reduces the need for it is welcome.

2. Added to which, my system is not well suited to adding guiding to photographs that are just taken with camera and lens.

3. ... get rid of PE (which I can apply equally to the mount whether I have a guided scope, or an unguided camera) ...

 

1. Don't worry, you'll get there. After the first successful try, it will be much easier.

2. After you'll succeed with guiding, I don't think you'll want to bother anymore with PEC unless you really really don't have the option to take your laptop and guiding camera with you. You could attach a short FL lens on the guiding camera on your DSLR's hot shoe. A 50mm lens for guiding should be enough for a 200mm lens for taking photos.

2 + 3. You still need to do the PEC training for each session if you rotate the mount after powering it off.

[Demonperformer]

3. EQMOD automatically records the positions of the stepper motors. As long as the mount is parked at the end of each session, and I do only move the axes with the clutches disengaged, it does NOT need to be redone every session.

2.Why would I want to attach another lens to the shoe of my camera (involving greater compexity and cost), if I can get rid of 88% of the problem simply by applying PEC?

1. Maybe I will, but by elimintating 88% of the problem, PEC should make a dramatic improvement while I am "getting there".

 

[dph1nm]

22 hours ago, Demonperformer said:

that would put it in the region of 30" of travel

Yep, seems about par for the course (assuming this is not a 10k high end mount!). My HEQ5 has about the same.

NigelM

[Demonperformer]

11 minutes ago, dph1nm said:

assuming this is not a 10k high end mount!

I should be so lucky ... just a lowly NEQ6

[alexbb]

20 hours ago, Demonperformer said:

3. EQMOD automatically records the positions of the stepper motors. As long as the mount is parked at the end of each session, and I do only move the axes with the clutches disengaged, it does NOT need to be redone every session.

Indeed, you're right. I've been slow yesterday and today even slower. I'll shove myself out.

[kens]

60" p-p is on the high side. My EQ6Pro has about 18" p-p. 30" p-p is more typical so you may be able to reduce it with some tuning.

You'll be unlikely to be able to eliminate PE using PEC, maybe halve it. A large part of my PE is from the transfer gear which is not a harmonic of the worm frequency and therefore cannot be removed. Even the fundamental cannot be entirely eliminated with PEC as so many factors come into play like imperfections on the worm wheel.

If you don't want to auto guide you could keep your subs to around 120s and try to take them at the peaks of the PE where the movement is at a minimum

[dph1nm]

47 minutes ago, kens said:

60" p-p is on the high side.

Isn't it 30" p-p? Looks like 2.5 pixels to me and its almost on the equator.

NigelM

[Demonperformer]

It's about 30" up and then 30" down in the 8-minute cycle.

[kens]

11 hours ago, dph1nm said:

Isn't it 30" p-p? Looks like 2.5 pixels to me and its almost on the equator.

NigelM

I think you are right. The up and down bit confused me. 30" p-p is absolutely typical of these mounts

[Demonperformer]

Having finally managed to get PERecorder to talk to the webcam (or should that be 'get the webcam to talk to PERecorder'?), despite being fairly knackered last night, I spent an hour recording a PE trace. Looks to be pretty much on a par with the dy curve in post 1 - possibly the PA was a bit more off than for post 1, giving a greater upward slant of the cycles. Taken at 2m focal length and a resolution of 0.66"/px, which I think should be sufficiently accurate!

Will get on an process this now and then repeat the exercise carried out at the start of this thread to see the difference (assuming there is one!).

Thanks.

[Demonperformer]

Quick calc: The first peak came in at about 12" above the line, the fifth peak comes in at just over 44", that's a drift of ~32" over 4* 8m cycles, which still works out at about 1" per min, so it does not appear that PA was any worse last night than for the original post.

Consistant PA from Polemaster ... that's encouraging.

Just as a matter of interest, how does this stack-up against other forms of PA that people use? (I know it means I am still going to have to get to grips with guiding - BOO!! HISS!!)