Need ideas for DIY dithering calibration.

[Andy56]

Hi,

I'm building a DIY ditherer that works on the DEC axis of a Sky Watcher Star Adventurer Mini.

It's based on https://www.cloudynights.com/topic/779709-star-adventurer-automatic-dec-dither/ but I have changed almost everything except the 3D printed bracket and the stepper motor.

Currently it uses an Arduino Pro Mini, DRV8825 stepper board, OLED display and two pots to set the dither frequency and amount and runs off 5V only. The aim is for it to run off a USB battery pack. I use a 5v->9v boost converter internally.

According to my calculation the camera will turn about 2.6 arcs secs per step BUT the hysteresis is about 360 steps. 

For a 135mm lens on a Canon 600d I need 2.53 steps per pixel so for 10 pixels I need 25 steps,  but if I want to change direction I need to add in 360 steps so if I have the hysteresis off by a few steps it could mean none or too much dithering.

Currently I'm finding measuring the hysteresis accurately is difficult. I'm using APT in live mode, full zoom with the camera pointing  to a mark on a fence 5m away and increasing the hysteresis steps until I can see movement, very inaccurate least ways I can't tell if it's good or not.

So what I would like is a means of analysing/comparing images to see how many pixels it has moved so I can set in a movement and measure the pixel movement and so see what the hysteresis is. 

I'm thinking something like the way PHD2 calibrates. I have though of using PHD2 but I would have to add a comms interface to the unit and I don't want to spend the time on the software at this time.

So does anyone know of any software tools that would give me a pixel shift indication by comparing adjacent images so I can increase the hysteresis slowly until I can see a change?

I've also looked at SharpCap but can't see an appropriate tool.

If you wish to see 

Cheers

Andy

 

Edited February 10 by Andy56

[michael8554]

PHD2 can measure Dec Backlash, akin to the hysteresis you're trying to adjust.

Failing that take before dither and after dither images.

Layer them on an image editing prog and measure the shift in pixels.

Michael

 

[Andy56]

Hi Michael,

I understand what you are saying.

Given the hardware I have at the moment I was hoping for some program that I could call programmatically to analyse the two images and compare this to a number of steps. I would increase the number of steps until there was a measurable pixel shift and this would give me the backlash number I need.

Since my first post I have decide not to be two precise at this stage and just give it enough shift to average out FPN but no so much to require significant image cropping. The SAM defaults to a minimum of 1' of dither (there's a trick to get this lower)  but I was looking to calibrate it to 1".

So now if I think of dithering in minutes of arc then at 135mm, 1arcmin will be approx 152 steps so if the backlash is 10 or 20 out it won't be a big deal at this stage.

So if I dither up to say 50 steps than I'll lose about 300 pixels off the frame in DEC. I can live with this at this stage.

My next stage is to have an ST4 port so I could use in on my Star Adventurer Pro (not Mini) and get tracking in RA an DEC.

Cheers

Andy