Guiding spikes after dithering. Can't find why.

[aleixandrus]

Hi! I'm having some issues with my guiding. From time to time, I have very large spikes after dithering and the mount takes a while to recover. My pixel scale is quite permissive but I'm planning go for a longer scope and I want to address this before. Here is the thing:

• I have an EQ6R, my OTA is a Samyang 135mm with an ASI183MM Pro and I guide with an 32mm/f4 guidescope with an ASI120M.
• Guiding goes well, stable at 0.6-1.2" depending on sky conditions. I dither every frame, usually 300s (narrowband monocrome sensor). But, suddenly and with no apparent reason, I get a very large spike in DEC just after dithering. This happens from time to time, usually few times in the night. Other nights, this get worse and they are more frequent. The screenshots below shows a worst case scenario.

This is my configuration for guiding in NINA and PHD2:

• Dithering: 5px with a scale of 6.45"/px (guiding), ie., 32"/px. My scope and main camera have an scale of 3.67"/px, so I'm dithering 8-9px. I saw many people moving 15-30px in the main camera but I think mine is just fine to fix de fixed pattern noise of the ASI183, which is  quite low.
• Guiding: I run the PHD2 Guiding Assistant and I accept all the recommendations. Each time I try to 'fix' it, results get worse. So I ended up trusting PHD2 and its recommendations.

My OTA is just a bit unbalanced in DEC as the Samyang is so small that the camera, filterweel, etc. almost touch the mount head, so it is a bit heavy towards ground (if pointing north). Not a lot, but noticeable.

What could be the reason of this spikes? Unbalanced OTA? Wrong configuration? Too much dithering? Backlash? I accept any ideas as I run out of them.

 

[ollypenrice]

This is a CMOS sensor so is it worth dithering? Personally, I wouldn't bother.

Olly

[Clarkey]

I am not exactly sure why your set up is behaving as it is, but the backlash graph is very odd. Initially the backlash is showing very little, then it stops responding. Normally the 'measured' line runs parallel with the 'ideal' once the backlash has been removed. This suggests the mount is no longer responding to the outputs from PHD2 - but I don't know why.

Personally, on a relatively under-sampled image scale, I would increase your dithering to a larger figure. It is possible PHD2 is struggling to detect the small movements? Alternately, you could try without dithering. (I tried with my colour camera and it had too many hot pixels so I stuck with it. Some you might not need to).

[Stefan73]

I'd try to eliminated all the obvious things which might stop the mount responding to PHD2.  Bad USB cables, cable snags, worm gears binding in certain positions maybe.

I had similar graphs with my Samyang / EQ6-R sometimes and I think a cable was getting snagged slightly as it went away with more careful cable management.

 

[michael8554]

The second guide graph shows Dec Stiction occurs at certain positions, not all as the first graph is fine. 

PHD2 struggles to correct the Dec Dither, taking  many seconds to correct. 

Then Dec overshots and PHD2 again takes many seconds to pull Dec back. 

If your sensor does not suffer from Fixed Pattern Noise you do not need to Dither. 

Michael 

[alacant]

19 hours ago, aleixandrus said:

any ideas

Hi

Loosen the circular nut which pre-loads the cone bearing at the base of DEC.

Make certain the bearing is clean and well lubricated. Use the white grease available at cycle repair outlets.

HTH

**EDIT: just seen the graphs on a bigger screen. It looks like RA would benefit from the same treatment.
In fact most SW mounts need pulling apart, cleaning, lubricating and adjusting, with new mounts being no exception. For guiding to be effective, the mechanics have to be better than perfect. Fortunately with a short lens, it doesn't matter so much.

Edited January 5 by alacant

[ollypenrice]

I'll say it again: dither is for DSLRs and, maybe, CCDs. 

CMOS?  Why?

Olly

[aleixandrus]

Thank you for your comments, really appreciate it. Some additional info:

• A dither to perform drizzle x2, as I undersampled.
• I don't see fixed pattern noise in my ASI183 when processing non drizzled master lights. I see some hot pixels, but darks seem to keep them off.
• My EQ6R is 2nd hand, about 4 years old or so. No maintenance was done as far I know. Although the mount has not many (real) work hours, maybe is time.
• USB cables are supposed to be "good" and they are less 1 year old. Yes, I know this means nothing but everything works except in that rare moments after dithering. No equipment disconnections at all, never ever.
• I take some time to route them carefully to avoid snags. I have a PegasusAstro Powerbox Advanced attached to the OTA and all cabling is tied and static except the one with power from the ground to the powerbox and both power and USB from the powerbox to the mount. To me, no cable failure is happening.
• The OTA is a bit unbalanced, down towards camera, as my filter wheel crashes with the mount head if I put the scope higher. The Samyang is so small it is not easy to handle or balance. Maybe I should add risers or additional weight to the front (not easy).
• Those spikes *only* occurs after dithering, randomly, without apparent reason.

As I said, given my pixel scale, I'm not worried by the current performance but for the future, when I go for a longer scope. If I can learn and fix things now, I prefer doing it before I'm having more problems. Based on your comments, I'll try:

1. Check cables again.
2. Increase dithering.
3. Re-run PHD2 Guiding Assistant to double check backlash values.
4. Study how to better balance the OTA (not that easy).
5. Check EQ6R mechanically (this will take longer, have to study how this goes, never done it to any mount).

[alacant]

11 hours ago, aleixandrus said:

If I can learn and fix things now, I prefer doing it before I'm having more problems

Hi

In that case, perhaps the best fix would be to be able to take the mount completely for granted, otherwise you're constantly fighting the unknown: The 6r is easy to dissassemble and upgrade. Details here.

Only then can you realistically expect the niceties of software and cables to have any effect.

Cheers, feliz año y ¡ánimo! 

Edited January 5 by alacant

[aleixandrus]

I'm not ready yet to perform that surgery to my mount 😅 I need get some experience before and try adjusting the backlash using the external screws or the software. In addition, I don't have Plan B if something goes wrong, there are not many opportunities here to repair the mount if I mess things up. You are probably right... but not yet 😁

[alacant]

53 minutes ago, aleixandrus said:

I'm not ready yet to perform that surgery

Ah, ok.

Luis at str-astrum will do it for you:

https://www.str-astrum.com/

You need send only the mount head and the handbox.

Cheers

[david_taurus83]

On 04/01/2024 at 19:12, ollypenrice said:

I'll say it again: dither is for DSLRs and, maybe, CCDs. 

CMOS?  Why?

Olly

And CMOS. I don't bother with the task of taking multiple sets of darks for different exposure lengths. Just a set of Bias frames for a master bias and dithering takes care of hot pixels and walking noise. That is of course with the newer CMOS with no amp glow.

[aleixandrus]

1 hour ago, alacant said:

Luis at str-astrum will do it for you:

Oh, thank you so much, didn't know about them. I'll consider get in touch once I could install my mount in my backyard observatory, which I should start building in the very near future, so this way the mount could stay untouched and adjustments should be easier.

[alacant]

On 08/01/2024 at 17:00, david_taurus83 said:

Just a set of Bias frames for a master bias

Modern sensor? Even cleaner: lose the in-camera bias and simply subtract the offset.
Cheers

**EDIT; apologies to OP. Whilst you may still use the synthetic offset, the 183 must be calibrated with light-matched dark frames.

Edited January 11 by alacant