NEQ6 Guiding - Rotational periods for belt mod mount

[mbalkham]

Hi All,

Doing some analysis of last nights guiding (the best I've achieved). Around 0.9" RMS over 5 hours and best period under 0.6"RMS with 20kg of OTA (and 20kg of counterweight). Heavily modded NEQ6 mount (wedge, Rowan Belt, Zero Backlash mod, ceramic bearings, hypertune, new SKF bearings throughout (yes they made a difference and yes I know they only rotate once a day!), many Geoptik mount upgrades for stability). 

I know that the 10.2, 122.2 & 478.7 periods relate to stepper gear tooth, stepper gear and worm drive gear (from EQMOD Appendix http://eq-mod.sourceforge.net/docs/PECPrep.pdf). Can anyone point me to information on the effect of the belt mod on these periods? Just looking to see if there is anymore I can do to improve tracking. RA still looks a little erratic. 

Guidelog, frequency analysis and sample guiding attached.

PHD2_GuideLog_2020-08-17_185448.txt

[mbalkham]

4hrs of guiding 

[spillage]

I have a similar set up minus the zero backlash mod. I usually use the target most in phd. I would say try shifting the ota back or forward a touch and see how the RA reacts (looks it may be east heavy). I normally find it is either balance or slight backlash. Normally get 0.7" ish, but have had it as low as 0.5" on a good night.

Of course I could be totally wrong but that is where I would start.

[bottletopburly]

I found lowering the exposure time to 1 sec from three made a big difference conditions allowing  that is and use the predictive pec algorithm.

[michael8554]

That large 10 second component is a worry often reported on the PHD2 Help Forum.

Their suggestion is to use the PHD2 PPEC as you are doing, but set the period to a fixed 10 seconds.

Also 3 seconds exposures seems slow ?

Long-term I would look at your guidescope in adjustable rings, which can lead to Differential Flexture.

Have you ever needed to adjust them ?

Probably not, so clamp rings are better.

Also I guide at 1300mm with a 200mm guidescope, you could save weight too.

The two wires coming off the mount control panel could do with supporting.

Michael

[blinky]

On 18/08/2020 at 10:40, spillage said:

I have a similar set up minus the zero backlash mod. I usually use the target most in phd. I would say try shifting the ota back or forward a touch and see how the RA reacts (looks it may be east heavy). I normally find it is either balance or slight backlash. Normally get 0.7" ish, but have had it as low as 0.5" on a good night.

Of course I could be totally wrong but that is where I would start.

I found the balance was critical once it was belt modded - did as you said and it made a big difference to guiding

[mbalkham]

Thanks all for the tips. Next clear night I'll give a few a go. 

Looking at OAG which could help me plus save some weight. Failing that might look at new guidescope rings. Prob move the onboard rpi and dew heater for better balance. Going to fiddle with ra meshing again. Try shorter guide lengths. Sort a couple of cables. Check balance with ammeter. Plus a couple of other things!

[blinky]

I would also say, are you getting nice round stars?  If so, then LEAVE IT ALONE!!!

[mbalkham]

6 hours ago, blinky said:

I would also say, are you getting nice round stars?  If so, then LEAVE IT ALONE!!!

Good frames are round bad ones are very stretched in RA (maybe twice as long in RA compared to Dec).

[michael8554]

On 20/08/2020 at 18:22, blinky said:

I would also say, are you getting nice round stars?  If so, then LEAVE IT ALONE!!!

Roughly speaking there are three shapes of stars due to guiding:

Small round stars, fat round stars, and elongated stars.

So if you want small round stars you'll have to work on your guiding.

RA RMS is 50% to 70% higher than Dec, which means elongated stars.

I was in the elongated stars situation.

Best RA I could consistently get with OAG was below 1arcsec at 1280mm focal length, , but Dec was even lower, so I set a small drift in PA and set Dec Min Mo to up the Dec RMS to match RA.

Michael

[mbalkham]

On 22/08/2020 at 06:22, michael8554 said:

Roughly speaking there are three shapes of stars due to guiding:

Small round stars, fat round stars, and elongated stars.

So if you want small round stars you'll have to work on your guiding.

RA RMS is 50% to 70% higher than Dec, which means elongated stars.

I was in the elongated stars situation.

Best RA I could consistently get with OAG was below 1arcsec at 1280mm focal length, , but Dec was even lower, so I set a small drift in PA and set Dec Min Mo to up the Dec RMS to match RA.

Michael

There's a sneaky trick. Very clever, I might play with this at some point if I can't get RA and Dec closer to one another my some mechanical means.

Just got my OAG and redistributed some weight onboard for better balance either side of the flip. Moved the computer/power etc. on top rather than on the side so much closer to 50/50 weight distribution now.

[masjstovel]

The significant response in frequency at 10.2s, indicates that the RA belt tension needs tightening. I have the NEQ6 beltmodded and SKF bearings myself, and had a oscillating issue in RA that it seems (?) you're having too. Your frequency response graph looks pretty simular to what mine did. 


I dont know if you did the mod yourself, but i handtightened mine first, but its very tight space between the housing and motor so it made a false impression of how tight it was. Used a flat screwdriver in between and sort of pushed it tighter with a light force. Made a significant improvement in RA.  Went from about 1.00" to ~0.75".

Next problem was that my stars were elongated even if guiding was satisfying. Now i see you have the zero-backlash mod, so i dont know - but:
That turned out to be a "too well balanced" setup. I had heard that perfect balancing was important with the belt-mod, but in my case that was not true.

Because of backlash and perfect balancing - the setup was sort of "floating" around too fast for PHD2 noticing it. 
Made RA east heavy, and DEC backend-heavy, and that improved the guiding further.  I also got a new guidescope for better pixelscale (from 5"/pixel to 3.2"/pixel).
Now my best guiding over time have been 0.45", but around 0.50"-0.60" normally. (Issues have recently been fixed so "normally" = last 20 hours or so of guiding). 

Now, Im much more a newbie than a pro, but these were the issues i had and how it was fixed at least.
 

BTW, thats an extremely sexy setup you have there:) 

Edited November 8, 2020 by masjstovel

[mbalkham]

Thanks for the detailed response. I've got guiding down to 0.5-0.6"RMS typically which I'm happy with (as my main image scale is 0.74"). I've switched to using multistar guiding in kstars and OAG. I'll try phd2 again at some point just for the log and analysis. I have found OAG helps in good conditions but I seem to be more wind effected now.