The EQ3 DSO Challenge

[Stub Mandrel]

Thanks Nige,

I think I have two problems, first is that t 1x sidereal it takes about ten seconds to clear the backlash no matter how well I set up, and second my stepper isn't powerful enough to shift the scope when the mount is significantly out of balance and most of the backlash is taken out. I'm going to try swapping in a bigger stepper (NEMA 17) .

[Nigel G]

Would you belive it, PHD has droped out at least 5 times tonight. I have cursed myself   dec everytime,

D'oh

[Stub Mandrel]

9 hours ago, Nigel G said:

Would you belive it, PHD has droped out at least 5 times tonight. I have cursed myself   dec everytime,

Sorry...

[Nigel G]

Its the position of the DSO  once it had passed the meridian everything worked fine  

Curse lifted

Nige.

[mikey2000]

Further to my recent desire to experiment, here is a little report.

1. About my field flattener - mixed results.  There are no adjustments available by the looks of things.  I've tried making my own shim out of cardboard but initial impressions are that I've got perfect stars top left and bottom right but the other two corners are even more stretched.  I must have made a wonky shim.

 

2.  Playing with PEC.  The eqmod pec works well.   The super zig-zag curve wasn't great as the 19s sawtooth isn't in phase with the main worm period so it soon goes awry.   I tried a eqmod pec curve for just the main worm PE and it does a good job (allowing PHD2 to deal with the 19s sawtooth). Not bad.  I'll stick with this.   Setting PHD2 PPEC algorithm to 19s (on top of the eqmod worm correction) didn't work well. It seems the PHD PPEC algorithm is designed for longer PE cycles and kept "growing" it's measured cycle length towards the main worm period.   No matter really - the eqmod Autopec brought the unguided PE down from 30arc sec to within 4 arcsecond, so PHD2 is less "busy" in RA.  I was getting RMS in RA about 0.6" for a while (with the scope aimed at about 20Dec so RA is quite sensitive that near the equator)

 

 

[Stub Mandrel]

I've found that backing the RS aggression right off gives me guiding better than 1" which is enough to keep me happy, it only goes significantly awry when the DEC does a big jump.

I think my DEC problems are largely from two things: 'stick slip' and inability to unbalance enough to remove backlash because of a weak stepper, I've replaced the stepper and ordered a pair of needle roller bearings.

[mikey2000]

But now I'm getting DEC issues again, two nights running .  I'd dismantled the scope from the mount to move it and after reassembly, I'm getting occasional DEC excursions (big ones too!)

 

i think maybe I unbalanced it a bit too much and I'm seeing slippage or stiction.  (And was aiming almost straight up...)

 

ive been trying the "bad polar align" to force unidirectional Dec guiding but still trying to get my head around which way to de-align and unbalance.  I'm reminded of the time I was learning the basics of drift aligning :-)

 

ill try try again a few times then will revert to using sharp cap to getting the best possible PA as that was giving good results anyway...

[Stub Mandrel]

I suspect that the EQ3's main weakness is the DEC arrangement.

The wormwheel is quite small, the adjustment arrangements are crude, and there is only a fibre thrust washer at one end!

It's clear that the RA isn't too bad as plenty of people are achieving 1-2 minute unguided subs with good polar alignment.

The challenge is that for guiding you need to adjust DEC as well.

10 minutes ago, mikey2000 said:

ive been trying the "bad polar align" to force unidirectional Dec guiding but still trying to get my head around which way to de-align and unbalance. 

You and me both!

[Nigel G]

I noticed last night that the mount was tracking good by itself and PHD was making things worse, so I changed the guide cam exposure from 0.2s to 1s ( less input from PHD ) which seemed to work well.

At 0.2 seconds PHD was going mental trying to guide, after I changed the exp to 1s it smoothed out a lot.

So after all the drop outs and fiddeling around I got 22x240s @ ISO800, dithered after each sub with flats and bias.

Modified 1200d, Equinox 80ED & guided EQ3.

NGC 7023 The Iris Nebula. good position atm, darkest part of my sky but seeing was quite poor with haze and a rising moon.

Looking forward to some good seeing for this to pick out the dark dust clouds.

Nige.

[wimvb]

5 minutes ago, Nigel G said:

I noticed last night that the mount was tracking good by itself and PHD was making things worse, so I changed the guide cam exposure from 0.2s to 1s ( less input from PHD ) which seemed to work well.

At 0.2 seconds PHD was going mental trying to guide, after I changed the exp to 1s it smoothed out a lot.

Nige.

 

I don't know if PHD2 has this feature, but at least in lin-guider, you can set it to average a number of exposures when sending a guiding command. If you set frames to, say, 3 lin-guider will average over three exposures before sending a command. This 'quiets' the guiding behaviour, as you're no longer chasing the seeing. Maybe PHD2 has a similar feature.

Just a thought.

[Nigel G]

1 minute ago, wimvb said:

I don't know if PHD2 has this feature, but at least in lin-guider, you can set it to average a number of exposures when sending a guiding command. If you set frames to, say, 3 lin-guider will average over three exposures before sending a command. This 'quiets' the guiding behaviour, as you're no longer chasing the seeing. Maybe PHD2 has a similar feature.

Just a thought.

I'm not sure, must check that out.

Nige.

[mikey2000]

With PHD2, I find 1-2s the best for my camera/guidescope.  There is a "guide assistant" which helps pick the best settings to avoid chasing the seeing (it's called "min move". I think the idea is that it measures the average wobble caused by seeing them set it so it only sends guide commands for movements larger than that average.

[Stub Mandrel]

The PHD tutorials say less than 1 second will 'chase the seeing'. A 1s exposure will have the same averaging effect as a few shorter ones. I imagine aaveraging is only needed if otehrwise the stars are getting over exposed.

What sort of FWHM are you choosing of guide stars (I've gone for stars between about 4.5 and 7.5)

[mikey2000]

I don't recall the FWHM exactly but those numbers seem sort of familar.  Looking at the PHD guide logs it records the star HFD which seems to be somewhere in the rande of 3 to 4 pixels (I use AutoSelect for the guide star) with a typical SNR of 20-something (or sometimes as low as 'teens')     Maybe the answer here varies depending on guidescope and guide camera.  My guide scope is only 130mm FL (30mm aperture) but my guidecam has very small pixels...

[Stub Mandrel]

1 hour ago, mikey2000 said:

My guide scope is only 130mm FL (30mm aperture) but my guidecam has very small pixels...

Wow, I thought mine was short at 183mm!

[mikey2000]

I read a while back that it's all about image pixel dimensions vs guide pixel dimensions (in arc seconds)   I further read that 1:4 is about as far as you can push it these days.  I'm attempting 1:3.9 :-)  If I downsize my main subs by 50%, they pixels double in size (each becomes a 2x2 superpixel) so 1:2.  no problemo!

Except the pesky sticky Dec axis...

[Stub Mandrel]

Just calculated my ratio and it's about 2.56:1

My first experiment was a success, it's my tries at real guided imaging which are turning out as failures!

I'm trying to pluck up the courage to stack my subs, I know the stars will look like nice biscuits (not even rugby balls!).

If my DEC bearing mod works for 'unsticking' it might solve your problems. Hopefully they will arrive tomorrow as Saturday Evening looks promising at the moment.

The spec is:

------------------------------------------------------
1 x AXK2035 Needle Roller Cage 20x35x2mm (AXK2035-MB) = £2.12
1 x AXK1024 Needle Roller Cage 10x24x2mm (AXK1024-MB) = £1.88
1 x Nadella CP1022 Thin Needle Thrust Washer 10x21.5x0.8mm (CP1022-NADELLA) = £1.22
1 x Nadella CP2035 Thin Needle Thrust Washer 20x34.5x0.8mm (CP2035-NADELLA) = £1.24
------------------------------------------------------
Items Sub Total: £6.46
(Royal Mail 1st Class Post (Ships Friday)): £1.50
VAT: £1.59
Order Total: £9.55

All from Simply Bearings. the thrust washers are packs of 2. Just spotted I hope those 1022mm ones aren't too small I should have got 1024. rats, not impossible to make some though.

My worry is that the small end may need me to fit a small spacer to centralise the bearing as the thread is heavily undercut.

[mikey2000]

Wow - that's some serious DIY.  I have managed to fiddle with the worm gears/shafts but I'm not sure I'm brave enough to start pulling the internals out....

[mikey2000]

One from last night.  I dumped 7 out of 41 subs due to the dec madness but at 2 mins each, that's not a bad keeper rate.   When I stepped outside to pack away, I was disappointed to see our old friend Mr Moon shining down on me so I wasn't expecting much.  My first thought after stacking was, 'nothing there, oh well.'  But when I stretched this one to pieces in startools, there it was!

 

I've seen hubble pics of this one - the dark halo around the cocoon is there in those pics as well.

IC 5146 Cocoon Nebula

[Art Gecko]

Nice, I like it Mikey!!

My best effort from last night here... The crescent nebula (finally had a go!)

41 x 60s subs, darks and flats as usual and stretched beyond imagination in Ps. I cropped it down a fair bit as it was a real struggle to get anything out of it in full frame.

[Stub Mandrel]

17 minutes ago, mikey2000 said:

Wow - that's some serious DIY.  I have managed to fiddle with the worm gears/shafts but I'm not sure I'm brave enough to start pulling the internals out....

No its easy, you just unscrew the four screws to take off the worm gear, undo the nut inside the mount (in the polarscope top hole) and pull out the DEC shaft. Replace the red fibre washer at the top with the big thrust bearing and the washer at the bottom with the small bearing. Clean off old sticky grease and replace with new grease.

The worm gear housing will need a simple spacer the same thickness as the top bearing MINUS the thickness of the red washer.

[Stub Mandrel]

5 minutes ago, Art Gecko said:

The crescent nebula

I'm learning to hate the Crescent Nebula, it's cursed!

[mikey2000]

2 minutes ago, Stub Mandrel said:

No its easy

 

 

:-)  When you make the mod, would you be kind enough to take a few step-by step pics?   I'm sure it would make a very popular post (alongside the EQ3 stripdown posts....)

 

[mikey2000]

Art Gecko - I think you caught some good data for your crescent but I'm not convinced Photoshop is the best software for the initial stretch.  I use PS for final tweaks but I highly recommend a free trial of astrotools.  Feed it a autosave.fits file from DSS and have a play.  Or if you like, dropbox a .fits file and I can have a try for you (for comparison purposes)

[mikey2000]

Here's a dec trace from last night.  I de-tuned the PA to try uni-directional guiding.

Mostly it's OK with PHD sending 'guide south' signals.   Then it goes genuinely south before recovering.

Here's my theory. As the scope is nose heavy, both the guide signals and gravity are trying to pull it down.  It's all ok then suddenly hits a slack bit in the worm and gravity moves the scope suddenly down into the slack.  The little overshoot in the recovery must be backlash as guiding changes direction.

 

So, I need to make sure the scope is nose heavy *and* that the PA offset is such that PHD sends Guide North signals, thus it only ever needs to lift the scope and won't ever flop into these dead zones.

 

Does that sound reasonable?