Drift alignment - what am I really seeing question?

[g__day]

I'm still trying to master this basic, but essential skill and I'm hoping someone can explain things before I go crazier.

To start with my permanent pier set-up seems to be within say a few arc minutes of the SCP according to MaxPoint with a 50 star model - so here's my puzzle:

Good news - after finally correctly pointing due North at the meridan - I adjust the equatorial mount so I see no drift of the star moving off the central dot of a Canon view finder attached to my C9.25 in about an hour - excellent!

Slewing due East about 25 - 30 degrees above the horizon I seem to see slight DEC drift (say 1 arc minute in 15 minutes) but I am also seeing maybe 1-3 arc minutes in RA drift (the always scope being ahead of the star).

How can this possibly happen - if there is no RA or DEC drift in an hour when the scope is vertical how can I see such significant RA drift but minimal DEC drift when adjusting my latitude elevation precisely?

I know advice says ignore all RA drift when adjusting for DEC - but I am finding it very hard to ignore that level of RA versus DEC drift!

Sanity restoring advice would be very welocme right now!

PS

As an example focused on Alphard at 11pm tonight when it was say elevated at 40 degrees above the horizon. After 45 minutes there was I estimate 5 arc second DEC drift North and the scope was ahead of the star in RA (West) by 45 arc seconds - how does that happen?

[Zuts]

Hi,

I am new to this, but are you worrying too much. maybe you are seeing periodic error and with guiding all would vanish :).

Also, my guess is that if its not polar aligned then its just not polar aligned. After you iterate a few times then except for periodic error it should stay exactly on the star you are on.

I tried a test last night. Usually i iterate back and forth a few times and it takes me around an hour to set up each time. I dont have a fixed place to image. The other night i just did a very rough RA align and assumed the DEC was OK from the previous session. I will try this again as in this case i cut down my setup time to only 20 minutes rather than over an hour. Now if i could only focus properly :)

With guiding using PHD i could see no drift using 8 minute subs with this rough align, this is about as long as i wish to go with a DSLR. Admittedly my focal length is only 600 mm, but hey it works for my setup :)

By rough i mean only minimal vertical movement after about 3 minutes, and also by rough i mean i didnt do an iteration with a barlowed reticle.

Paul

[g__day]

PEC is on - should be under 3.5 arc seconds peak to peak if the Vixen gear operates to spec.

PEC also should not see a net and consistent RA drift and it shouldn't be of that magnitude - that's way too large and it shouldn't vary in low horizon versus elvated shots without a clear mathematical reason.

I do 10 minute shots at 2300 mm its alot more demanding than 600 mm shots.

With PHD I am seeing issues recently - Craig Stark (who authored PHD) has given me some suggestions - Rod Wodaski says check differential flex, the SkySensor2000 user group suggests isolating the mains power source - even if its run through a UPS into a regulated power supply.

The stats I have achieved suggest that East West my misalignment error is now probably down to low arc seconds given there is practically no drift of any kind in 3600 seconds at 2.3 metre focal length. But for an elevation adjustment run to see a relatively large RA drift when there is practically no DEC drift in a 45 minute run frankly really puzzles me and deserves analysis and elimination.

As you say - you devote an hour to this each night - I'm on a fixed, permanent location pier - I've spent 100+ hours trying to really nail this!

[Merlin66]

Frustrating!!??
Looking at the symptoms; if there's no appreciable Dec drift then I would suggest it's either a balancing/ flexing issue with the scope, or maybe atmospheric displacement ( abberation, NB at the horizon it's almost 30 min arc).
Is the movement always in the same direction? ie RA appears to run faster/ slower?

Have you tried checked this RA movement for an object on the celestial equator at varying altitude both on the East side and the west side of the meridian??? Does the "error" definately vary with altitude?

Maybe with some more analysis we can crack it?

I had a similar but different problem; only showed at high hour angles, where my spectroscope would appear to loose the star???? Finally traced it down ( or the major effect) to movement in the adaptor tubes between the scope and the spectro ( infact one of the issues was the use of T2 threaded spacers... I thought they would be more robust and solid rather than 2" sleeves with grub screws, BUT they unscrewed themselves whenever the "load" was offset CCW!!)

[g__day]

Balance seems fine and in fact RA "appears" fast and even when you move counter-weights to off set this - it still seems fast. So good part of the challenge is its consistent.

I am sure it is mainly polar mis-alignment I am experiencing - my 'internal' issue is I am surprised at how it is manifesting itself.

I think what I must do is babysit it and see how the drift is manifesting itself over 10 - 20 minutes at the eye piece. Normally I set a stop watch - go and watch and episode or two of the West Wing - then when this is over - say 47 minutes - see how much drift and in what components I have experienced - adjust and re-centre - then re-set the timer and go again!

I try and track very close to the meridan (to amplify any apparent error) and would expect the error to vary (decreasing) with altitude. So far I have only tried this on the East - not yet the West.

More experimentation to come!

Matt

[edwardsdj]

I'm certainly no expert in the precision you are trying to achieve but atmospheric refraction could certainly be a large part of the RA shift you are seeing.

My understanding of drift alignment is that if you could point at a star on the Eastern or Western horizon (altitude 0 degrees) and there was no atmosphere then any declication drift can be compensated with a change in mount altitude. Because in practice there is atmosphere and you have to look at a star many degrees above the horizon, you can never purely isolate the altitude error.

Good luck with it!

Have fun,
Doug

[sheeny]

Yes... try the west. It will be interesting to see if the error remains fast or if in fact it turns slow... If it is a balance related problem it should go slow.

Al.

[g__day]

Personally I'm also wondering if I expected mis-alignment drift to manifest in the same way for alt versus az.

Remember I am looking through a canon 400D DSLR viewfinder - which has 9 focus points. Each focus point (I use the central one) has a box around it. From the central point to the box is probably 1 arc minute in angular distance.

My expectations - possibly wrong!

The camera is attached to the SCT so RA drift is always Up / Down and conversely DEC drift is Left / Right.

I expected Polar alignment drift to be predominately in the DEC left / right axis - and for there to be minimal RA drift.

If the slow drift is equally in RA and DEC - say one arc minute in 40 minutes magnitude - so 45 degrees down and right - this may well be normal by geometry (shperical trigonometry). What I never expect to see if drift say 4 parts RA and only 1 part DEC - I don't see how that can correctly happen only in Altitude adjustments if everything else is correctly set-up?

Are my expectations valid?

[Merlin66]

Instead of using the camera viewfinder do you have access to a guide eyepiece with a cross wire reticle? I think using the camera focus squares is not accurate enough!
On my eyepiece, an illuminated 12mm with fine cross wire, mounted on a 12" LX200 it is very easy to see dec movement in minutes, this then speeds up the whole alignment operation.
When looking at the East or West, if there has been any RA movement on my set up, its never been of sufficient magnitude to be noticed or recorded in my brain.
If you see significant movement RA/Dec when on the celestial equator in the East/ West, then I can only think the Azimuth setting of the polar axis must be incorrect.
Somethings not right here?????

[edwardsdj]

Thinking about this some more, I'd completely eliminate all Dec drift without any regard to the RA drift. When there is no Dec drift wherever you point in the sky the mount must be correctly polar aligned.

Any remaining drift in RA must then be a separate problem unrelated to the polar alignment and I would investigate it as such.

Have fun,
Doug

[g__day]

Doug,

That is the same conclusion I have reached. I checked with a mate who is an electrical engineer - to check power purity is not an issue - the thing is if it where it should manifest in only one sky position.

It could be the SkySensor2000-PC somehow adjust for refractive index incorrectly - why the problem dissappears as elevation climbs - but I want to totally dismiss polar mis-alignment before I look for exotic issues!

Merlin

Yes I do have a 12 mm illuminated reticle, it should aid qualifying exactly what I am seeing. I have spreadsheet now where if you enter a stars elevation and a specific angular degree of drift over a specific period of time it calculates the precise mis-alignment in both axes - I think I will make use this to esitmate how much of an adjustment is required!

The Canon viewfinder has very fine lines around the boxes - so in reality they do exactly the same function as the reticle - given I've aligned the X and Y axes to be RA and DEC drift there is little room for mis-interpretation.

It could easily be that there is a subtle error and/or I am making a basic mistake.

[g__day]

The other thing I was considering - could this be some sort of wierd mirror flop - that is only discernible at low elevation - and doesn't seem to distort stars?

If so I guess you would expect to see repoical behaviour when point due West.

[edwardsdj]

It's a shame you can't just use a laser collimator like in a Newtonian to see any effects of primary mirror movement with altitude.

[g__day]

I hate weird behaviour that is hard to diagonise and correct. It takes so much patience to sally forth trying to get the imaging platform to its full potential. Even the clouds are annoying me now - you get things set up - like tracking a bright star at say 3pm - come sundown - or an hour before - and its cloud central!

grrrrhhh!

[Merlin66]

After all the "inputs" did you manage to resolve the problem??? If so, what was the answer?? Love to know.

[g__day]

Not yet - upgraded the SkySensor2000-PC from V2.05 to the latest V2.10 EPROMs mid week - now waiting for skyies over Sydney to clear - it's been lousy with clouds for most of this month.


Went I get somewhere - or more data - will keep folks posted!

[g__day]

Very interesting read here http://tech.groups.yahoo.com/group/ss2k/message/7076

Saying suppling any inverted A/C power into the Atlux / SS2K-PC be it the mount Mount or indirectly via an A/C powered PC via auto-guider cable into the hand controller can cause this - more investigation when the weather clears.

* * * * * * * * * * * *

--- In ss2k@yahoogroups.com (http://tech.groups.yahoo.com/group/ss2k/post?postID=a7T6WSrnghwfE2LbKFxIyKk Sa8U3wgfMmNTRdP_jVwnl-wfq5KarQVNg1lbsXE4DF1u66kyWqTjkz4_B 8uxU), "JAMIE FLINN" <jamiecflinn@...> wrote:
>
Have I got some info for you!!!!

I have been dealing with this for a few years and last month isolated the problem (at least for me)....AC current:

perform the following control test using a guide camera if you have one (reticle eyepiece if not)...all of this assumes you are using "house" power to some kind of 12v inverter...

1. hook up your system as normal - AC to inverter to ss2k - rs232 between ss2k and computer (also on house power) if you use that - there are various combos here to try
2. hook up your guider as normal
3. polar align - do 3 star alignment - make sure you are in polar mode
4. while watching the guide star slew your mount just a bit to make sure you know which way the star moves to be too fast (sidereal + whatever)....center your star
5. watch closely and make sure without touching anything you are seeing RA drift only (at least not massive dec drift because dec drift also induced ra drift)
6. *****very quickly, swap house power to the mount for a battery pack....turn on ss2k and press ESC (your 3 point alingment will still be there)
7. centre and watch the star again......if you see drift in RA still and have the computer attached via rs232 to the ss2k PULL THE PLUG on the computer (assuming laptop which has battery)
8. watch the star - at this point you may see the drift stop dead - it did on mine - I found that house ac connected in any way - even through a line filtering UPS (which made the drift slightly less but still there) -
I had drift - BUT NOT ALL THE TIME > This lead me to believe that the power supply to my house was subject to "sour power" - noise from airconditioners, electronics etc...

Once I switched to all battery power with no ac contact, I have been
steady as a rock. This problem also plagued me at StarFest every year
using the campground power - by the time the place was full and there
were fridges and fans and scopes and who knows whatelse, I would be
drifting almost out of control

Here are my combos

AC -> INVERTER -> SS2K = DRIFT
AC -> COMPUTER->RS232->SS2K = DRIFT
COMPUTER ON BATTERY->RS232->SS2K + SS2K ON BATTERYPACK->AC (AS IN CHARGING THE PACK) = DRIFT (I COULD NOT BELIEVE THAT ONE)
AC->UPS->COMPUTER->RS232->SS2K = DRIFT
BATTERY->COMPUTER->RS232->SS2K = NO DRIFT
BATTERY->SS2K (NO COMPUTER) = NO DRIFT
BATTERY->COMPUTER -> RS232->SS2K + BATTERY->SS2K = NO DRIFT

I also played with shielding every cable attachmednt to no avail....I now own a huge 60amp/h battery supply for my laptop and cameras, and use 30 amp/h battery packs for the ss2k and heaters

[g__day]

Well I did more drift alignment until I could see only RA drift - no DEC drift at all.

Well when the scope is pointing straight up - I see no RA drift, when its lower to the horizon - it seems like 1 arc minute of drift per 30 minutes.

Well I checked by GPS co-ordinates and tried running everything off DC power (a Celestron 17 amp hour power tank) and did this with no PC or auto-guider connections to the hand controller.

So I have a few things extra to check -

1. Is RA too fast both East and West?
2. If I lower the hand controllers Lattitude adjustment will this slow down the RA rate?
3. If I switch to Polar unaligned mode - with just one star selected will RA be correct?
3. Same as above - but do a two star align and try adding a small delta to the RA - will this slow down the RA?

Also I found there is both a SkySensor2000-PC and a SS2K Yahoo group - where folks have seen my situations before.

So I hope to progress matters further!

[edwardsdj]

Hi Matt,

Glad to hear you're making some progress :)

Did the move to DC power help?

Are you using some sort of King tracking rate that takes atmospheric effects into account?

Have fun,
Doug

[g__day]

Doug,

The move to DC power did nothing. The SkySensor2000-PC equates for elevation related diffraction (but I wonder if its spot on correct).

More testing shows me at elevation of 27 - 55 degrees I see no DEC drift but 1 arc second of RA (too far West = too fast) every 30 minutes. At elevations above 80 degrees the I see less than 1 arc minute of drift in RA (too far West) every hour.


alah - my notes:

http://tech.groups.yahoo.com/group/ss2k/message/7240
http://tech.groups.yahoo.com/group/ss2k/message/7238
http://tech.groups.yahoo.com/group/ss2k/message/7237

Matt

[g__day]

Just thought - if I check elevation on every 20 minutes and re-centre target - and if the the SCP alignment was out by the average elevation of the star Sirius but West and below the SCP by say 37 degrees - I could hit a Larange point where at 20 minute intervals there would be no DEC drift in elevation between 30 - 50 degrees - but the longer angle could cause a speed up or RA - relative to the angular rotation from the SCP alignment position. If the angle was just right and the delta was the right size I would see an apparent 1 part in 300 speed up in RA!

More research still!

[edwardsdj]

Maybe you would but that's hardly your problem :)

I'd be tempted to try without that SkySensor2000-PC altitude based atmospheric refraction model and see if the RA drift gets any better or worse.

[g__day]

Doug,

Its hard wire in and you can't turn it off. From memory its only 1 arc minute at Sydney latitude at elevation 20 degrees and it decreases as you elevate the scope (helps having a dad who was a navigator and surveyor - amazing the old school has all this knowledge at their finger tips).

So refraction alone isn't enough to account for this degree of tracking error - unless they modelled it really badly.

I think I will experiement with maxpoint to build a 50 star sky model and see what it thinks my alignment errors are. My gut tells me I somehow just slightly blew polar alignment in an esoteric way that is producing what I am seeing naturally!

[edwardsdj]

Hi Matt,

Sorry to misunderstand. I just thought you had eliminated dec drift in all sky positions. I agree that a polar alignment error will manifest as combined RA and Dec drift.

This is why ideally, drift alignment would be done on a star on the horizon due East or West (where altitude error is maximally manifest as drift in Dec) and at the intersection of the equator and meridian (where azimuth error is maximally manifest as drift in Dec).

If there is no Dec drift for all sky positions (East, West, North, South, Zenith) then the polar alignment must be correct. I'm sure the effect you are referering to would only exist in two particular positions in the sky. Maybe I misunderstand?

Good luck with it!

Have fun,
Doug

[g__day]

Doug,

What I refer to should only occur at one point in the sky - so my thinking or observations must be a tad wrong!

I did a 50 star model with Maxpoint - raw pointing error was 158 arc seconds that Maxpoint brought down to just 34 arc seconds since I've tuned DEC. So for a 12mm eye piece nearly every target was close to dead centre.

MaxPoint diagnosed my alignment to be about 50 arc seconds West and 30 arc seconds lower than the SCP. So being within one arc minute is so far not bad.

If I could just cure this rotten tracking rate - arrrgggh!

I did take the chance when I removed my Canon 400D to visually observe NGC 2070, Dor and Saturn with Vixen LVW eye pieces 22mm, 17mm, 13mm, 8mm and 5mm - about the first visaul astronomy I've done in 6 months! They looked good but I think my collimination is just slightly out! In the next two months I will probably ask Bintel to service - clean front corrector plate - check focuser backlash and re-colliminate my main OTA.

I finally got the chance to mount the Meade DSI on the main OTA whilst the scope was targeted high at the Meridan (where drift appears minimal) I trained PEC using PHD to auto-guide on the main OTA. The worm was pretty smooth bar once spot - corrections were at peak under 10 bar one spot that reached about 20, finally summation of the 300 step 8 minute process was a net delta of about 2 - so that's looking good too.

So next clear night I plan to drift align with my piggy backed 80mm Megrez refractor and/or with my 127mm MAK and see if I can see RA tracking error there. If I don't that kinda says its the mirror or imaging train of the main OTA that is giving my observed shift.

Finally I may track Sirius on my main OTA from elevation 30 - 90 or 100. If in 4 hours I see a drift of say 4 - 12 arc minutes which increases then decreases I will have more data to deduct with. I can't see how pointing can be superb when tracking appears to have problems - something doesn't add up there.

On night two I might centre on Sirius on my main OTA and every hour into the tracking issue a goto Sirius and see if it moves at all.

Thirdly I may try a 3 star alignment and see if tracking improves or worsens when its running in polar aligned mode.

I do feel all the diagnostics are helping to hone in on the cause of the error!

thanks all,

Matthew

[Merlin66]

Going back over this thread, and remembering it's an SCT; I'm beginning to think it may just be mirror flop movement?????

[EzyStyles]

i just want to say ..... AUTOGUIDING :)

I can't even do 30 sec exposure shots without trailing. No perm observatory, a 'really' rough polar alignment and plonk the EQ6 into the marked concret ground thats it. PEC what? never heard of it ;)

I remember i spent heaps of time adjusting drift alignment, getting polar alignment to be exactly in the end, problem all solved with Autoguiding. :thumbsup:

Anyway g'luck with it.

[g__day]

Well I put the illuminated reticle on my refractor and and tracked Canopus for an hour - it barely moved off the cross hairs - and at high elevation it was excellent.

So after all this hassle - it starting to appear that the mount is doing all its supposed - gotos from the hand controller are landing very close to dead centre - and from MaxPoint - well its five times better!

Auto-guiding therefore isn't really going to help me if its mirror flop in the primary - unless you do it at prime forcus with a S-BIG!

So I will experiment more to confirm these findings - and I'm happy that my most expensive piece of gear is performing right to spec - but now I have to seriously think how do I solve mirror flop.

I guess on the cheap I could try some sort of off-axis guider - but my shopping list at the moment is looking like:

1. Off axis guider
2. 4" - 5" APO
3. S-BIG
4. Electronic, computerised focuser for the APO

Oh dear! There goes any sanity or fiscal responsibility!

[tempestwizz]

Personally, I agree with EziStyles. Autoguiding fixws most , if not all problems. I was a little dissapointed when I took delivery of my Tak EM200 mount. There is no PEC feature. PHD guiding for my system was/is plug and play. I have not (yet) found the need to make any adjustments. I am up to 5 minute exposures with no noticable (to me at least) problems.
Mind you, I am only using a 500mm focal length scope, with a DSLR imager.
Higher fl's and smaller imaging areas will magnify any tracking/guiding errors.

HTH
BC

[EzyStyles]

If your getting mirror flop then polar alignment, autoguiding, pec, drift etc etc will make no difference. Guess fix that problem first.

ps.. mmmm. sbig...mm..drool.

[g__day]

Yep - mirror flop can't be fixed other than by adjust how the mirror moves on the tube - or on axis guiding.

[Karls48]

Just for interest – you can do it with GStar. This camera got two outputs (Composite Video and S-Video) that can be used simultaneously. So you can guide with AVI and capture with S-Video. You would need fast laptop with two frame grabbers. One PCMCIA card and other USB.

[g__day]

Yes - given you use an on axis shot!

I have done great 10 minute shots unguided at the full 2.3 metre focal length - so in some parts of the sky the mirror isn't moving too badly on the baffel tube - in others it is.

What I have to work out is why and if its addressable!

[g__day]

Hmmm,

Spoke to BinTel today about Lumincon on axis guiders and what I thought was mirror flop - they say that's unusual because usually mirror flop is:

1. a sudden shift - not smooth throughout the movement hour by hour and
2. most often seen during a meridan flip

More research required still - will move the weights a tad and see if this worsens or improves matters!

[edwardsdj]

I agree. That is when you see mirror flop. You also see it when changing the direction of the focus knob. If you always find focus by making the final adjustments with the mirror moving against gravity, the mirror position should hold unless you move east/west to the other side of the meridan or north/south to the other end of the sky.

[g__day]

Moving the weights down has improved things a fair bit - so maybe things where out of balance! Which would account for poor RA rate at some elevations, not all.

Plus drift guiding with a MAK took some learning - it requires moving the scope in the opposite directions to my SCT.

[g__day]

How switching from polar unaligned mode to polar aligned mode affects tracking


I initiated my mount in polar aligned mode and synchronised on one star. I thought my polar alignment was good because of several factors: pointing was superb, MaxPoint modelled alignment on fifty stars and said I was within 30 arc seconds of the Celestial Pole and I could see no minimal DEC drift on 1 hour shots.

But the mount appeared fast in RA to the tune of 60 – 120 arc seconds an hour – more at low elevations and less at higher ones.

After seeing most people’s comments I ran through several configurations of isolated DC power, filtered and conditioned mains power – but nothing helped.

Next I tried more carefully balancing the counterweights – this slightly improved things. Then I adjusted the DEC gears to reduce slop – this brought pointing to within 34 arc seconds. But tracking persisted to be too fast – all over the sky, not just East or West.

So I reverted to polar unaligned mode and added 1 star at a time up to 3 stars and checked the tracking at each point – still no improvement.

Then I switched from unaligned mode to polar aligned mode and executed a goto on one of the sync stars I was currently pointing at – and it moved over 30 arc minutes!

Suspicious of this I ran long sessions with PEMPro V2.0.39 which finally has Meade DSI support directly and it informed me I had to move my mount Eastwards about 40 arc minutes. This puzzled me given the other checks and balances said I was fine, but I experimented and suddenly tracking was a whole lot better.

Now I noticed pointing was a bit off, and I went to inspect my settings, three things surprised me. The first it said my polar elevation was 33.9, which about 0.1 degree too high for my precise latitude. Second Az said it was 180.1, but I expected for Sydney SCP is 11.4 degrees East of magnetic South – so I expected to see this. But the big one was my polar alignment coordinates. I expected them to read something like:

0 +123 + 123 +0 +0

But they didn’t, they read:

0 +120 +75 +54 + 48

Now the first two three digit terms are simply adjustments required from the scopes initiation position in case it’s not perfectly pointed East and level at power on. But the last two coordinates are more cryptic, poorly documented axes adjustment parameters. In polar aligned mode I expected these to me zero.

It appears switching from unaligned to aligned mode left these parameters non-zeroed. The result – well tracking Sirius (due East) elevation 30 degrees – 45 degrees for one hour show no RA drift and maybe 10 seconds of DEC drift – sensational!

I ran several more runs of PEMPro – which generally makes clouds appear, until its corrections was coin toss +/- 0.2 arc minute adjustments over a 20 minute run, which I imply means get a better CCD!

I also synced several times on each of three stars – until goto’s where spot on for them, and noticed they weren’t the first few iterations (about 2 arc minutes off in DEC). Also after each alignment on Sirius, Canopus and Acrux I check the alignment parameters – and noted the last two digits of the polar alignment coordinates were varying wildly – from -300 to +300, generally with only a few points separating the two values.

So let me sum up where things are at. I was surprised too see Maxpoint and PEMPro disagree significantly – but I guess PEMPro was correct. It was too late to confirm tracking accuracy – but if its under 20 arc seconds an hour, and mostly in DEC – then I’m smiling. I am still suspicious pointing has detoriated (but MaxPoint should resolve this) but will live with it if tracking is back to superb!

Lastly I really would like to know what those two parameters of the alignment model are. I ponder if it was that – rather than the rotating the mount East 30 – 40 arc minutes that significantly improved tracking!