Help: HEQ5 pro guiding gremlins frequent 12 arc second spikes
I have a heq5 pro mount (with rowan belt mod) and have all but given up on astrophotography because of this guiding issue. Scope is a 100mm Apo. asi cameras (asi 533 and 120mm for guiding) using eqmod pc to mount (and also asi air plus).
Essentially I have these random spikes that show up on the guiding graph that result in a double image of stars a few star widths apart. They happen frequent enough that shooting at over 30 second exposures results in half the shots being ruined
PHD2 guidelog image attached. (note the final part of the image I changed dec aggressivenes to 5% hence why that correction takes so long)
So things of interest, spikes are on RA and Dec! mostly RA though
Spikes are both above and under the graph (so tracking can be ahead and behind). Between spikes guiding is quite good.
(errors on Dec and RA on either side is what really confuses me)
Errors are fairly consistent in size 12.5 arc seconds. (Note <for scale> if I tap the guide cam with my finger nail I get a 4 arc second spike.)
Its possible this started after a night of the telescope mount being stuck grinding the telescope against the mount.
Things I've mostly ruled out.
Power supply, tried 3.
USB cables, tried 2 (no hub in use)
Computer, tried PC and ASI Air plus
Wind/wires/wobbly random bits.
I've opened it up and cleaned/regreased
I've tried east heavy/west heavy
I've tried minimise backlash, I've tried loosening effectively adding backlash.
I tried switching the RA and dec motors
I can hand turn the pullies freely through full 360 on RA and dec without binding or overly heavy resistance
Slewing is fine
Balance is good.
Calibration is good.
I'm kinda thinking main pcb issue bit I dont know.. and its an expensive guess.
How did the steppers feel, could you turn them easily a full 360 with no binding? I had something similar when my stepper motors had failed. They would miss steps making guiding unreliable. When I removed one I could feel it grinding rather than turning small steps at a time. Also check Eqmod logs make sure there aren't any errors over serial.
Its possible this started after a night of the telescope mount being stuck grinding the telescope against the mount.
That doesn't sound good.......
The only thing I can add is reset PHD to default, but such a big movement in RA and Dec probably not.
What happens if you choose the guiding assistant and let it run ?
Are there any bad, odd noises coming from the mount, especially when you get these big spikes in the graph?
How did the steppers feel, could you turn them easily a full 360 with no binding? I had something similar when my stepper motors had failed. They would miss steps making guiding unreliable. When I removed one I could feel it grinding rather than turning small steps at a time. Also check Eqmod logs make sure there aren't any errors over serial.
Am not sure how it should feel but its not grindy. And thats also why I swapped the dec and RA motors to see if there was any different on the assumption it was hitting RA all night)
I do get occasional eqmod comms errors, but I always have ,which is why I had to disable the limits check, and how I had the 'accident' in the first place
Also when tracking there isnt actually any commands being sent the on board PCB handles tracking pulses eqmod just sends commands to start and stop tracking and any change of tracking rate.
That doesn't sound good.......
The only thing I can add is reset PHD to default, but such a big movement in RA and Dec probably not.
What happens if you choose the guiding assistant and let it run ?
Are there any bad, odd noises coming from the mount, especially when you get these big spikes in the graph?
Jeremy
I spent about 20 minutes watching the gears turn (with little bits of tape stuck to them) and listening to the steppers and I couldnt notice anything untoward when a spike happened. Am using asiairpro now which doesnt have guide assistant but calibration looks good. when my laptop was being used(its died) I did try resetting to default and doing full guide assistant run without any improvement.
I don't have a Skywatcher. If it's happening regularly what does that period relate to - one rotation of the worm shaft, stepper motor, etc.
rotating the steppers manually should feel like even clicks. If not that, can you disengage the steppers and manually turn the worm shaft - it should feel smooth while turning.
I don't have a Skywatcher. If it's happening regularly what does that period relate to - one rotation of the worm shaft, stepper motor, etc.
rotating the steppers manually should feel like even clicks. If not that, can you disengage the steppers and manually turn the worm shaft - it should feel smooth while turning.
As far as I can see theres no set period to the spikes, they can be 5 seconds apart several minutes apart and anything inbetween.
Yeah hand turning the drive shafts is smooth and non binding.
I don't have a Skywatcher. If it's happening regularly what does that period relate to - one rotation of the worm shaft, stepper motor, etc.
rotating the steppers manually should feel like even clicks. If not that, can you disengage the steppers and manually turn the worm shaft - it should feel smooth while turning.
I’ve owned a HEQ5, one worm cycle is 9.6mins
IMO a regular short period 12arc sec spike is either intermittent spurious data drop out during a guide pulse cycle or physical damage to the worm gear or main gear on the Ra and Dec.
My hunch is the latter due to the scope hitting the mount and grinding
You couldn’t visually see damage to gearing causing a 12arc sec spike ( it’s microns )
It’s a tough one to find.
I’ve owned a HEQ5, one worm cycle is 9.6mins
IMO a regular short period 12arc sec spike is either intermittent spurious data drop out during a guide pulse cycle or physical damage to the worm gear or main gear on the Ra and Dec.
My hunch is the latter due to the scope hitting the mount and grinding
You couldn’t visually see damage to gearing causing a 12arc sec spike ( it’s microns )
It’s a tough one to find.
so I did a run with no dec guiding, and I still saw Dec spikes of about 12'' so that kinda ruled out pulse data issues for me (Unless theres something really bad happening inside the pcb with data crossover) its annoying asiair wont let me disable RA guiding pulses too.
As I understand it the large brass cylinder of the worm is brass specifically to be the weaker element that might get worn away (compared to the steel(maybe) worm screw part). My assumption is that if this was worn (as you say I cant see anything) it would be worn in one place not worn all 360 degrees, at worst 8 hours out of the 24 hours for the full circle might be worn, so I could expect to at least have a 25% chance of it not be using a bad bit of the worm gear (assuming I've randomised it enough).
I find the dec spikes the most incomprehensible. Without Dec guiding, none of the dec should be doing anything! which implies somehow a fault is in the RA that somehow produces an error in the Dec (sometimes) but no spike in RA sometimes. Which seems unlikely.
I guess if comms were intermittently scrambled an RA pulse could be received as a Dec pulse. I think i need to get my laptop back online so I can run a test with PHD2 just monitoring but not issuing any pulses. Having said that though, I did set the RA only guiding to be 5% aggressive, and didnt see any changes in the sizes of the 12"" spikes. I wonder what the data packets for pulses look like.
Hi Michael,
In order to rule in OR out mechanical versus electrical or software issues affecting the guiding I would first do a back to back comparison, during the same session, of images with guiding versus images without guiding and look to see if the double stars you saw previously in even 30 second images are present in the unguided images. If they are not present I would focus my attention on the electronics and software rather than mechanical drive or wear issues.
So initial tests (with asiair plus) doing 60 second subs guided and unguided have some interesting results.
Unguided appears to NOT have the double stars, but does have RA streaks with a similar frequency to guided (approx every other image).
Also trying different guide rates the spikes are proportional to the guide rate
0.5 guide rate produced 12.5'' spikes 0.9 guide rate produced 25 '' spikes and 0.25 guide rate produced 6'' spikes (Not sure what to make of that)
attached graphs and small subsection of the image of M20 as well as a full image.
Note scope is f9 100mm apo, so 900 FL no field flattener. using l extreme filter, cam is asi 533mc pro
The intermittent streaks in unguided imply to me its mechanical. But the guide rate being proportional to the spikes makes me thing electrical/comms So
So initial tests (with asiair plus) doing 60 second subs guided and unguided have some interesting results.
Unguided appears to NOT have the double stars, but does have RA streaks with a similar frequency to guided (approx every other image).
Also trying different guide rates the spikes are proportional to the guide rate
0.5 guide rate produced 12.5'' spikes 0.9 guide rate produced 25 '' spikes and 0.25 guide rate produced 6'' spikes (Not sure what to make of that)
attached graphs and small subsection of the image of M20 as well as a full image.
Note scope is f9 100mm apo, so 900 FL no field flattener. using l extreme filter, cam is asi 533mc pro
The intermittent streaks in unguided imply to me its mechanical. But the guide rate being proportional to the spikes makes me thing electrical/comms So
Maybe theres 2 issues..
The main determinant for me is the obvious star trailing (RA streaks) on UNGUIDED images 3,5,7 and 8, but NOT, despite some minor star ovality, on the other UNGUIDED images 1,2,4,6 &9. It clearly demonstrates a loss of drive to the RA axis, probably for just under one second*.
I suspect, when the guiding is TURNED ON, the guiding system tries to compensate for the apparent star movement (caused by this intermittent loss of drive) and the guiding system injects that large 12.5 arc second correction pulse.
*. It would be good to know the length of the star trail in UNGUIDED images 3,5,7 and 8 in original image pixels. Can you zoom in on the RAW 3008x3008 ZWO533 image in Photoshop and measure the length of the star trail in pixels, either diagonally or as so many horizontal pixels x vertical pixels. I suspect it to be around 14 pixels based on your image scale and the possible tracking intermittent failure time associated with the 12.5 arcsecond correction pulse, whilst tracking at the sidereal rate, but who knows? All info helps solve problems.
Electrical issue more likely, given the apparent randomness/Irregularity(?) you reported of the fault, who knows.... more info
BTW - have you ever tested with everything PLUGGED OUT except for the mount, a camera (preferrably DSLR) and power to see if there's still star trailing?
Best
JA
Last edited by JA; 16-06-2022 at 11:37 AM.
Reason: typo
Things I've mostly ruled out.
Power supply, tried 3.
...
Can you advise the specs (Output Voltage & Max current) of the 3 power supplies you tried? Or pictures of the makers impu/output rating label on the devices.
These mounts are somewhat marginal at a power supply voltage of 12Volts and many suggest their use at higher voltages, typically 13-14Volts. I believe even the manufacturer's spec includes such a range. I will check and repost a reference.
EDIT: I just checked and Skywatcher specifies 12 to 16 Volts DC 3A, see red circled spec in pic below...
EDIT2: I also found a Skywatcher USA site that specified it as 11 to 15 Volts DC 2A (so they're not even consistent ??)
The main determinant for me is the obvious star trailing (RA streaks) on UNGUIDED images 3,5,7 and 8, but NOT, despite some minor star ovality, on the other UNGUIDED images 1,2,4,6 &9. It clearly demonstrates a loss of drive to the RA axis, probably for just under one second*.
I suspect, when the guiding is TURNED ON, the guiding system tries to compensate for the apparent star movement (caused by this intermittent loss of drive) and the guiding system injects that large 12.5 arc second correction pulse.
*. It would be good to know the length of the star trail in UNGUIDED images 3,5,7 and 8 in original image pixels. Can you zoom in on the RAW 3008x3008 ZWO533 image in Photoshop and measure the length of the star trail in pixels, either diagonally or as so many horizontal pixels x vertical pixels. I suspect it to be around 14 pixels based on your image scale and the possible tracking intermittent failure time associated with the 12.5 arcsecond correction pulse, whilst tracking at the sidereal rate, but who knows? All info helps solve problems.
Electrical issue more likely, given the apparent randomness/Irregularity(?) you reported of the fault, who knows.... more info
BTW - have you ever tested with everything PLUGGED OUT except for the mount, a camera (preferrably DSLR) and power to see if there's still star trailing?
Best
JA
Am at work at the moment but I did zoom in on the really bad star trail (image 3) and it did look a similar length to the 12.5 arc second errors.
The short bursts of no tracking is what I assumed the original spikes to be. sidereal being 15''/s so I guessed at just under a second of no tracking to get the spike.
And I have had cases of just tracking the moon (during a lunar eclipse) over hours the mount falling behind where it should be
I guess in asiair after I changed the guiderate I should have recalibrated, but that doesnt really explain the doubling in spike, the spike should happen first (2 second exposure on guidecam) then the correction should happen, and if it needed a recal the correction should just be over or under.
Re power supplies am at work at the moment so will have to get exact details later. But I know the one on the mount is 14V and at least 3Amp (and was bought from possibly RS components) the power isnt shared its dedicated to the mount. The asiair has its own power supply shared with the asi 533 , some ebay 12v 6a power brick. the 3rd unused is also an ebay job.
Are you using a guide scope or oag? Random spikes can be caused by differential flexture between cameras when using guide scopes.
Guide scope but my tapping on the guidescope produces a 4 arc second spike which I figure has to be more than any flex would produce. and am seeing 12 arc second spikes
As far as diagnosing the root cause of the intermittent star trailing problem you are "LUCKY" in a way that the problem relates to an apparent intermittent, but recurring, lack of tracking. Why do I say "lucky"? Because you can dispense with the time consuming testing required with mount setup, aligning, guiding, imaging etc.... All you need to do is focus on the mount itself for the moment. Later when the cause is narrowed somewhat, and some changes potentially made, retest it for tracking using imaging.
We only need the mount, its power supply and means of starting its tracking. Perhaps a hand controller would be easier than the ASI Air which might inject more variables. Later we could retry to initiate tracking with the ASI Air to confirm.
Call this Home diagnostics 101 for your issue:
1. Remove the HEQ5 Pro side panel to expose the RA and DEC drive motors and gears.
2. Inspect the gear drive for anything untoward, excess wear, tooth problems, debris in the drive train,etc...
3. Check that the RA pinion gear and RA driven gear are locked tight on their drive shafts.
4. Check that the DEC pinion gear and DEC driven gear are locked tight on their drive shafts.
5. Check any backlash present.
6. Power the mount ON and go through the hand controller start up procedure and set TRACKING ON.
7. FOCUS your attention very closely on the RA pinion gear. Based on Skywatcher's specified 705:1 gear ratio I would expect the RA motor pinion to make 705 revolutions per 24h in which time the RA axis will revolve once. If we factor that down further the RA pinion gear will move ~29 revs per hour or ~0.5 revs per minute or ~3Degrees every second. And that's the kicker. We know from previous tests that the mount looses tracking for about a second every few minutes. I think that , although difficult, and perhaps with a magnifier or video you should be able to see the RA motor pinion STOP for about a second every 2-3 minutes as per your previous findings. Bear in mind the RA motor pinion is, I'm guessing, something like 10mm in diameter then if it stops for ~ 1 second a point on its circumference will not have moved the requisite ~0.3mm. That's the size of the change you should focus on, hence possibly use a magnifier.
8. Observe whether the issue also exists with the DEC pinion as per step 7.
9. If you have been able to see the fault as described (the intermittent stopping of the tracking), then you are lucky because you now have a test bed on which you can make changes to help you search for the root cause. Try different power supplies, reseating any motor connectors You need to remove another cover for that). If you are electrically inclined or have a friend who is you could look at the motor drive voltage/current and see what happens to it during a stoppage of drive. Also check the power supply voltage as a function of time. A storage oscilloscope function would be great for those tests. I could imagine that the mount crashing in to the tripod may have overloaded the drive electronics and potentially damaged the output drive, fried an output transistor(?) depending on the circuit protection.(???)
10. Retest above with ASI Air control of tracking ON to the mount incase somehow there is an intermittency to the cable????
The fact that this loss of tracking is presumably occurring in RA and DEC suggests a common (probably electronic) cause in the drive circuitry or problems with the internal mount power or power supply.
It has a rowan mod.
I happen to have been down this road before back in april, where I tried to use a time lapse mode on my phone to look for issues filming the gears.
I think that was like 20 mins of realtime but cant remember.
I think I need to try doing this with the mount loaded up, maybe for a few more hours. I think I came to the conclusion that the errors are too small to see anything going wrong, even timing rotations.
Will absolutley double check the grub screws on the motor drive shafts are tight. But I did try (months ago) swapping the ra and dec motors without any benefit seems unlikely boath would have been slipping but who knows.
@JA I'm skeptical I am 'lucky' because I cant see how the tracking issue would cause such huge spikes in dec. I think maybe theres multiple issues at play here. But yeah solving RA would be a great great start and you never know might mysteriously fix dec.
That may have been helpful to know, but mostly the same applies in terms of physical checks, other than to suggest checking the belt tension is not too low or some possible belt damage.
Quote:
Originally Posted by dizzy2003
I happen to have been down this road before back in april, where I tried to use a time lapse mode on my phone to look for issues filming the gears.
I think that was like 20 mins of realtime but cant remember.
I think I need to try doing this with the mount loaded up, maybe for a few more hours. I think I came to the conclusion that the errors are too small to see anything going wrong, even timing rotations.
Will absolutley double check the grub screws on the motor drive shafts are tight. But I did try (months ago) swapping the ra and dec motors without any benefit seems unlikely boath would have been slipping but who knows.
Thanks for all the advice so far BTW..
Mike.
I agree that's why I suggested it is more likely there's a common cause such as electronics / power or supply
because I cant see how the tracking issue would cause such huge spikes in dec. I think maybe theres multiple issues at play here. But yeah solving RA would be a great great start and you never know might mysteriously fix dec.
, but mostly the same applies in terms of physical checks, other than to suggest checking the belt tension is not too low or some possible belt damage. 
