Mountings and Controls For Widefield Imaging Rig

[Gina]

No, I haven't.  Good point - could be earth loop problem - I'll try that - thanks

[Gina]

As a slight diversion from the focuser problem, I've been looking at the rotation drive and the possibility of an INDI driver to control the stepper driver directly as the Arduino approach is not very nice and the upshot is that the present rotation gear ratio of 10:1 is simply not suitable, resulting in very difficult control ratios in the code.  As stated further back in this thread, a 10:1 ratio results in a degrees to steps ratio of 5.55 recurring.  OTOH a 9:1 ratio would give 5 steps per degree - much better

I could stick an extra gear with 90 teeth on top of the current 70t gear and use a 10t pinion to give a 9:1 ratio.  I would need to alter the end bracket design and reprint it but that's no problem and well worthwhile to make other things easier.

[Gina]

The present gears have mod 2 teeth which seem rather large.  If I were to reduce the tooth size a bit I could make the ratio 18:1 giving higher torque to the cage and smoother movement.

With the number of teeth on the pinion needing to be a whole number, the number of teeth on the large gear needs to be a multiple of 9 or for the smoother drive, a multiple of 18.  1:18 gives 5:90, 6:108, 7:126, 8:144 or 9:162.  I've found around mod 1.5 is good for 3D printed gears and to clear the teeth of the present large gear, the new one wants a diameter of about 150mm or more.  Gears of 6 and 108 teeth mod 1.4 gives a large gear diameter of 151.2mm.  The pinion would be 16.8mm pitch diameter.

[Gina]

4 hours ago, tekkydave said:

Have you tried running the logic ground directly from the RPi?

3 hours ago, Gina said:

No, I haven't.  Good point - could be earth loop problem - I'll try that - thanks

Hmm... Motor doesn't rotate at all now   Checked power and ground connections on driver pins and +3.30v on logic and +13.6v on motor supply.  Rechecked all data connections.

[Gina]

Well, I don't know - tried everything I can think of   Except for starting again with a new circuit construction!  I guess that's what I'll have to do. 

I suppose I could just make a box for a simple setup without camera rotation.  ie. replace the Arduino focuser with the INDI RPi version on my earlier simple rig.  I have a HAT with a stepper driver and three power MOSFET heater/cooler circuits that I no longer need as I've done it with separate boards in my mono ASC.  I can take the MOSFET circuits off to make room for the buck converter.

[tekkydave]

The circuit I built a while back for experimenting with Indi used 3.3v for the motor driver and I had no issues. Its a weird one. Have you tried manually forcing the DIR line hi/lo to see if the stepping direction changes? Also is the current limit pot set at a sensible level?

[Gina]

Yes, tried DIR line to 0v and +3.3v one rotates the other doesn't!  See a little way back in thread.  How weird is that!!  Tried several drivers including one I've used before and know works.  I find the preset pot setting is fine.

[Gina]

No widefield imaging tonight - can't get anything to work   Sky seems to be clearing.  I have the Esprit 80ED scope rig still indoors mounted on my NEQ6 and pointing out of the east window so if the sky is clear when it gets dark I can see how well auto-focus works.  Assuming the focuser is still working - it was fine a few days ago.  I just have to accept the fact that I'm not as good as I was and if I can't get things to work, so-be-it!  There's plenty I can occupy myself with.

[Gina]

Been designing the gears for 3D printing.  The pinion from the calculation works out at 8.4mm not 16.8 as I posted above.  This makes the pinion too small for the 5mm shaft of the motor.  Basically this means that the pinion needs to be bigger at the base of the teeth and hence smaller teeth.  With 7 teeth and 4.5mm pitch radius the base circle of the teeth is more than 5mm diameter.  Keeping the 1:18 ratio gives 7:126 and a modulus of 9/7 = 1.3.  The large gear will be 162mm diameter pitch circle.

[Gina]

Printed the gears but I think the tooth pitch is too fine really, particularly for the pinion.  I might try reprinting the pinion at higher printer resolution though before I redesign for 9:1 instead of 18:1.  The large gear looks fine.

[Gina]

Pinion still not very good.  I could try using the UP Plus 2 printer and 0.1mm resolution.  These small prints are really pushing my Pilot printer.  Or I could try different filament as the problem is slight blobbing on the sides of the teeth - I'm currently using very cheap filament.  Resolution at 0.2mm seems adequate as the teeth are well shaped except for tiny blobs in places.

[Gina]

Pinion printed much better with different filament but the teeth on the pinion seem rather frail so I shall go back to 9:1.  I've looked at the possibility of something between 9 and 18 but with micro-stepping always being powers of two and 9 not divisible by 2 we are left with multiples of 9 as the gear ratio.

[Gina]

Gear sizes can be 153mm and 17mm diameter hence radii of 76.5 and 8.5 with 90 and 10 teeth.  Now printing...

[Gina]

Gears printed and seem fine   Now printing a modified end bracket and stepper motor holder to cater for the larger distance between centres of the gears.

[Gina]

Bracket printed, holes tapped for the fixing screws and attached to dovetail bar.  Cage part assembled with new gear glued on an fitted into supports.  The gears mesh very smoothly   I'll take and post a photo or two later.

[Gina]

Back to trying to get stepper motor working from Astroberry Focuser.  Tried a brand new RPi 3 B but still no joy.  Also, tried a NEMA17 that I know works.  Now to try a new micro SD card with the working backup image written to it.

[Gina]

OK - New RPi, new micro SD card with known working software, new stepper driver, HAT that used to work, stepper motor that used to work, 5.1v Raspberry PSU, 13.8v bench PSU for motor supply.  Now to test...

[Gina]

No joy

The software seems to be working as the green light on the RPi flashes as the steps occur.

[Gina]

If my fingers worked better I'd take up knitting again!!!

[Gina]

Shut down, removed stepper driver and been checking continuity of all the connections with DMM - no problem found.  Pure magickkkkkkk   Black magic at that!!!

[Gina]

Maybe tomorrow I'll do some serious faultfinding.  Clear a nice big space on the table, use a laptop to access the RPi so that I have everything to hand and connect oscilloscope to each data line in turn, That has to show where the problem is surely!  I need to sort this out before I go completely mad!!!

[NEM]

A wild guess here, from someone who's never even used an A4988, so it's probably not the answer, but...  Are you sure there's a negative connection from the Pi to the driver chip ground?  What if, when the DIR pin is at 0V it's providing the ground to the driver through internal leakage in the A4988, so the circuit works?  Then, when the DIR line goes +ve there's no ground and the circuit stops.

Noel

[Gina]

Thanks for you suggestion Noel   But yes, I've checked that, both as ohms and when working, the voltage on the Gnd pin of the driver.

[Gina]

Thought I'd just check the stepper motor for coil resistance (including continuity and shorts) fine - 55ohms.  But while checking that I've found a mistake on the HAT, though why that didn't stop it working before beats me (it even worked for auto-focus from KStars/Ekos).  One of the motor outputs is connected to Gnd   One little link!!  Of course, that doesn't explain why the other circuit (for rotation rig) doesn't work but I suspect that is something else.

[Gina]

Phew!!!!  It's working   Rotates both ways.  So I'm not completely mad - just a little stupid!  Now to the board for the rotation rig...