Stepper motor mount for Losmandy GM-8


John Scherer
 

Anyone have a stepper motor mount design for the Losmandy GM8 they would be willing to share?  I've looked around and I see there are a few using the GM8, but no real detail on the motor mount.  I'd like to add it to the File section.  I've got my design mostly done, a 2:1 belt drive but I need to add an idler pulley for belt tension.  I've not actually built mine yet so I'm uncertain if the motors are clear of any collisions with other areas of the mount or ota.


Khalid Baheyeldin
 

On Wed, Dec 18, 2019 at 03:27 PM, John Scherer wrote:
Anyone have a stepper motor mount design for the Losmandy GM8 they would be willing to share?  I've looked around and I see there are a few using the GM8, but no real detail on the motor mount.  I'd like to add it to the File section.  I've got my design mostly done, a 2:1 belt drive but I need to add an idler pulley for belt tension.  I've not actually built mine yet so I'm uncertain if the motors are clear of any collisions with other areas of the mount or ota.
In case you have not seen it, in the show case page, there is a GM8 by Klaus Steinen. There are links to pictures there.


John Scherer
 

Thanks Khalid.  I did see that.  I may take the time and model his design in Fusion, though I'm a bit confused by his setup.  He mentions that on his GM8, the gearbox on the motors are 12:1, with 3.8 deg/step motors?  Seems like the steps/deg on that would not be enough for astrophotography.  I came up with 691 deg/step with that design?  Is that correct?


John Scherer
 

sorry, not 691, but 19200 steps/deg with 32 microsteps, NOT 691.  Spreadsheets don't work well if you can't enter numbers correctly ;-).  But even so, at 19200, that still seems a bit low astrophotography, is it not?  Above I mentions my motor mount us belt driven at 2:1, that was a typo.  Its actually 4:1. 


Khalid Baheyeldin
 

On Wed, Dec 18, 2019 at 07:59 PM, John Scherer wrote:
But even so, at 19200, that still seems a bit low astrophotography, is it not? 
19,200 steps per degree is above the minimum threshold of 12,800 which translates to 0.28 arc second per step.
0.28 arc seconds is between the 0.15 and 1.25 specified as limits in the spreadsheet.
So it is adequate, specially if it is for a short focal length optical tube.


Howard Dutton
 
Edited

On Wed, Dec 18, 2019 at 05:39 PM, Khalid Baheyeldin wrote:
19,200 steps per degree is above the minimum threshold of 12,800 which translates to 0.28 arc second per step.
0.28 arc seconds is between the 0.15 and 1.25 specified as limits in the spreadsheet.
So it is adequate, specially if it is for a short focal length optical tube.

That 0.28" is the calculated step angle the estimated resolution is 0.58".

~1.2" res - Some of the Instein OnStep kit users run the GM8 at 180:1 driven by a 400 step motor so evidently that works though I wouldn't recommend it except perhaps for visual only use.

~1.0" res - The original Losmandy Digital Drive on GM8's.

~0.6" res - My G11 w/400 step motors (32x microstep mode) into 360:1 lots of folks use this with good results.  Maybe there is room for improvement maybe not.

~0.5" res - The original Losmandy "Digital Drive" on G11's.

~0.3" res - Your proposed GM8 w/timing belt 4:1 driving the 180:1 worm/wheels would come in at 25600 steps per degree.


~0.2" res - This brings me to another project I'm working on... over the next few months I will swap out the bipolar 400 step motors on my G11 with 5-phase 1000 step motors.  Will they really track at 2.5x higher resolution?  According to the datasheets slew speeds should be much higher than with my 0.9A bipolar motors; a couple of times higher?  I'd guess these would work a treat on the GM8 too since resolution is so high and torque is up there too and the usual G11 mounting plate/etc would be all that's needed from what I understand.  Not the cheapest option though...  ;)

I'm trying these (the oriental drivers were just too much $$$):
https://www.ebay.com/itm/Autonics-MD5-HD14-5-Phase-Micro-Stepper-Motor-Driver-1-MHz-20-35-VDC-3A/182328977502?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649

With these:
https://catalog.orientalmotor.com/item/pk-series-42mm-5-phase-stepper-motors/pk-series-42mm-5-phase-stepper-motors/pkp546mn18a?__hstc=62166835.b0060da8a1a769f933a4748aebb7f157.1576771487602.1576771487602.1576771487602.1&__hssc=62166835.9.1576771487602&__hsfp=122382122


Khalid Baheyeldin
 

On Thu, Dec 19, 2019 at 01:00 PM, Howard Dutton wrote:
On Wed, Dec 18, 2019 at 05:39 PM, Khalid Baheyeldin wrote:
19,200 steps per degree is above the minimum threshold of 12,800 which translates to 0.28 arc second per step.
0.28 arc seconds is between the 0.15 and 1.25 specified as limits in the spreadsheet.
So it is adequate, specially if it is for a short focal length optical tube.

That 0.28" is the calculated step angle the estimated resolution is 0.58".

~1.2" res - Some of the Instein OnStep kit users run the GM8 at 180:1 driven by a 400 step motor so evidently that works though I wouldn't recommend it except perhaps for visual only use.

~1.0" res - The original Losmandy Digital Drive on GM8's.

~0.6" res - My G11 w/400 step motors (32x microstep mode) into 360:1 lots of folks use this with good results.  Maybe there is room for improvement maybe not.

~0.5" res - The original Losmandy "Digital Drive" on G11's.

~0.3" res - Your proposed GM8 w/timing belt 4:1 driving the 180:1 worm/wheels would come in at 25600 steps per degree.
Add to the above this fact:

Atmospheric turbulence (seeing) limits resolution to 0.4" - 1".
The former being the best there is on earth, i.e. Hawaii, Atacama, Canaries ...

So, 0.3" should be more than adequate.

~0.2" res - This brings me to another project I'm working on... over the next few months I will swap out the bipolar 400 step motors on my G11 with 5-phase 1000 step motors.  Will they really track at 2.5x higher resolution?  According to the datasheets slew speeds should be much higher than with my 0.9A bipolar motors; a couple of times higher?  I'd guess these would work a treat on the GM8 too since resolution is so high and torque is up there too and the usual G11 mounting plate/etc would be all that's needed from what I understand.  Not the cheapest option though...  ;)

I'm trying these (the oriental drivers were just too much $$$):
Very interesting. I have seen these 1000 step 5-phase motors on eBay, used. Seems that Oriental Motors are the only one that makes them (for now). But the lack of a suitable driver was an obstacle.
Good to offer support for them, for those who need the extra resolution.


Howard Dutton
 

On Thu, Dec 19, 2019 at 11:20 AM, Khalid Baheyeldin wrote:
Very interesting. I have seen these 1000 step 5-phase motors on eBay, used. Seems that Oriental Motors are the only one that makes them (for now). But the lack of a suitable driver was an obstacle.
Good to offer support for them, for those who need the extra resolution.
I nervously await testing to see if this is worth the effort/expense, experimentation like this is as likely to fail (i.e. similar performance) as succeed.


Khalid Baheyeldin
 

On Thu, Dec 19, 2019 at 02:36 PM, Howard Dutton wrote:
On Thu, Dec 19, 2019 at 11:20 AM, Khalid Baheyeldin wrote:
Very interesting. I have seen these 1000 step 5-phase motors on eBay, used. Seems that Oriental Motors are the only one that makes them (for now). But the lack of a suitable driver was an obstacle.
Good to offer support for them, for those who need the extra resolution.
I nervously await testing to see if this is worth the effort/expense, experimentation like this is as likely to fail (i.e. similar performance) as succeed.
The good news is that it is a STEP/DIR driver, and will be configured as an external driver (i.e. microstepping is in the driver, not in OnStep), so that part is taken care of.

However testing is two faceted:
- Testing that it works, which is easy.
- Testing whether the resolution is better, that is much harder.


Howard Dutton
 

On Thu, Dec 19, 2019 at 11:48 AM, Khalid Baheyeldin wrote:
- Testing whether the resolution is better, that is much harder.
I think the thing to do is a test similar to the hack-a-day one but with more motors, more drivers, and torque loading similar to what moves my G11 around.  To that end I also ordered a 116 oz-in holding torque NEMA17 400 step motor as I feel it might turn out to be the best performer.  I could use a laser diode spot across the room/ruler to take angular measurements but probably will use my 200k count encoder instead since it'll be quicker and more accurate probably.  I can hack the WiFi addon to gather the measurements.


John Scherer
 

On Thu, Dec 19, 2019 at 01:28 PM, Howard Dutton wrote:
To that end I also ordered a 116 oz-in holding torque NEMA17 400 step motor
That's a mighty big nema 17!  Who makes that?  Linn Engineering?


Howard Dutton
 

On Sat, Dec 21, 2019 at 09:51 PM, John Scherer wrote:
That's a mighty big nema 17!  Who makes that?  Linn Engineering?
Oriental:
https://catalog.orientalmotor.com/item/2-phase-bipolar-stepper-motors/42mm-pkp-series-2-phase-bipolar-stepper-motors/pkp246md15a2


Khalid Baheyeldin
 

On Thu, Dec 19, 2019 at 04:28 PM, Howard Dutton wrote:
On Thu, Dec 19, 2019 at 11:48 AM, Khalid Baheyeldin wrote:
- Testing whether the resolution is better, that is much harder.
I think the thing to do is a test similar to the hack-a-day one but with more motors, more drivers, and torque loading similar to what moves my G11 around.  To that end I also ordered a 116 oz-in holding torque NEMA17 400 step motor as I feel it might turn out to be the best performer.  I could use a laser diode spot across the room/ruler to take angular measurements but probably will use my 200k count encoder instead since it'll be quicker and more accurate probably.  I can hack the WiFi addon to gather the measurements.
An advantage of those 5-phase motors is that they could be connected with just 1:1 pulleys for EQ4 (144 teeth at 1/64 microstep).
That can be helpful in cramped situations where a bigger pulley is obstructed (like my mount).

The disadvantage is that the external drivers are bulky and unlikely to go into the box.
And the cost of it all is significantly higher than bipolar options.
Another disadvantage is the lack of on the fly mode switching.

Maybe size and cost will go down in the coming years.


Alexander Varakin
 

I used 400 steps NEMA 23 for my GM8 with direct drive using spring coupler.
Here is the motor:



John Scherer
 

Did a 3D print of my 4:1 stepper mount for the GM8 mount.  It’s printed in PETG and came out pretty good. 


Howard Dutton
 

Nicely done John.


Chris Vaughan
 

What material are you using for the 3D print John? And do you reckon it'll be okay with a warm stepper in the mix. 


John Scherer
 

Hey Chris, it’s PETG and prints @ 240-260c and has great thermal properties and should be fine with the heat from
the stepper.  If your stepper ever got hot enough to soften PETG, you’d have other problems ;-)


John Scherer
 

I should add, my plan is to machine both the plate and the cover from aluminum.   3D printing just allows for rapid prototyping and iterations in the design.   


John Scherer
 

I'm mostly done with this 4:1 belt drive for the Losmandy GM8 and you can look at the design here.  I'm just waiting on pulleys and belt coming from china and I can test this, though I have no reason to believe it won't work. If anyone wants STL's for printing let me know.