Losmandy G11 OnStep Conversion


GuitsBoy
 

I just finished converting my new (to me) Losmandy G11 to OnStep in just under 3 days.  This is my fourth OnStep conversion, and so far the easiest.   Cross posted to CN.  Posting here too just in case its useful to anyone.

Tuesday morning, I picked up the G11 locally.  I knew the mount was missing the gemini controller, no big deal since I planned to switch to onstep anyway.  What I didnt realize was that this supposedly "tested" and "great condition" mount had a completely seized up and frozen DEC spindle.  Even fully removing the clutches, I could not get the DEC to turn.  I had to wedge a long piece of scrap lumber in the saddle plate to get it to break free.  I spent a few hours with solvents and a rubber mallet to finally free the spindle.  I let the DEC housing soak overnight, and after a few more hours worth of work, finally freed up the gummed bearings.  I re-greased with superlube and reassembled the mount, and everything seems fine.  The RA side seems to be in great condition, only the DEC side was severely neglected.  Not sure why the disparity between the two.  But regardless, it moves nice and smooth now.

 
onstep_g11_worm_bearings.jpg
After trying to pull the stock bearings out for 45 minutes, I decided to pop them in the toaster for 5 minutes at 325.  The bearings dropped right out, and new ones slipped right in place.

onstep_g11_worm.jpg
I ran into so much trouble trying to turn down the flexible helical shaft couplers I bought for the project.  I dont have a lathe, and couldn't quite do the job on a drill press with a file.  I decided to go with some rigid couplers I had on hand, and they did the trick.  I modified some NEMA17 stepper motor brackets I found on thingiverse, and 3D printed them up.  I drilled the holes just a touch oversized to let the shaft and coupler do the centering, then snugged into place.  It spins smooth, and I dont think there's any runout or shaft play.

onstep_g11_RA_ext.jpg
I needed to add a RA extension, but didn't want to shell out 300 bucks for the Losmandy one.  I found a five pack of 1/4" ID, 1/2" OD x 1-1/4" long aluminum standoffs on amazon for $6.  I 3D printed a basic cylinder to work as a housing (126mm OD, 122mm ID, 31.75mm height) and it fits perfectly over the standoffs.  I dont hate the look, honestly.

onstep_g11_ma.jpg
I purchased the MA adapter to mount up to my old Meade tripod.  The holes didnt line up, so I had to drill and tap a few M6 cap screws to keep it in place.   I printed up that plastic thumb nut to snug the tripod spreader bolt, but I since decided against it, and put a metal nyloc nut in place.  Better to be overbuilt than not strong enough.

onstep_g11_RA_motor.jpg
Here's  the RA motor and the printed bracket.  I punched down he stepper motor cables into an RJ45 (ethernet) jack, then modified and printed a small keystone box I found on thingiverse.  Without getting too fancy, a couple of zip ties hold things together.  Not exactly elegant, but it is quite functional.

onstep_g11_dec_motor.jpg
Here's the DEC motor and bracket.  Again the same RJ45 jack for quick disconnects.   Up top, you can see I had to make an adapter to install a vixen dovetail clamp, since the mount came with a Losmandy D style only clamp.  Eventually, Ill pony up the $130 to get the proper dual clamp from Losmandy.  Its a gorgeous piece of machined art, and would really help the aesthetics.

onstep_g11_controls.jpg
I found these aluminum project boxes on amazon for $4.50 each.  I picked up ten or so for various projects, OnStep being one of them.  Up top, the 12v input and USB connector.  On the bottom are four RJ45s, one each for the RA motor, DEC motor, Focuser motor and the Temperature Sensor.

onstep_g11_back_profile.jpg
Here you can see the motor outputs on the bottom of the OnStep enclosure.  The controls are temporarily mounted to a piece of scrap wood until I can figure out something more permanent.  You can also see the anemic saddle plate adapter.  Its pretty snug, but I wish I had some wider mill stock on hand.  This will do for now. 

onstep_g11_side_profile.jpg
Here's the side profile of the completed mount.  The RA extension looks right at home, and you can see there's plenty of room to clear the motor now.  You can also see my hop trellis that I've been putting off stringing up the hops so I could tackle the G11 first.  What was I thinking?  You can't do proper astrophotography without a homebrewed IPA, can you?

As for the nitty gritty OnStep details:

I'm currently running OnStep 4.23.r on a Wemos R1 ESP32 board and CNC Shield V3.  The stepper drivers are LV8729's pushing around .7 amp to NEMA17 0.9º (400 step) motors good for up to 2.0A.  The focuser control has an A4988 driver.  Each of my three main telescopes has a DIY focuser motor with RJ45 jack to plug into the mount controls.  Lastly, I have a DS18b20 temperature sensor that plugs into the fourth RJ45 jack to handle temperature compensated focusing.

The geartrain is pretty simple on the G11.  For  the first iteration, I've opted for a direct driven worm without any gear reduction at the motor.  I may eventually go belt drives, but the brackets and design will be far more complicated.  I don't really know if a little additional mechanical gearing is worth the added complexity.  The 1:360 worm gearing is already pretty robust.  For comparison my heavy meade LXD650 had 1:180 worm gearing, but 1:2.25 belt drive reduction, so the final reduction was quite similar.  I never felt the meade was underpowered with the same motors.  I'm running the RA and DEC steppers at 1/64 microsteps, which gives me a final resolution of 0.14 arc-sec per step, and a max slew rate of 2.4º per second.  Not exactly speedy, but it doesn't feel terribly slow either.

My imaging gear:
Meade 127 ED/APO refractor
Explore Scientific 208mm f/3.9 Newt with ES HR Coma Corrector
AstroTech AT72EDii with .8x flattener/reducer and 1.0x flattener
SVbony 60mm guidescope / Meade LPI-G guidecam
Canon 77D unmodified, and a bunch of 2" filters.
I just received my off axis guider, but I'm not gonna play with that until I've tested the new mount with the familiar guidescope.

Thanks again Howard and all the OnStep contributors!  What a fantastic project!  

Clear skies to all,
-Tony




Khalid Baheyeldin
 

Thanks for sharing this.
It is in the showcase page now.


George Cushing
 

If you ever want to try the spiral couplers I can turn you some.

But first you have to get some 1/4" stainless nylocs for the tripod😀


GuitsBoy
 

Thank you for the offer, but my dad has a lathe, and amazon just delivered a bag of helical couplers to him yesterday.  By chance, do you know why the flexible, bellows or oldham couplers are so strongly recommended around here?  Is it simply for accommodating defects in the brackets, wrong angles, off center holes, etc?  Is it for runout in the stepper motor shaft?  Is it to save wear and tear on the worm shaft bearings?  Is it simply to smooth the steps of the motors?  Ive been wondering why this whole time, and even more so now that I don't see any obvious issue with the rigid coupling.

As for the tripod, did you mean because the acorn nuts are rusted on the legs?  Or because I used a taboo metric m6 fastener to attach the Losmandy MA?  The tripod is 30+ years old, and is just a temporary solution while I decide on the most functional location for a permanent pier.  The tripod lived a long and good life.  Soon it gets to retire and live out the rest of its days comfortably tucked away in the garage or basement .

Thanks,
-Tony


On 5/1/2021 2:22 PM, George Cushing wrote:
If you ever want to try the spiral couplers I can turn you some.

But first you have to get some 1/4" stainless nylocs for the tripod😀


adraasch
 

Any lateral, or angular, misalignment between the motor shaft and the worm as well as run out on both the motor shaft and worm is the reason for flexible motor couplings.  Bearing life will suffer if one has misalignment(s) and uses fixed couplers.

It also could lead to motor stalls as the motor would be working harder as it was rotating through the area that binds.

Arlen


On Sat, May 1, 2021, 3:50 PM GuitsBoy <guitsboy@...> wrote:

Thank you for the offer, but my dad has a lathe, and amazon just delivered a bag of helical couplers to him yesterday.  By chance, do you know why the flexible, bellows or oldham couplers are so strongly recommended around here?  Is it simply for accommodating defects in the brackets, wrong angles, off center holes, etc?  Is it for runout in the stepper motor shaft?  Is it to save wear and tear on the worm shaft bearings?  Is it simply to smooth the steps of the motors?  Ive been wondering why this whole time, and even more so now that I don't see any obvious issue with the rigid coupling.

As for the tripod, did you mean because the acorn nuts are rusted on the legs?  Or because I used a taboo metric m6 fastener to attach the Losmandy MA?  The tripod is 30+ years old, and is just a temporary solution while I decide on the most functional location for a permanent pier.  The tripod lived a long and good life.  Soon it gets to retire and live out the rest of its days comfortably tucked away in the garage or basement .

Thanks,
-Tony


On 5/1/2021 2:22 PM, George Cushing wrote:
If you ever want to try the spiral couplers I can turn you some.

But first you have to get some 1/4" stainless nylocs for the tripod😀


GuitsBoy
 

Well, the first shakedown run shook out a few bugs, but worked through them.  Anyone else ever run into an issue where PHD guide pulses seem to try to guide at full slew rates?  Even if you manually guide at .5x, or even set the guide rate manually with :R1#?  I had to restart everything and all was well.  Ive run into this a few times on other setups as well, and not sure how to reproduce it or avoid it.

Anyway, I was shooting m63 and m101, both at pretty close to zenith, and well north of the celestial equator, so perhaps not a very challenging guide target.  Still, I was pretty happy with a VERY consistent .75 to .85 arc-sec RMS for a first try, especially considering I was shooting through thin cloud cover the entire night. 

Full size image

There's some very obvious periodic error on the RA worm.  Lets throw the log through PEC prep to find  the period, shall we?

The sharp spike is 119 seconds, or half the worm period.  A quick search online shows this is most likely slipping of the oldham coupler.  I dont have oldham coupler, they are rigid, but they could easily have been tightened down out of alignment, or even slipping as  the worm rotates.  But other than the 119 second period, nothing stands out at me as an issue.  I'm a bit relieved that I dont see the dreaded 75 second spike.

Arlen, as you mentioned, this looks to be exactly the reasons  you mentioned NOT to use rigid couplers.  Thank you for the info, by the way.

Anyway, I will attempt to re-align everything on the RA drivetrain to try and center it better with the rigid coupler.  In two weeks I hope to see my parents again for the first time since covid, and if I beg and plead enough, maybe my dad will bring a couple of those turned down helical couplers with him when they arrive.

But all in all, seems like a good start for this mount.  The guide error was incredibly regular compared to my last mount which was totally random, so nothing to tune out.  Considering the clouds and he fact that I didnt bother setting up a wind break (< 5 mph), I think this thing shows a lot of promise.

Any thoughts or suggestions?

Thanks all,
-Tony



George Cushing
 

Yeah, just teasing you about the tripod. Every time I think I don't need a LX tripod I find another use for one. I was down to none and then the local observatory lost its lease and for want of storage space re-donated a couple of LX3s to me. Both were in beautiful shape (except for the famous silvered secondary) including the tripods, so I'm good again. Now I'm shipping LX3 parts as far as Bangkok.

I use old Meade DS/ETX worms to make a 1/4"/5mm mandrel to turn the couplers. Tell your dad to get some 4mm long set screws before starting. You can't get to 0.61" w/o the tooling hitting the stock 6-8mm long screws and they are very hard. IIRC they are 4mm threads. 

Howard used them to deal with any minor error in centering the worm and motor shafts or runout in either. With 28,600 µsteps per degree the frequency for the pulses (106.77Hz) is too high to have much impact unless some resonance sets in. Then try another stepping rate.


GuitsBoy
 

Thanks for the advice.   I got tired of waiting.  Two weeks is a long time....

It's ain't pretty, but it don't need to be.  Got the job done, taking a couple thousandths off at a time with the cross slide vice.  I removed the set screws during machining, but I could see how a mandrel could come in handy.  I just used a snugged long m4 cap screw.  I did have smaller 4mm m4 set screws which I could have used to keep things steady, but I didn't see your reply until afterwards.

Anyway, its in there now.  Fits great.  Doesn't seem to spin any smoother, but I guess its not something you'll feel by hand.

I just did the one on the RA side.  I'll wait for my dad's version, and swap to his on both sides when I see him.  Hopefully this will get me by for a couple weeks.

Thanks again for your help and suggestions,
-Tony




On 5/2/2021 11:06 AM, George Cushing wrote:
Yeah, just teasing you about the tripod. Every time I think I don't need a LX tripod I find another use for one. I was down to none and then the local observatory lost its lease and for want of storage space re-donated a couple of LX3s to me. Both were in beautiful shape (except for the famous silvered secondary) including the tripods, so I'm good again. Now I'm shipping LX3 parts as far as Bangkok.

I use old Meade DS/ETX worms to make a 1/4"/5mm mandrel to turn the couplers. Tell your dad to get some 4mm long set screws before starting. You can't get to 0.61" w/o the tooling hitting the stock 6-8mm long screws and they are very hard. IIRC they are 4mm threads. 

Howard used them to deal with any minor error in centering the worm and motor shafts or runout in either. With 28,600 µsteps per degree the frequency for the pulses (106.77Hz) is too high to have much impact unless some resonance sets in. Then try another stepping rate.