Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Harmonic drive EQ mount project

  • Please log in to reply
16 replies to this topic

#1 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 16 October 2019 - 09:53 AM

Finally - I'm FREEEE.  Well sort of.  Last December I got two HD component sets for under $200.  They're used from an industrial robot, and look to be in excellent condition.  One set also had a serious crossed roller bearing on the output.  I bought another bearing to match for the second axis.  It took me 10 months to grind through the design as I picked away here and there when I could.  I want to do this project really well including finishes and all, so not to be hurrying.  I now have some 40+ sheets of detail drawings and a wee bit of machining to do - no CNC - just a manual lathe and mill.  The plan is to run it all with OnStep which I tried briefly in 2017 - those folks have really done impressive things with that platform!  I designed in axis release clamps outside of the HD drive trains to mimic having clutches like a typical worm drive might. The whole mess should weigh about 30 pounds (without tripod or counterweights).  Should easily carry 50 pounds or more depending on the placement of the load and acceleration rates - more if i use a counterweight.  HD's have pretty big periodic error so guiding will be needed for long focal length / duration imaging.

 

I think I'll super clean the shop before making chips - lots of chips.

 

Cheers,

 

Bob

Attached Thumbnails

  • Harmonic Drive Mount FINAL 4b.jpg
  • Harmonic Drive Mount FINAL 3b.jpg
  • Harmonic Drive Mount FINAL 2b.jpg

  • PrestonE, gregj888, kingjamez and 11 others like this

#2 macdonjh

macdonjh

    Fly Me to the Moon

  • *****
  • Posts: 6,022
  • Joined: 06 Mar 2006

Posted 16 October 2019 - 10:45 AM

I'll take one.

 

No, wait, I just bought a mount.  Oops.  Bad luck for me.



#3 theropod

theropod

    Sputnik

  • -----
  • Posts: 46
  • Joined: 01 Jan 2019

Posted 21 October 2019 - 09:16 AM

So, belt drives in the finals? My phone just won’t really let me see detail in the drawings. Yes, I’ll be watching this thread.



#4 Aleksandr Naumov

Aleksandr Naumov

    Mariner 2

  • *****
  • Posts: 207
  • Joined: 23 Jun 2014

Posted 21 October 2019 - 09:25 AM

Bob,

 

I would like to ask you:

 

- what is the reduction of HD ?

  (I see stepper motors on your drawings, think there is no internal gear drive)

 

- could you estimate backlash on these HD ?

 

- how do you estimate "pretty big periodic error" ?



#5 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 22 October 2019 - 09:40 AM

Hi Macdonjh - thanks.  Hopefully I'll end up with one!

 

Hi Theropod - the belt drives are on the input side of the harmonic drives:  90T and 17T for a 5.294: reduction from the stepper motors.

 

Hi Aleksandr - The HD sets are listed as 100:1.  Because of the mounting arrangement the RA axis may actually be 101:1 * 5.294 to give 534.7:1 overall.  The DEC axis will be 100:1 * 5.294 to give 529.4 overall.

 

The output backlash for the HD drives is zero.  However there is a small backlash on the input side due to the Oldham coupling used in the wave generator assembly. The units I have are spec'd at 9 arc-sec max backlash on the input.  So that equates to .09 arc-sec backlash between the motor and the output if you assume zero backlash in the belt drives.  HD drives are often described as "zero backlash", but there is that tiny bit that will depend on the ratio of the drive.
 

Regarding the periodic error:  I first became aware of this reading a thread in the mounts forum about the RainbowAstro RST-135.  Lots of good discussion in this thread.  My sets are either 30 arc-sec peak to peak if standard, or 15 arc-sec peak to peak if special.  I'd sure like to feel special, but don't really know what I have in hand.  This info all comes out of the HD component sets catalog.

 

I'll update as I successfully make some parts.

 

Best Regards,  Bob


  • Aleksandr Naumov likes this

#6 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 10 February 2020 - 11:13 AM

Well in spite of life's demands finally a bit of progress.  Finished up the machining of the core box structure and did a test fit of the parts.  Looks very solid and square - so far so good.  Lots of finishing work to do before anodizing  - that's a long way off.  Here's some pics with the HD parts stuffed into place on the RA axis.  Once I do the side covers I'll start on parts that need lathe/mill/rotary table work.

 

Best Regards,

Bob

 

Attached Thumbnails

  • IMG_20191228_135320.jpg
  • IMG_20200208_134403.jpg
  • IMG_20200208_134424.jpg
  • IMG_20200208_134337.jpg

  • PrestonE, brave_ulysses, don clement and 5 others like this

#7 theropod

theropod

    Sputnik

  • -----
  • Posts: 46
  • Joined: 01 Jan 2019

Posted 10 February 2020 - 08:34 PM

I am amazed at the work done here. Stunning.


  • PrestonE and gregj888 like this

#8 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 01 September 2020 - 10:56 AM

Thanks Theropod,  sorry for not replying before.

 

I finally finished up the lathe work on all these parts that form the basic drive trains internal to the box, as well as the output shafts.  Still need to buy some gearbelt pulleys to modify, but I'll wait until i see all this go together.  Now all these guys need to go to the mill for keyways, radial holes patterns, slotting, and such. 

 

Thanks for looking.

 

Best regards,

 

Bob

Attached Thumbnails

  • Lathe Parts.jpg

  • PrestonE, gregj888, R Botero and 3 others like this

#9 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 17 September 2020 - 10:48 AM

Howdy,  finished up all the mill / rotary table work on the above parts.  The right hand group is the RA axis, and the left hand group are declination parts.  Also in the mix are the spline cups and wave generators to start showing how these guys go together.  The top four parts on the left are the stainless DEC output shaft and pieces of the coaxial DEC axis release - sort of a bearing supported shaft clamp.  Not really a clutch at all.

 

Cheers,

Bob

Attached Thumbnails

  • Drive train parts machined.jpg

  • PrestonE, brave_ulysses, dmcnally and 2 others like this

#10 mic1970

mic1970

    Apollo

  • -----
  • Posts: 1,225
  • Joined: 09 Aug 2018
  • Loc: St. Louis Metro Area

Posted 17 September 2020 - 10:54 AM

popcorn.gif popcorn.gif popcorn.gif popcorn.gif



#11 PrestonE

PrestonE

    Vanguard

  • *****
  • Posts: 2,302
  • Joined: 29 Apr 2005
  • Loc: San Miguel de Allende,Mexico

Posted 17 September 2020 - 01:44 PM

Hola Bob,  It is looking really nice.

 

One though.

 

On your plates that cover the box frame that is holding the HD's

I would recommend putting a couple of alignment pins in each

side to help take up the load and keep things aligned when it is 

under load...

 

as, just bolts holding things together are not really an accurate 

way to locate precision parts.

 

Drill under size and ream to fit...

 

Looking forward to your finish...

 

Best Regards,

 

Preston



#12 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 17 September 2020 - 02:26 PM

Hi Preston,  I wholeheartedly agree!  There's a few places that need some doweling before it's all over.  Just tried to test assemble the DEC axis - it's tricky to get the wave generator inserted correctly into the flex spline cup.  I need to slow down a bit and do this carefully.

 

Best Regards,

 

Bob


  • PrestonE likes this

#13 LarsMalmgren

LarsMalmgren

    Messenger

  • *****
  • Posts: 407
  • Joined: 22 Jan 2010
  • Loc: Denmark

Posted 17 September 2020 - 04:21 PM

Did you make the harmonic drive parts or are they bought ?

 

If you did make them, what material is used for the spline cups?



#14 PrestonE

PrestonE

    Vanguard

  • *****
  • Posts: 2,302
  • Joined: 29 Apr 2005
  • Loc: San Miguel de Allende,Mexico

Posted 17 September 2020 - 04:47 PM

Lars, he says in the first post that he purchased them for under $200...

 

Best Regards,

 

Preston



#15 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 17 September 2020 - 06:25 PM

Hi Lars,  yeah making functional harmonic drive components is way far beyond any skill or capabilities I could hope to have.  I've seen some mention of a 3d printed drive.  Such a thing wouldn't be very accurate or have much capacity.  I don't know the metallurgy of the spline cups, but I'll bet it's quite specific.

 

Best regards,

 

Bob


  • PrestonE likes this

#16 LarsMalmgren

LarsMalmgren

    Messenger

  • *****
  • Posts: 407
  • Joined: 22 Jan 2010
  • Loc: Denmark

Posted 18 September 2020 - 01:16 PM

Thanks both,

 

somehow i missed that you bought the HDs :)

I've seen several 3D printed on youtube, pretty fun to see how the spline cup flexes.

That made me speculate on what type of metal alloy that is used.

 

Looking forward to see you finish the mount waytogo.gif



#17 lukasik

lukasik

    Vanguard

  • *****
  • topic starter
  • Posts: 2,383
  • Joined: 11 Oct 2005
  • Loc: Dacula, Georgia

Posted 30 September 2020 - 11:07 AM

Hi there,  a bit more progress.  I've finally seen it all fit together (not any of the base parts - just the main drive box).  It will all have to come apart again numerous times to add dowels locations, tweak fit ups, finishing, etc.  But it's encouraging that it all just works as planned so far.  Here's a couple more pics, and a link to the first test with motor power.  The stepper is howling horribly - no optimal motor control here.  Just quick and dirty 1/8 stepping with a Pololu TIC 500 to watch it run.  Even with that said it has gobs of output torque.

 

https://youtu.be/O7NZB3SvN-I

 

Best Regards,

Bob

Attached Thumbnails

  • First rough assembly.jpg
  • Two Drive Trains assembled.jpg

  • PrestonE, R Botero, Yanouk and 3 others like this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics