Turning a worm?

[NoelC]

Anyone had any experience at trying to make a precision worm drive?

My old HEQ5 Pro has (what I consider to be) unacceptable periodic error (+/- 15 arcseconds), so I spent quite some time researching and learning to turn up a worm as that was obviously the component at fault with the idea I could improve it for a fraction of the cost charged by Dark Frame Optics. If successful I could then upgrade the worm on my EQ6-R Pro.  However my various attempts have turned out far worse than the original and I'm starting to think it cannot be done.  I measured the eccentricity on the original worm outer shoulders at up to 60 microns (that's a lot) - foolishly thinking I could do better I manufactured one (after 3 attempts) with 14 microns eccentricity but it's producing a PE of +/- 140 arcseconds.

Initial rough measure up of worm using vernier gauge.

Having looked at this worm again, can now say it was turned between centers (hence the flats on each end), probably roughed out first then mounted and the shafts turned down (how would you do that between centers?) and the thread cut very precisely.  There are a few faults with the design; it does not stand the bearing off the shoulders sufficiently; so that under tension the bearing outer could rub on the worm, the flats are an obvious problem introducing eccentricity onto the gear of belt pulley mounted on the end, and the dimensions are unnecessarily sloppy (8.95 shafts should be closer to 8.99).

Doing some basic trigonometry, the wormwheel on an HEQ5 Pro is 70mm dia (PCD approx 68.8mm).  Teeth are around 1.1mm deep, with 20 degree pressure angle (40° teeth) set on pitch of 1.6mm (16tpi) I'm coming up with calculated figures of 59" angular motion per micron radial motion - does that sound reasonable?  Which gives (according to my flaky trig) a PE of around +/- 150 for an eccentricity of +/- 7 microns on 20° teeth.
If that's correct I will need to work to tolerances of less than 1/10 micron (which I can't even measure).

I'm diamond lapping and all of that old cobblers, but I'm starting to wonder if this is impossible.  I note that most ATM'rs tend to use much larger wormwheels for RA drives and don't publish their PE.  I guessed the manufacturers at Skywatcher new their business, but an initial quick view of their products lead me to believe the worms weren't that precise.

[Carole]

I know some-one who does, if you don;t get any joy on here, E mail me and I'll see if I can put you in touch.

Carole

[Roberto]

Hi Noel

I've only screws on my lathe but never a worm gear.  Are you sure the thread pitch is in TPI? This is all Chinese, would expect it to be some other - metric? - pitch.  Maybe that's part of the issue?  I wonder how many teeth the worm wheel has.

Roberto

[NoelC]

Thanks Roberto.
Like you I assumed it would be metric, but the pitch was awfully odd (certainly not a standard metric thread)- so I ordered a thread gauge and it says it's 16TPI (suspect a throwback from American influences/copies).  Pressure angle is 20° as close as I can measure it.  Teeth on gear have been turned with a 3/4" hob - aha I thought - 16TPI X 3/4 - but no, the pressure angle on a 16TPI  thread is 30° - so no cheap and easy way to hob a gear wheel.   But since you mentioned it; I will strip it again and re-check the thread to make sure it's not that simple.  The HEQ5 Primary gear is listed as having 135 teeth on the EQMOD website (http://eq-mod.sourceforge.net/reqindex.html) - I've not counted them but that would tally with the PCD I observed (68.8mm).

Rechecked my maths on the PE (dropped a decimal):
For a peak to peak eccentricity of 14 microns, I now calculate the radial displacement at 5 microns, which equates to an angular displacement of 31 arcseconds (i.e. pretty much what I'm seeing with the Skywatcher worm, and nowhere near the 248 I'm measuring).
So there IS something wrong with the worm other than PE, suspect it's the tooth profile.  Carole's contact (Mark) recons it can be corrected with lapping the worm into the wheel; I'm a little reticent to do that as it offers no way back if I'm wrong...
I'm tempted to knock up a worm and wheel at 30° pressure angle (as it avoids making a custom hob)  and lap it to within an inch of it's life.  But at £66 a shot for the brass and goodness knows how much for the tooling to set up for gashing accurately, maybe I'll just stick the mount back together and flog it to put the proceeds towards a better mount.

Obviously not as easy to eliminate PE as I had hoped...

[Roberto]

Strange angle between teeth.  Could it be a Whitworth at 55 degrees which pressure at 27.5 instead?  I know old scopes (1800s, early 1900s) had worms with Whitworth threads but would be surprised if a Chinese one had same!  Whitworth would certainly had TPI pitches.

Lapping is an option but as you say - no way back.  Although most of the PE is in the worm, you want to make sure you lap to a good wheel also.

Roberto

[MarkH]

Noel, do  you have a  set of Zeus tables ? They would be very helpful if you haven't.  I'm  pretty sure  from memory that  Roberto is correct  with 55° being whitworth.  As for the Chinese  using it  all the slr cameras  manufactured  in China  use a whitworth  thread on the tripod mount. One thing I would  seriously consider in manufacturing is temperature variation.  I'm  sure that would far exceeded some of  the variations  you have mentioned.  I remember  Guy Martin  handling a 2" pin for a spitfire wing spar for a few minutes while giving camera  commentary.  His body  temperature in that short time was enough to  prevent that pin being fitted. You might give a wider consideration to material  selection
Hth Mark.