Possible permanent 76 Second error fix found, looking for opinions/input


Allen Ruckle
 

Steve,

      No I did not use any type of lapping compound.  There are several posts on the forum cautioning against the compounds.  My G11-T does not have any 76 second issues, however I have seen a G11-T that did and the owner reduced the error by installing higher quality bearings, but the error was not eliminated totally.

aruckle


 

It is interesting the 76 period was still there.  


 

Did you use lapping compaound?


Allen Ruckle
 

Dwight,
      I totally agre with you about exercising the mount to break in the gear train, which rubs all the gears together which smooths  the surface of the gears as they rub each other smooth.  

       When I purchased the G11-T RA upgrade I set up the mount with a minimal load and ran it for about Six hours at Slew speed of 800 x Siderial rate.  Initially the peak to peak error before break-in was above +5/-5 seconds of arc.

       I have been using the mount moderately for four years and my latest periodic error readings from PemPro are +1,6/-1.2 seconds of arc with PEC off.


aruckle


Allen Ruckle
 

Brenden,
      There is one and it is made and sold by Ovision.   I bought one and paid $ 540.00 for it. ( I don’t recall if that included shipping ).
it was very nice and included a test report indicating 3” peak to peak periodic error. 

aruckle


Peter Boreland
 

These tadpoles did not put up much of a fight! Second light (first real image) using Onstep with nema23 Clearpath servo direct driving the Ra worm after tuning the motor correctly. Used the photon sponge that is the ASI 294 to get this image. Ha 39x400s, Oii 23x40s, Sii 10x400s over 1.5 nights. Overall the mount performed very well. The meridian flip work flawlessly, both nights, and It actually parked it self where it started in the CWD position! I very much like the way Onstep first goes to the CWD position before it moves onto the target when flipping. Less likely to snag something.  I'm super pumped to finally get back to imaging. Clear skies everyone!



Peter Boreland
 

First light on the new setup early this morning. There is only a little good and a lot of bad sadly. First its seems my setup sustained real damage when Gemini II decided to strangle my mount by rapping the umbilical cable 1 1/2 times around the dec axis on Nov12th last year.  Last time in use until yesterday, and the final straw in fact. The tightly pulled cables placed great strain on the scope imaging train.  It now appears I have elongated stars across the image field regardless of exposure length. The same elongation shows in the OAG image also. I'm hoping that the problem is sag in the flattener connection at either the scope or OAG ends, and not damage to the focuser mechanism. I plan to contact Sky Watcher in the morning.  This is a real bummer!

I was able to get a few minutes of PHD2 data (11min in fact so not to draw too many conclusions). Looking at the Dec graph seeing was horrible. Dec error was 0.45" rms, but used to seeing 0.25" rms or less. When allowed to drift, polar alignment was good (spot on in fact) in Dec, but there was a lot of drift in Ra. Having viewed frequency spectra now of  mine and other peoples mounts, I was entirely struct by the lack of any perceivable 240s error in the drift graph. Well virtually no periodic error at all in fact. What is somewhat puzzling is the 1.2" 120s component, and we have our old friend the 0.7" 76s error.  The bane of Losmandy worm design. These guided out reasonable well using hysteresis guiding, but the higher frequency stuff appears about the same magnitude in both the raw and guided response, as was pointed out to me by Bob, so I'm wondering if this could be PHD2 generated noise, perhaps trying to correct for the Ra drift.  There is a lot of food for thought here and comments would be appreciated.

Image: Raw on top and corrected on the bottom to same scale. Just in case the numbers are hard to read the first horizontal in 0.25". 


Peter Boreland
 

Completed the direct drive one piece worm for the Ra axis. Spec:

* Nema23 Motor with 6400 optical encoder . 
* Indexed PEC (1/6400 precision)
* Ceramic bearings
* 5/8" Bellows coupler
* Torque limit safety shut down.
* 24V DC 



Peter


Robert Benward
 

Peter,
What I did below was to minimize the impact the motor weight would have on the movement and balance of the SLW, as per Henk's comment.  Implementing it is another story.  A "real ball bearing" refers to the pivot point.  You just can't get close tolerances and expect it to be smooth and loose.  What was a tight fit for the left bearing block really needs to be a ball bearing (which is not shown).  You have a lot of weight hanging off that little bearing point, it really should be a ball bearing.  And yes, you would have to do some machining to the main plate, and you would need probably two bearings with preload. 

The SS strap is to insure the Bellville spring that pre-loads the bearings do not bend the OPW bracket (which is does on mine, I had to put an offset bend into it).  It insures the physical connection between the two blocks is symmetrical.  This arrangement still needs to insure that the right block does not rise off the plate as the worm tries to climb the wheel.  More friction!

Regarding your DEC, if you have 20ms of backlash, isn't that considered successful?  It sounds like you have the DEC covered.  I may have to re-assess my planetary gearbox, but not now, when it gets warmer.

Bob

Bob


Peter Boreland
 

Bob,

Please explain further SS strap and what do you mean by real ball bearing? It looks like I would have to cut into my base plate? How do you get the pivot to work? Also for dec need to accommodate 48T pulley, is there room? Perhaps a top view would help.

Peter


Peter Boreland
 

Bob,

This is going to take some studying🤔 It looks very interesting. I appreciate you and Henks concern about my design. I increased  my spring tension to over 4lbs. I also polished the block bearing surface. The canter leaver problem is my biggest concern. Historically, I have set my worm ring gear mesh spacing using feeler gauges. I do a full 360 degree Dec rotation. I measure the current to three decimal places. I do not get bidding, and my backlash is under 20ms. This is fixed not SLW. I will have this installed hopefully over the weekend and will report on performance. I am very open to other designs if this is a fail.

Peter


Robert Benward
 


Peter, Henk,
Here you go:


Motor rides at near center of gravity. Has little impact on spring force necessary to push worm into wheel.  Note, with SS strap, top OPW bracket can be reduced to 1/8 material.  A real ball bearing supporting the pivot bearing bracket a real plus in this application.  Not the main plate needs to be slotted to clear belt.

Bob


Robert Benward
 

Peter,
I concur with Henk, the motor weight, bracket, and pulleys can be significant.  Alex's design is well executed, but was the friction measured?  Although not accurately quantifiable, a while back, I proposed a method to examine the friction.  Disconnect the belt from the motor and get a 3-6ft piece of open ended belt material.  Put it over the worm pulley, and draw the ends down and away, at right angles to the worm movement.  Alternately pull on one side then the other, kind of how you would use a wire saw.  Just feel the friction; you can quickly get a feel for the friction produced by a particular arrangement.  My spring plunger is 2.7lbs max, and I probably have it half way in.  Any more and I get a lot of friction.  The weight of the motor becomes a significant fraction of the soring force.  There is a bearing surface somewhere and that has to be addressed.  The finish on the main plate is for looks, and was never intended to slip bearing surface.  Also, Alex's spring looks large, and it is quite extended.  I would really like to know what the friction on that worm feels like.  Due to the motor mass, I think floating the motor on the worm is more deleterious to performance than having a moderate force from the belt pulling down onto the bearing surface.

There are not a lot of ways around this.  We want a light action worm, minimal force in to the wheel, but enough that we take out most of the backlash.  And ZERO forces from the belt.  A tall order to fill.

Another variation would be to put the motor and pulley on the worm pivot side.  That would move the mass to the center of rotation, minimizing its effect.  But that would require a significant redesign of the main plate.  I see a big slot to pass the pulley through the main plate to the motor below.  I need to draw this up...

I had considered a flexible cable drive or even a universal joint to direct couple the motor to the worm.  My biggest concern was the torque would wind up the cable, especially if it was long, creating backlash of sorts.  A pair of universals would have their own backlash issue.  The motor axis would have been at 90deg to the worm axis.  I even considered a 180 deg turn, putting the motor where it is now.

As for Losmandy's back off screw, once you engage it, I see it as defeating the purpose of the spring loading.  The spring will push the worm against the screw and not the worm wheel.

Right now, other than Peter's concern with my planetary gear box, I am happy with my spring loaded variation.  

Bob


Peter Boreland
 

Henk,

I know I'm crazy right? Like a fox!  I very much appreciate your thoughts. I do not see the additional motor weight as significant. Please also note I'm not a fan of spring loading the Dec axis. So if I lock down the screws some of your your augments against this design are perhaps not valid? However, I want to be fair and open minded, so want to give Bob's spring loaded worm design a fair go.  You can adjust for the extra weight with a heavier spring, which is the direction I decided upon. There is no such thing as high frequency noise when it comes to the Dec axis. This is not my implementation for Ra, perhaps there is some confusion? One final thing, I've never found use for the back off screw. I do preserve the functionality for nestaga reasons:) If you get the spring tension right, WHY would you need the back off screw? I do see the need to add support for the motor when spring loaded, as it cantilevers the worm. I plan to add an outrigger and a nylon screw to support the motor at the horizontal pivot point.

Peter


 

If I get your design correctly,  you have attached the motor to the moving worm block on the non-pivot side.  Personally I would never do this because the additional motor weight will require much larger forces to move the entire worm/motor assembly, which will generate way too much friction.  It is also partly uncontrollable because of the variable angle and gravity.  This can lead to all kinds of unwanted vibrations.  Is the reason for this that you want to avoid a motor axle coupler?  Your cure might be worse than the disease.  A spring loaded worm should be simple and light weight.  Even then you have to be careful, for instance I suspect the back-off screw of the Losmandy SLW design was added to relieve the pressure, just guessing.  Anyway good luck and let us know how it goes.


Peter Boreland
 

Bob,

The one thing I have already noted with this design, being that the weight of the motor is above and forward of the worm, is that the worm block will want to cantilever forward. While it may be possible to tighten the two bottom screws enough where this is not a problem, and allow free movement, I have decided to add an outrigger fastened to the bottom of the worm plate to provide support for the motor. An adjustable screw will make contact with the motor body with low friction plastic. 

Peter


Peter Boreland
 

On Tue, Jan 11, 2022 at 01:26 PM, Peter Boreland wrote:

Hank and Bob,

I've completed the mechanicals for the Dec axis. Really too cold to spend anytime at the mount fitting stuff. I've narrowly avoided disaster on several occasions here with the dimensioning of the plate and relative placement of stuff. The saddle knob clears the top of the bracket by about 1/16".  I think you can see better now why I placed the spring piston where I did.  The OLED display of SERVO42C, at the end of the Oriental Motor, is visible from the east side will in operation so that could be a good thing. The whole worm assembly is self contained, and lifts off easily, just have to loosen off the spring piston. It's very compact! It does look like a similar design would work for Ra, while preserving access to the altitude knob. The 24V DC is supplied on a seperate cable (2 pin, 24V, GND). The other connector supplies STEP, DIR, and EN signals. The cable will be CAT8, so each signal is shielded. Shield ground connection at the controller box. Pulley is 3:1 ratio (16T, 48T). Belt is 164mm G2T. The wiring tucks nice behind the motor housing so is out of sight.

Peter





Peter Boreland
 

Bob,

Does your Ra bearing block have one or two bearings on the driven end? My Ra block has a double set, but  Dec does not. When I squeeze the belt belt together and apply pressure on the pulley, I can see movement of the large pulley. This deflection is I believe coming from the bearing which is set quite a ways back as it is on the inside of the block. For the dec axis this is likely not a major issue, but for Ra it's a killer. I can not see how one would get a low 240s error. Moreover, one may also see a significant 76s error. If your Ra block has two bearings (inside and outside) the worm may be better supported. Even so I am not optimistic. I am direct driving my worm in Ra so this issue is not one I'm worried about. This discovery is a result of being able to easily apply pressure and seeing the result.  Not possible when it was buried in worm housing. The bearings are ABEC7 and are new. 

Peter


Robert Benward
 

Peter,
My only suggestion is to move the spring plunger higher (longer bracket) so its force is directly inline with the worm and worm wheel.  That hole for your cap screw is convenient, I assume it was part of the Losmandy worm assembly?  I had to drill my hole...

Bob


Peter Boreland
 

Bob,

It's not something I would do. My friend is 3rd generation machinist. He has lath so big it has crane/chantry set up above it. Last time a saw a lath this big was on the Battleship NJ in Philadelphia. 

Inspired by you SLW design, I decided to implement my own version using an 8-32 spring loaded plunger. They too come in two spring strengths. What I really like about this design, apart from how incredibly compact it is, is the ability to easily remove the worm block. The plungers arrive Tuesday. The locking nut does not, as it may appear in the images, get in the way of the worm block moving as it should.

Peter