Building an H2H mount

[Greg Mueller]

I'd really like to have an h to h mount. So much so that I'd build one. I have a machine shop so I have the tools.

I was thinking an old aluminum flywheel for the actual worm gear (wheel) but getting a worm gear to match could be a trick. Also flywheels usually have a big gear ring which is heated up and slipped on the flywheel, so If you cut a flywheel in half, it would seem the gear ring would fall off.

I wonder what the original manufacturers did for gear segments and worms. Anyone know?
If someone has an H2H mount, could you measure the diameter of the gear segment?

The rest would be pretty easy

 

[1captain]

Greg I don't know if a fly wheel gear will be strong enough or not, an most of them their groves for the teeth are not near as deep as the H to H mounts gear that I have here. have 4 H to H dishes, 1-Ajax this gear is 1/2 moon an is the widest of all gears about 1 1/2" wide, have a birdview an this gear is about 3/4" wide also a 1/2 moon gear, now also have a Channel master an this gear is a full 360 or round gear also fairly thick maybe about 7/8", an have a old CFI on a 1.2m dish with a round gear, now this 1 is the smallest in diameter about 6" an about 3/4 of a inch thick. now on the all of the drive gears here, execpt the bird view gear, the middle part of the gear is like machine down to stop bindind from the worm drive.

now you can still buy H to H mounts but they have goten pricey, but when you start looking at what a machine shop charges, it just may be worth it in the long run, an a lot less headach. but I agree with you nothing is better that a H to H mount, stronger, very acurate on stopping at a satellite with out actuator slop, an very dependable also.

 

[Greg Mueller]

Someone is still selling them?

 

[1captain]

Someone is still selling them?

Check you're PM

 

[Anole]

motors, Motors, MOTORS:

I have also been fascinated by the H-H mounts.
And, over time, I've collected many good photos of their inner workings.

The AJAK came in three models:

posted by Glater
- light - only suitable for little light weight mesh dish

posted by Satfaca
- mediocre - not confident it'd handle the 100 lb dish I had in mind...
(but I'd test it if I could find one!)

posted by Linuxman
- battle-tank! - - suitable for flipping over a small car

Stogie posted some good pictures of the motor/mount on his Paraclipse.
Don't recall what that was, but it's a big sucker!
Had what may have been a chunk of motorcycle chain wrapped around a large semi-circular form, then a cog engaged it from the outside.

LoneWolf once posted some pix of an unknown H-H.
It was built a bit differently. Smaller than some, but still looked sturdy.


What makes an H-H (or any motor) perform well, is to have lots of counts between birds.
On the forum somewhere (likely in the BUD section), there is a long discussion with Linuxman on magnet wheels and total counts across the sky.
It was likely Birdview-related, as he was making a multi-magnet wheel for his motor.
It'd be a good read for anyone refurbishing an old mount or making a new one.


Now, if I can bring up a favorite subject of mine (which seems to fall on deaf ears), how about a controller that drove the H-H, but with USALS input? -
I don't think it would be very difficult.
Don't know if it'd become a commercial success, but that's only due to the small installed base of H-H BUDs! -

 

[Lone Cloud]

Well, supposedly a G-box works with USALS, but I have never found it to be accurate. I change satellites in my receiver and who knows where the G-box is going to send the dish?!

Maybe it's because I rename all of my satellites to reflect orbital location rather than their current designations (whatever those are this week). For example, G10 (18 now?) I rename as Galaxy 123 - I don't have to indicate east or west because I can't see any eastern hemisphere satellites)

Maybe that renaming is screwing things up for the G-box. I don't know why it should, because when I hook the receiver back on the 1 meter Ku, with motor, it doesn't care what I named the satellite, it just goes there.

In any case, I have gotten to the point where the G-Box is not connected to the receiver at all. It is a stand alone positioner.

I think a USALS big dish mover would meet a need. Especially if the bugs could be worked out.

 

[pendragon]

Now, if I can bring up a favorite subject of mine (which seems to fall on deaf ears), how about a controller that drove the H-H, but with USALS input? -
I don't think it would be very difficult.
Don't know if it'd become a commercial success, but that's only due to the small installed base of H-H BUDs! -

I performed a rough assessment of this by example a month ago. I recently picked up a couple of GBOX V3000s From Sadoun to run my 3m on a Ajak H180 and my 1.8m on a Moteck H180. After diagnosing a bad sensor on the Moteck, both drives now regularly run back and forth between 37.5 and 137 without problems or resyncing.

Before this I had employed an analog receiver and a VBOX. To save time setting up the GBOXes, I decided to calibrate the angle offsets for each count on the V3000s. Translating the orbital positions into motor angles for my lat/long, I came up with roughly 0.05 degrees/count on the Ajak and 0.09 degrees/count on the Moteck. I then did a 'spreadsheet USALS' to construct nominal motor positions for each bird.

For the most part this got me pretty close on both dishes. For laughs I went through and tuned each position with my spectrum analyzer and compared the 'best position' to the calculated. Now finding the 'best position' isn't an exact science, and admittedly I didn't pull the TLEs for each bird to find where each was at the time in relation to their center-of-boxes. These are both unconsidered error perturbations on the following.

What I found was the Ajak was reasonably constant in its angular increment across the range, but there were local variations. The Moteck was nowhere near as accurate. Part of this may be play/backlash in the Moteck drive, which you can feel twisting the dish by hand. By contrast the Ajak has almost none. My conclusion was that although one could calibrate a nominal motor angle for a HH drive, there would be a loss in accuracy if this was built into a positioner and run by USALS. By contrast, one can control the tolerances of building a USALS motor, and potentially burn a calibration table into the microcontroller for each unit if needed.

Conceptually it would be easy to implement a USALS capability into positioner, but the variations in positioner accuracy for the installed user base could be a big headache.

 

[Greg Mueller]

Fascinatin stuff !

 

[Greg Mueller]

Is there a limit to how many counts there can be between positions? I'm not exactly sure how the system works. If you weren't constrained by having to make a profit, you could make a little geared gizmo that could have a lot of counts in a postioner.

 

[Anole]

Pendragon -

Thanks for your USALS comments.
Didn't mean to drag Greg's thread too far off course.

Regarding Gboxes and magnet counts, I refer you to a thread by 1captain, where he put what looks like a ChannelMaster - Primestar 1.2m dish on a real BUD H2H mount.
(after adapting the mount to an offset dish with some beautiful metal work)
Quite a lot of valuable info posted by him, that is directly pertinent to this thread!
It's a must-read. :up

There's also an old thread by Linuxman, where he first tested a Gbox, and we discussed how many counts he got across the arc.
I should refresh my memory by re-reading that thread.
I know we discovered quite a few valuable lessons.

But as for gears to increase your count, just remember that positioning accuracy is dependent on backlash, and slop...among other things.

 

[pendragon]

Is there a limit to how many counts there can be between positions? I'm not exactly sure how the system works. If you weren't constrained by having to make a profit, you could make a little geared gizmo that could have a lot of counts in a postioner.

We ultimately are faced with resolution vs. accuracy, and in this case significance. One could increase the angular resolution by increasing the number of counts, but unless the drive mechanism accuracy was similarly increased, the ability to steer to exactly the same position would not improve. We are also dealing with non-rigid structures, horribly imbalanced mechanical systems and significant dynamic loads, like wind, to boot.

Perhaps more pertinently, the beam pattern of a BUD feed is not a smooth, symmetrical shape that reaches a clear peak dead on-axis. And this all varies over the frequency band(s). The strongest peaks may not be on-axis. The sidelobes may pick up adjacent orbital positions at significant levels. The birds are not stationary and wander in position throughout the day. Thus the 'best position' may only be for a particular transponder on a particular satellite for one instant in time. My measurements bordered on the subjective because it was difficult to reliably identify what position was the best, even given multiple trials.

Increased angular resolution would have a negligible effect in improving BUD reception. Tracking of the orbital arc on the ground is only approximated by the traditional azimuth and declination adjustments. Even if these are set perfectly, the orbital positions will only cross our track at a few points (sorry, I'm too lazy to do the math).

I expect most BUDs can be positioned far more accurately in the motor angle than cross-track. For example, let's say I want to adjust my 3m to true south to an accuracy of 0.05 degrees, similar to one count on my GBOX. At my latitude the dish's polar axis is at about a 45 degree angle. If I loosen the polar mount on the pole, and it doesn't wobble in the process, a change in 0.05 degrees corresponds to rotating the 3m dish about 1mm at the dish rim's lowest point when it is at true south. For a smaller dish, we would be talking fractions of a mm. Even though I know my true north to better than 0.1 arc seconds and can make deadly accurate measurements on the dish itself with my theodolite, that's still not an easy adjustment In reality my mount's bearings and the positioning of the feed simply are nowhere good enough for this to matter.

 

[Greg Mueller]

In CNC milling machine land, better accuracy is obtained when you put the resolvers on the table (of the machine) itself, rather than on the motor. That way you can get true position of the table rather than counting "steps" (pulses etc) on the motor with all the associated errors, as small as they are.

So I'm thinking it would be more accurate to put an anti-backlash geared counter directly on the polar axis, rather than have it driven off of the rear end (etc) of the motor.

 

[Anole]

Greg -

Yes, that's true, but the dish only swings less than 180°, so you're faced with some really close magnets! -

One of the guys did find some sort of shaft encoder which was:
- inexpensive
- high resolution
It was in one of those Linuxman/Birdview threads.
Price was in the $40 range, and was dip-switch selectable for (as I recall) 500, 1000, 2000, and maybe 4000 counts per revolution.
I had personally favored a home-made hall-effect sensor, but either would require another (non-standard) wire and power to the sensor.

Trying to put feedback on an existing mount, I feel somewhat comfortable with a magnet wheel on the final worm to the gear that swings the dish.
But, perhaps with some other mechanism, or sensor, you could come up with an other workable idea.
You know, it's not always the best or most accurate solution that's best for the job.
It's the one that is good enough, easy to build 'n service, that wins the race.
So, I'm open to all sorts of imaginative ideas.

 

[14karat]

Ok... I've wondered for a while, so I'll just ask...
Anyone attempted a hydraulic actuator?

And I agree with you Greg about driving the counts directly off the dish.
My thought is mount an engine flywheel on the pivot axis, with a pinion gear driving the counter. For the counter, instead of reed switch / magnet issues, use a starwheel with infrared led & receptor (works like a mouse wheel)
Dirt cheap and should work better than the mags... drawbacks would be in the circuitry...

 

[Greg Mueller]

I was thinking of a couple of anti-backlash gears mounted to the polar shaft. You could set up any ratio you wanted.
I got some little optical type sensors once that I was playing with. They were encoders off the back of a small (but high speed) electric (dc) motor.
Naturally I took it apart to see how it worked.
It looked like they printed a series of "spokes" on a plastic disk which interrupted a little led light source. At the time I was interested in encoders for an astronomical telescope (real high resolution encoder stuff).

 

[14karat]

And I'll go one further with the flywheel thought...
Use a long shaft motor with a wormgear to drive the flywheel... there's your H-H actuator...

 

[14karat]

Greg, those are used in cheap PC mice and in 2-axis... extremely hi-res & the electronics are all there, just need to be powered somehow.

 

[Greg Mueller]

I was wondering about that. The resolver would give you pulses, but you'd have to convert them in something that would make it all the way into the house

 

[14karat]

I think the PS2 mice have signal outputs of +-5V (or maybe it's 0-5?)
Interesting being TTL signal... wonder how the mover would interpret it since it's just looking for a closed circuit...

 

[ECruzBUD]

My H-180 gear motor & BUD

Hi Folks,


This is Eugene, from Seattle, Washington.

I read all of your posting in regarding H2H mount issue.

Me too, I do have problems. Not realy. I haven't even try it yet.

I have 2 of H2H mounts.

1. is KTI H-180 mount with a ring-mount that goes onto the
back of the 10 foot BUD.

2. I have a AJAX H-180, but no ring mount included. Just the
gear motor thing.

Where do I find a ring mount that attaches to the AJAX
motor onto the back of the BUD???

Where do I find an extra internal gear motor part for the KTI H-180 gear
motor?? A motor itself, it's hard to find one because KTI is no longer
in business. I would always like a spare part before I do put up
a new part to be use.

So, I have 2 of H2H gear motors in box, one of them without
a round ring mount, they're sitting around since 2003 and 2004.

I would LOVE to replace from the actuator arm from one BUD,
and replace it with the KTI H-180 mont with it's round ring mount
and gear motor. But I would like to be prepaired for an extra
motor for a back up. I could send you photos of the motor that
I'm refering to.

I hate that Superjack 24 inch actuator arm that I have now.
It's weight 19 pounds, versus 113 pount dish. Not very strong
to keep it study. But whenever there's a wind storm, it shook
or wabbles my BUD. Someone in California coast told me that
thier KTI H-180 mount is capable of wind storm at 50 mph.
on his 10 foot BUD.

Anyone could help me with referals or suggestions, or ideas???
Or photos of my H2H gear motors, or has a round ring mount for
the AJAX H-180? You're welcome to contact me.

Thank you.
Regards,
Eugene