USALS on CBand project FAILED

[Magic Static]

When I started in FTA some years ago I had a Ku dish and motor. I learned about USALS and used it. This was very cool and then found out it was not available in C-Band equipment. Later while reading thru some German forum I ran across a picture of a modified polar mount for c-band dish that was powered by a Ku motor and used USALS. It was one poor quality photo and no explanation of its working details other than a brief description. That stuck in my mind and I thought I would like to try something like that. Then I finally took the time to understand this forum was not a mega-commercial site of little interest to me. And I joined. The world of FTA was opened to me and I saw unlimitless possibilities happening here. WSI held a special on 6’ Dishes and I saw my opportunity to experiment with my USALS idea.

My idea was to modify the polar mount somehow and attach an SG2100 motor to it. I put the mount together and studied it. It was soon apparent there was no way I could make this mount work. In fact one had to take extra measures just to attach this mount to a pole. The polar mount is a simple design so I decided to build my own mount to suit me. The Ku motor is not a powerhouse so I needed to make the dish easy to move. The dish was very light made from (.02thin) gauge steel. I wanted bearings on the axis of the mount not just a long rusty bolt. This prompted me to base the mount on the headset of a bicycle. I went to a bike junkyard and was given an old frame and fork to work with. I cut out the headset and set up my mount. I went to a local metals store and bought many of the pieces I needed for scrap prices. Having a fully equipped auto shop at my disposal I had all the tools needed. I made a rolling satellite dish pole to work from. The SG2100 had been modified to use an external power source for the motor but still accepted DiSEqC commands thru the coax.

Winter weather set upon me and my satellite projects were put on hold. This gave me time to iron out a few more details on this dish. Indoor testing continued. I soon was able to make this Ku motor move this dish H-H with the help of springs. I was excited. This might work I thought. Of course I was always told it wouldn’t by smarter people than me. In fact I purchased the dish mover and VBox X at the time I bought the dish. But I wanted to see for myself. I needed a location for the dish and this turned out to be on top of a new storage building. And that was not here yet. So I had to put a lot of pieces together to get this dish going. Winter was long and hard but we survived. The shed was here.

Finally! (Where have I seen that before?) The pole was up and the dish came home from the shop. With my son’s help I soon had it secured to the pole. I couldn’t aim it yet. Still too damn cold and I didn’t really have a good way of doing that yet. But it’s up and running the final test, wind loading. Spring time is a good time for wind testing. And this is where we see the total futility of this project. That motor will never be capable of handling even light winds. It wobbled so badly I thought it might actually break the case and blow its guts all over the place. After a week of observation I pulled the motor off and installed the jack. After all I still wanted the C-Band stuff. Good weather showed and I aimed the dish. The Dish works well but no USALS L

 

[SatelliteAV]

Very innovative project and hardware design!

Thanks for sharing the project and photos.

 

[phlatwound]

Wow, what a nice attempt, great fabrication, sorry it didn't pan out for you.

Might I suggest removing that ugly Raiders logo...it still may not work, but at least it would look better?

 

[nicknjen]

Does anyone know how many pulses per degree are sent back on the KU motors? I have modified my birdview and taken off the magnetic reed switch and installed a rotary encoder. I get back 48 pulses per rev which comes out to about 20 pulses per degree of movement. I assume if you could match the pulses per degree of movement with the small Ku motors on a BUD mover than USALS should work...

So if anyone has that number please let me know, I have an assortment of TTL type encoders and would love to see if I could get that to work!

 

[Pixl]

Magic,

You are about 3/4 of the way into the project I've envisioned.
Using Usals would be cool because WHEN it is working it is very accurate. More so than my actuator arm.
I had bought a used Usals motor to take apart and see what makes it work. I had hoped to use the Usals motor to control my actuator which has decent power and speed. I expected to find some sort of accurate position sensor on the main shaft. Not so. Instead I find the the sensor is counting pulses on the tiny little motor that is then geared down many times to the main shaft. You can't push the whole thing back wards thru all those gears, too high of ratio. However with your setup you might be able to do this.
A command to move is sent to the Usals motor.
You tap in to the voltage that goes to the small motor and use it to also pull in a relay that switches a high current 36v. power supply to the actuator arm. The actuator arm would be moving the dish and the Usals motor would be spinning the gears, they would work together and when the Usals position is reached it would park. A second relay and some diodes would be needed for the opposite direction.

I never got past the part where the Usals motor would be mechanically coupled to the polar shaft. Looks like you already did that.

 

[Magic Static]

The rev counter on a Ku motor is on the armature shaft the little electric motor. This gives you thousands of counts. Many more than a CBand mover. This enables the USALS calculations to be very precise, no matter what the arc may look like from your vantage point. Ahh Pixl beat me to it

 

[Magic Static]

What I need was a way of locking the dish down once it moved into place. Like a screw jack. Kind of a vicious circle here. Leads me to believe if it was easy enough for a non-engineer type like me to figure out, somebody would have already. What made it easy for the KU motor was the fact it's also the mount.

 

[Pixl]

Does anyone know how many pulses per degree are sent back on the KU motors? I have modified my birdview and taken off the magnetic reed switch and installed a rotary encoder. I get back 48 pulses per rev which comes out to about 20 pulses per degree of movement. I assume if you could match the pulses per degree of movement with the small Ku motors on a BUD mover than USALS should work...

So if anyone has that number please let me know, I have an assortment of TTL type encoders and would love to see if I could get that to work!

I don't think the pulse count from the actuator will work correctly due the "pulses per degree" changing as the actuator changes it's leverage ratio as it gets near the far East side of the swing. The Usals motor gear ratio is consistent across the whole arc.

 

[Anole]

I too thought about this sort of project.
Took apart a dead SG2100 to have a look, couple of years ago.

However, I didn't see a convenient (and accurate) way to multiply the power of the 2100.
Nor a way to push on the motor and have it move the pulse counter.
Maybe I'll give it a re-think.

One of our intrepid members -did- hang that same 6' dish directly on an SG2100 without the aid of better bearings.
He located the dish beside his house and behind a fence as I recall, low where it wasn't subject to the wind.
It ran, but I believe he abandoned the idea. (?)

Anyway, what I came up with, was the need for a modified super-Zbox to run a BUD which had an H-H mount.
Yea, not everybody has those, but the two BUDs I chose have H-H: Birdview, and Ajak-180!
I did post about it back around the time Sadoun was talking about adding servo control to his Vbox.
Nobody seemed much interested in my mod at the time.

ps:
nickngen has a birdview which is H-H and high resolution pulse counter, so he/she's thinkin' along the same lines...

Oh, and one last thought: brainstorming these project on the forum, often lead to a multitude of fresh new ideas to try.
I have a few buddies I always run crazy ideas by first... or post 'em here for comment.

 

[linuxman]

I think the answer is really pretty simple.

The receivers all have the logic built into them. The motors all know how to respond to the logic.

The answer is figuring out a gear reduction system perhaps involving a sprocket/gear on the shaft of a SG2100 or similar motor, with a sprocket/gear mounted on the polar mount on a C-Band dish with the same precision that the little motor has.

The smaller motor would then be powerful enough to turn the big dish with the same precision without having to re-invent the wheel.

 

[Anole]

And people say their Ku dishes are too slow now... :up -

One idea my local buddy Joe and I came up with was a tracker/servo-follower idea.
We'd sense the movement of the 2100 (optically, or however), and move the BUD with a big motor, to match position.

 

[Hermitman]

If you're looking for more pulses per degree of movement, why not use the actuator motor itself. While DC voltage is fed to the motor to make it move, the voltage at the motor itself is actually pulsating DC. This is due to the communtator contacts moving under the brushes. Someone should be able to come up with a simple pulse detector circuit connected to one of the brush wires and send the pulses to whatever pulse counting device you're using. Good luck.

 

[linuxman]

If you're looking for more pulses per degree of movement, why not use the actuator motor itself. While DC voltage is fed to the motor to make it move, the voltage at the motor itself is actually pulsating DC. This is due to the communtator contacts moving under the brushes. Someone should be able to come up with a simple pulse detector circuit connected to one of the brush wires and send the pulses to whatever pulse counting device you're using. Good luck.

That's really quite a good idea, but the whole theory of USALS won't work on a normal linear actuator system, simply because the amount of pulses between satellites is not uniform across the arc.

The USALS idea will only work on AJAK, Birdview, or similar true H-H mounts/motors that have a uniform distance travelled between degrees across the entire arc.

 

[mr3p]

I've been moving my WSI 6ft dish with an HH motor since the day I installed it using an HH120 and SG9120. I've posted pictures of the modified mount and everything assembled in a couple of threads if you do a search. It defintely can be done.

 

[Magic Static]

I really didn't have a problem moving the dish either. It just wouldn't hold still in windy conditions. I would like the try the SG9120 but it doesn't fit the bracket for the SG2100. I would have to remake that part of it. I may try that at some time later. Or I may get off on another crazy idea. One never knows.

 

[nicknjen]

As lunixman mentoned above, I have a Birdview.. everything is liner across the arc. As for the Vbox with USALS it HAS TO KNOW how many pulses or closures or counts it takes to get from Satellite "A" to Satellite "B".. it is going to move and count pulses or closures across the connectors marked "sensor" on the Gbox.

For example does anyone who uses USALS know how many counts it takes to move their dish from Galaxy 28 89W to Galaxy 17 91W? All I want to know is how many digets does your V or Gbox counts to move from one satellite to another whom is using USALS.

As for getting the feedback, I have that already taken care of with a rotary encoder.

 

[nicknjen]

I should elaborate a bit more on my above post.. what I would wand is also need to know is the two birds you are moving between and the USALS information. What I am trying to do is get a PLC to talk USALS (getting one of my EE guys from work to help me) and move the dish rather than using at V or G box. I know my Vbox does not do anything with USALS but was not sure if anyone had one that did.

But what I was above mentioning is I was trying to find out the resolution it needed to be.. or what the receiver is telling the mover to do...

Figuring out the USALS "lingo" will be a project for my EE buddy.

 

[Magic Static]

The geometry involved in calculating USALS is a little more involved than a certain number of clicks between sats. The distance between two sats can be different depending on where on the earth you are looking from and where in the arc they are. These are not big differences but they prevent accurate tracking without taking them into consideration. The large amount of counts provided make the calculations very precise and compensate for these differences.

 

[Anole]

For all the people who put 24-magnet wheels on their Birdview, AND used the Gbox which counts on both the make and break of the reed switch, we get 48 counts per revolution of the main worm gear.
nickngen says he's got a 48-pulse shaft encoder (though I don't know where he stuck it), so presumably he's getting 48 counts, too.

The USALS calculations are done in the receiver.
And it needs to know where on earth you are located, as well as where the birds are on the arc.
The motor needs only the most simple intelligence to do the job.

If I remember correctly, what I proposed was a Zbox that could be told the number of counts per degree.
(or maybe it was fractions of a degree per count)
All Birdviews with the same magnet count would be the same, so we could publish the calibration number for all to use.
AJAK and other H-H mounts would be different. But again, all alike.
The mythical Zbox would report that to the receiver, then carry out movement commands given it.
But my memory is a bit fuzzy, as it's been a few years ago since I discussed the idea.
There was little or no interest from the forum for the plan.

 

[Magic Static]

Now owning one of those fine Ajak 180s, I am particularly interested in giving it USALS cabability. If you are sure that USALS is calculated by the receiver, then I think we can implement this. I will have to give this some thought for awhile. If USALS originates in the receiver than the receiver would have to query the motor as to the number of counts per degree. Unless this is a standard the receiver would not know this.