Problem with VBox III Accuracy

[linuxman]

Currently with the 24" superjack, I have 913 counts from just on each side of 37.5W and 139W. That comes out to 38 counts per inch.

Thanks for the extra data dem0nlord!

I am sure Gabshere is seeing your info.

It seems we have all faced the problem outlined through this whole thread.

I think you are right. It is a problem with the actuator (speed, slop, length), sensors (magnets, gear ratio), the positioning box GBox/VBox, and the receiver itself.

The best thing to do is try and eliminate all the mechanical problems possible, then find the right positioning box to maximize accuracy.

No positioning box will operate efficiently if mechanical problems still exist.

That's what I am trying to do here. Eliminate everything that might cause a problem and go from there.

I am not going to build the extension until sometime next week. I need to finish testing the GBox on this dish, move it over to the Birdview, test it there, and write a review before continuing the experimentation.

 

[gabshere]

ok i swapped out the von weise 24 " just to test it out to see if it could count high enough
i set 137 at zero and moved to 58 and its at 499 on the counter
but i stopped at 72 also to check out the count it was 430 ( from west)
but comming from the east it was at 435 eventhough there are more counts its still not accurate. i will have to look back at the counts on the 18 inch plus i took pictures and went beyond the limit and dish fell lol but its done that before. no damage done. i do have one problem and its with both arms....... when i set it up for my far west satellite i have to move the mounting bracket high up on the arm housing this puts me at the fully retracted position , no room for adjustments on the arm and limiting my length in reach. so when this 24 inch actuator reaches my failure point ( where the arm looses connection with the acme screw) its only 16 inches extended.

 

[linuxman]

so when this 24 inch actuator reaches my failure point ( where the arm looses connection with the acme screw) its only 16 inches extended.

Two things:
1. The inner arm is not being retracted all the way towards the motor before starting.
2. It isn't a 24 inch arm.

If it is a Von Weise, measure from the gearbox to the end of the outer cylinder and see what the distance is. It should be about 29 inches.

If it measures only 22 or 23 inches, it is an 18" actuator.

 

[dem0nlord]

Anybody have a 36" Von Weise? If so, could you measure the distance between the 2 mounting points if it is set to the minimum?

This was a problem I had with the Thomson-Saginaw, the thing was huge and I couldn't get that distance short enough to put it on the east side of the dish. If the Von Weises can go shorter like the Superjack, I'd be tempted to try one, maybe that would get me my 30W-139W travel. I assume the Von Weises do have a max extension limit switch that can be set, correct?

 

[gabshere]

i did take it apart clean grease it and put it back together ( the outside measured 28 inches roughly) the 18 inch actuator measures 21 inches.( 23 all the way to the motor)

now i probably don't have it set properly but it retracts almost all the way to 2 3/4 inch of shaft.

so i will have to look at the mount on the motor arm and adjust the shaft out so it can extend its length. my problem is placement of the mount so i can get my last west satellite.

i will have to check a couple of things

added pics

 

[dem0nlord]

Never mind, looks like Von Weise doesn't make a 36". Bummer.

 

[linuxman]

I assume the Von Weises do have a max extension limit switch that can be set, correct?

All the Von Weises I have do have a limit switch for both directions of travel.

so i will have to look at the mount on the motor arm and adjust the shaft out so it can extend its length.

Inner shaft all the way in with only the little pivot arm exposed. Then adjust the clamp on the outer tube to set it on the mount.

 

[gabshere]

Inner shaft all the way in with only the little pivot arm exposed. Then adjust the clamp on the outer tube to set it on the mount.

i have tried this before if i set it like that i only get about 12 - 14 inches of extension
if i set it to where 6- 8 inches are exposed then i can get 18-22 inches of extension
sounds weird .... i have limits switches but they are not set properly.

 

[linuxman]

i have limits switches but they are not set properly.

That's where the problem is.

But be careful not to let it extend more than about 18 to 20 inches. It will lose it rigidity if you go any further.

Edit: BTW the pictures look great!

After reviewing the pictures, you don't need to extend it any further. You aren't going to go any further East with it because it will hit on the mounting cap.

 

[gabshere]

i have overextended it several times lol it just looses connection with the screw and sometimes flops over but it doesn't fall far and no damage occurs its just hard for one person to reseat the arm lol lol

i will set the limits but i have to think things through mounting location on the arm and extension of the arm so it will extend. also my mount for the actuator doesn't have a pivot bearing like the superjack so this it also something i need to look at.

that first pic is with it extended too far and it doesn't have contact with the acme screw although the arm is still intact .. there is also a binding point which causes a stop point for the motor .

 

[linuxman]

Hey George,

In post #18 of this thread, I posted some pictures of the VW-24LC and along with them at the bottom are two pictures of the V76-5 24" actuator that I have a couple of too.

If you look at that picture, on the left is the limit block, and there is a round shaft with lobes on it like a camshaft. It has two screws that you can gently loosen and move to allow more travel or less.

One is to limit how far in, and the other is to limit how far out.

 

[gabshere]

i would suspect that one that has the sensor running to it looks like it has places for several magnets but i would have to check it out maybe your von weise guy could shed some light on it. maybe a way to make this one count twice as much

 

[linuxman]

Got the GBox V3000 switched over to the Birdview today.

Did a reset on it, and started as always on my TS satellite 91W.

Data from Before:

With the VBox II on my Birdview dish running it all the way from 30W - 142W with a 34:1 gear ratio motor I get 907 pulses with it.

With the GBox V3000 running pretty close to the same parameters (give or take a few as there is no hard limit) I am now getting 1977 pulses.

That is from negative 1101 on the East to positive 876 on the West. Which tells us that it will go past the 1000 mark and actually display 1101 on the front of the GBox and I assume the same thing would happen on the other end to measure a total of 4000 counts.

The accuracy has improved as far as landing in the correct spot, but is not quite as dramatic simply because the VBox II was giving me over 900 counts already.

To describe the magnets on the Birdview, it is a solid metal wheel with 8 embedded magnets, so yes Anole, you were correct in our most recent phone conversation, it is counting on and off of each magnet.

The drawback here is that the magnets are so close together that even a single click on the GBox remote (not the receiver remote) moves the dish 2 pulses. It will not give me a single pulse movement.

One other thing to take not of is I am using a Traxis 3500 on the Birdview, and was using the Pansat 3500 on the Unimesh, so the counting sensitivity works across different receivers.

I'll try to go up and get pictures soon of the proposed modification to the Unimesh dish we discussed earlier.

Fred

 

[Anole]

Got the GBox V3000 switched over to the Birdview today.

With the VBox II on my Birdview dish running it all the way from 30W - 142W with a 34:1 gear ratio motor I get 907 pulses with it.

With the GBox V3000 running pretty close to the same parameters (give or take a few as there is no hard limit) I am now getting 1977 pulses.

Yea, okay, that sounds like a 2X improvement.
Pretty much what we were expecting.

The accuracy has improved as far as landing in the correct spot, but is not quite as dramatic simply because the VBox II was giving me over 900 counts already.

Being twice as accurate as not good enough got you an improvement.
Being twice as accurate as pretty good isn't so spectacular, eh?

The drawback here is that the magnets are so close together that even a single click on the GBox remote (not the receiver remote) moves the dish 2 pulses. It will not give me a single pulse movement.

Well, the question is, can you move three counts or five counts?
If you move away, and come back, can you hit that odd count?
A better way to say it is this: if you are at 500 and can only go to 502 or 498,....
... but now you move to 450 and come back, can you sometimes land on 499 or 501?
If so, then it's just a simple matter of having too low a gear ratio, much as we thought from the beginning.
That's probably where the 70:1 or 72:1 gear motor ratio would improve the Birdview.

BTW, what's the deal with the receiver remote versus the Gbox remote?
I want to operate from my receiver's remote only.
What does that do for you?

I guess the more important question is, can you select any Gbox count for your landing position, then have the Gbox motor your Birdview over to it and land accurately?
If so, then I guess that's about all one could ask.

One other thing to take note of is I am using a Traxis 3500 on the Birdview, and was using the Pansat 3500 on the Unimesh, so the counting sensitivity works across different receivers.

Yes, it does.
Any other differences or shortcomings of one receiver over the other that might contribute?
I'll assume not, so just asking.

I'll try to go up and get pictures soon of the proposed modification to the Unimesh dish we discussed earlier.

The extension to the motor mounting point?
Well, are you pretty well done with this phase of testing?
I suppose so. Can't think of much that's left out.
Do you need to run it this way for a few days to look for gotchas?
Great job! - :up

 

[linuxman]

Well, the question is, can you move three counts or five counts?

So far it has been a pretty consistent 2 counts.

If you move away, and come back, can you hit that odd count?
A better way to say it is this: if you are at 500 and can only go to 502 or 498,....
... but now you move to 450 and come back, can you sometimes land on 499 or 501?

I have not noticed such a phenomena.

If so, then it's just a simple matter of having too low a gear ratio, much as we thought from the beginning.
That's probably where the 70:1 or 72:1 gear motor ratio would improve the Birdview.

I would agree. The only way to improve accuracy on the Birdview at this point is with the higher gear ratio motor.

BTW, what's the deal with the receiver remote versus the Gbox remote?
I want to operate from my receiver's remote only.
What does that do for you?

The remote for the GBox or for the VBoxes for that matter is only used for fine tuning the satellite in. (can be used separately if not using the receiver to move the dish) Once programmed with the proper landing spot, it is just a matter of choosing the satellite and channel you want with the receiver remote and off you go.

While I am on that subject, I want to express my two biggest complaints about the VBox III other than it's lack of accuracy.

There have been many times when I would select a satellite and channel on a different satellite and the VBox III would not respond causing me to have to choose another channel on the same bird and hit it again. Sometimes even have to select a different satellite and channel in close proximity and then move to the original selection. happens about 20% of the time.

The other complaint about the VBox III is changing channels on the same satellite would cause the box to click as if it was going to take off. I never detected any movement, but it was still annoying. Happens probably 40 to 50% of the time.

I guess the more important question is, can you select any Gbox count for your landing position, then have the Gbox motor your Birdview over to it and land accurately?
If so, then I guess that's about all one could ask.

So far in my limited tests today, it lands as programmed with pin-point accuracy.

Any other differences or shortcomings of one receiver over the other that might contribute?
I'll assume not, so just asking.

None that I can tell.

The extension to the motor mounting point?

Yes, but supposed to get nasty tonight and then much colder for about a week.

Well, are you pretty well done with this phase of testing?
I suppose so. Can't think of much that's left out.
Do you need to run it this way for a few days to look for gotchas?

Yes, I am pretty much finished. I would like to run it a couple of day like this, and then I will write the review.

Just another note, I have the VBox II hooked up to the Unimesh now, and the counts are the same as the VBox III. The big difference is that it lands as programmed with excellent accuracy as the VBox II's always have. And no annoyances.

 

[Anole]

Some of the Vbox 3 accuracy problems might be fixed, but if it just goes randomly weird on you, then that's not something I'd be interested in addressing.
Probably why it has been withdrawn from sales, too.

Now, if you want to have a discussion on pulse-doubling for the Vbox 2's, that might be worth a conversation.
But for the less intrepid, just installing the Gbox sounds like a good solution.
I think I can predict how your review will go, already.
"Buy it. Install it. Program it. Love it!" - :up

As to why the Birdview does not benefit . . .
That's not at all what I predicted.
You are saying basically that on the Birdview, you can only hit even numbers?
I was so sure with the 8-magnet wheel, you'd have better results.
Here is what I'm thinking the pulse must look like from your eight magnets passing the reed switch:

[FONT=Fixedsys];  
         _[/FONT][FONT=Fixedsys]          _  
[/FONT][FONT=Fixedsys]________| |________| |________
        [/FONT]

Where what we would prefer to see is more like this:

[FONT=Fixedsys];  
       ____[/FONT][FONT=Fixedsys]       ____  
[/FONT] [FONT=Fixedsys]_____|    |_____|    |_____|
        [/FONT]

In the 2nd case, we could count on the up and the down edges of the pulses (I'm still assuming that's what is going on) and get double resolution.

If the Birdview mechanism is too touchy to actually stop with the magnet near the reed, then that could just be the gearing.
Moving the reed closer to the magnet might affect the pulse width, but that would be an whole 'nother thread of discussion.
Likewise, we could whip up some hall effect sensors that did the trick but again it'd be a whole 'nother thread, too!
And we'd also have to determine if the current gear ratio would even benefit from pulse-doubling or not.

 

[linuxman]

If the Birdview mechanism is too touchy to actually stop with the magnet near the reed, then that could just be the gearing.
Moving the reed closer to the magnet might affect the pulse width, but that would be an whole 'nother thread of discussion.
Likewise, we could whip up some hall effect sensors that did the trick but again it'd be a whole 'nother thread, too!
And we'd also have to determine if the current gear ratio would even benefit from pulse-doubling or not.

I like your diagram, and I would agree that the wave isn't long enough for the gear mechanism to take advantage of it.

I also agree with your assessment of my review.

"Buy it. Install it. Program it. Love it!"

I like that and will probably use it.

 

[linuxman]

The Review of the GBox V3000 has been written and is in the Equipment Review section of the FTA area.

 

[SatPhreak]

If so, then it's just a simple matter of having too low a gear ratio, much as we thought from the beginning.
That's probably where the 70:1 or 72:1 gear motor ratio would improve the Birdview.

Annole, I know what you meant but a 70:1 is a lower gear ratio than a 35:1. The first number is revolutions of the drive gear compared to the driven gear. (A 70:1 would have twice the counts as a 35:1 would (assuming the same number of magnets on the drive gear).

Sorry, I don't know how to quote less than the whole thread.

 

[linuxman]

Place the two tags in quotes below before and after the text you want to quote:

but put a / just before the q in the second tag.