I have had a problem over the years of drift of my satellite locations necessitating my reprogramming their positions. My masts are vertical to a fraction of a degree.Dishes are mechanically solid. Actuator cable is shielded and grounded at one end only. One end of the cable has a capacitor across. I've had this problem in 4 locations with with 7 dishes. Many V Boxes and Titanium ASC-1's; many actuators; many receivers. I'm asking, is this a common occurrence? Or is it something peculiar to me. Thanks.
Actuator Miscounts
- Thread starter rikoski
- Start date
- Latest activity Latest activity:
- Replies 14
- Views 1K
What is commonality between these locations and/or equipment?
I typically run the shop dish over 1000 counts per day between 5 or 10 satellites and haven't had the need to adjust for a miscount in at least a year or more. The last adjustment was due to a failed reed switch.
Over the years of using a mix of V-boxes, G-Boxes, ASC1s and many brands of IRDs , I have found the most common cause is to be satellite or channel surfing using a STB where navigating a satellite or channel list signals a dish "GoTo" with each item passed. If you instead use a STB where the user selects a satellite or channel then OK to "GoTo" the selection, there rarely are miscounts.
I typically run the shop dish over 1000 counts per day between 5 or 10 satellites and haven't had the need to adjust for a miscount in at least a year or more. The last adjustment was due to a failed reed switch.
Over the years of using a mix of V-boxes, G-Boxes, ASC1s and many brands of IRDs , I have found the most common cause is to be satellite or channel surfing using a STB where navigating a satellite or channel list signals a dish "GoTo" with each item passed. If you instead use a STB where the user selects a satellite or channel then OK to "GoTo" the selection, there rarely are miscounts.
Hey Hi. I recently added Ku to my C Band dish. "Bumping"" my ASC-1 E & W to peak signal caused accumulated lost counts over time. Check your actuator motor for slop. Also check the motor to actuator coupling for slop. I had to cut strips of brass shim stock and bend pieces into a u-shape and fit in the coupling. That fixed most all of the missed counts. Check your dish pivots too for slop.
This right here is what I have seen in the past. When everything else is checked and is top notch. I have found where a dish has sloppy pivot points that had worn out. Seen one time where the dish was actually rotating on the pole just a mm or so but enough to add up over a few moves.
But are drifts common? I've had this problem since I came back to satellites in 1983. Brian does 1000 counts on 5 or so satellites. I do a lot more than that. I use my 5 dishes to watch tv. Maybe that alone suggests that there will be drift regardless about how careful I am in setting things up. The dish driven by the Ajak HH180 drifts as much as the ones with Venture arms. Is drift common?
Drifts can be common. They happen for a variety of reasons and can be very difficult to track down. Especially when you have more than one cause. This is exactly why there is a "shift" function on a VBox or ASC1. One of the biggest problems I had was having too much of a load on the dish motor because of heavy snow or extreme cold. This would cause voltage irregularities in the logic side of the DiSEqC box and cause miscounts. So between correcting issues I can and avoiding ones I can't, I don't adjust my positions very often.
A timeline since getting my dish running again around 2 years ago.
Installed a new LNBF from Titanium and a new receiver.
A well used 12' dish with bolt and bushing pivots. Used an old GI DC2+ to move the dish.
Saginaw actuator not moved in ~10 years. Lubed it, rebuilt motor. Installed a new reed switch.
GI box used as E-W control. Arc swept was 131W-55W.
Noticed that when peaking signal there was a 'dead band' of several counts where signal would fluctuate.
Non repeatable positioning (count log for a sat was 550 for peak, next time 445, then 553 for example).
Count errors would accumulate. Got worse when peaking signal for each sat over time. Lots of eraser use on paper.
Replaced/modified dish pivots with pillow block bearings. 4 bearings and hardware around 60 bucks.
Dish was now solid mechanically. Grabbing the dish edge and bottom rim and letting go would return signal to peak.
Ordered and installed an ASC-1. Programmed in all sats for DISEqC control.
Still got count errors. Not as bad and mostly from adding new sats and bumping the dish W-E to peak sig. before storing.
Pulled the motor and benched it. Ran it with 24 VDC with an oscillosope hooked to the reed switch with a 5 VDC supply loop.
Running the motor showed erratic pulses with varying pulse length. Ex: pulse, pulse, pulse, pulse, pulse, pulseeee, pulse.....etc.
Magnet was a 6 sector and attached to the actuator coupling gear. Tried varying the reed-magnet distance for best pulse width.
Put up with that for awhile. Found a new never used SuperJack 36". It was reed sensor equipped. Installed and used it.
Still got count errors. Benched the motor as above. Pulses were erratic.
Fabricated a hall sensor pickup and interface for the ASC-1. Oscilloscope showed consistent pulses but a long trailing pulse when the sensor was shut off by the magnet.
Looked up the data sheet for the sensor. Built an encoder wheel with mini neodymium disc magnets from Amazon.
Data sheet said that the sensor would be driven 'hard-off' by the presence of a S pole from a magnetic field.
So the encoder wheel has 20 magnets with opposing poles around the edge of the disc. Giving 10ppr.
Made the hall sensor distance adjustable from the encoder wheel. Set the distance so 'scope showed a clean square waveform.
There are hall devices with a Schottky output (this one doesn't have..but will be soon) which will assure a square pulse.
I am now getting very accurate dish positioning with awesome repeatability. Adding a Ku LNBF though has given counts errors as I've scanned the arc for sats and had to bump the actuator to peak signal before storing the position in the ASC-1.
It's when doing the bump or suddenly changing motor direction is where counts errors show up now.
A week after getting all usable satellites stored in that stinky ASC-1 and I get no errors.
Blaming the controller added up to just being able to give it a good signal and eliminating dish slop.
Industrial background and knowing that a bad encoder/drive and coupling and other mechanical tolerance defects will make a CNC operator go nuts.
I'm all out of ideas. Except for some wire brushing and Rustoleum.
Installed a new LNBF from Titanium and a new receiver.
A well used 12' dish with bolt and bushing pivots. Used an old GI DC2+ to move the dish.
Saginaw actuator not moved in ~10 years. Lubed it, rebuilt motor. Installed a new reed switch.
GI box used as E-W control. Arc swept was 131W-55W.
Noticed that when peaking signal there was a 'dead band' of several counts where signal would fluctuate.
Non repeatable positioning (count log for a sat was 550 for peak, next time 445, then 553 for example).
Count errors would accumulate. Got worse when peaking signal for each sat over time. Lots of eraser use on paper.
Replaced/modified dish pivots with pillow block bearings. 4 bearings and hardware around 60 bucks.
Dish was now solid mechanically. Grabbing the dish edge and bottom rim and letting go would return signal to peak.
Ordered and installed an ASC-1. Programmed in all sats for DISEqC control.
Still got count errors. Not as bad and mostly from adding new sats and bumping the dish W-E to peak sig. before storing.
Pulled the motor and benched it. Ran it with 24 VDC with an oscillosope hooked to the reed switch with a 5 VDC supply loop.
Running the motor showed erratic pulses with varying pulse length. Ex: pulse, pulse, pulse, pulse, pulse, pulseeee, pulse.....etc.
Magnet was a 6 sector and attached to the actuator coupling gear. Tried varying the reed-magnet distance for best pulse width.
Put up with that for awhile. Found a new never used SuperJack 36". It was reed sensor equipped. Installed and used it.
Still got count errors. Benched the motor as above. Pulses were erratic.
Fabricated a hall sensor pickup and interface for the ASC-1. Oscilloscope showed consistent pulses but a long trailing pulse when the sensor was shut off by the magnet.
Looked up the data sheet for the sensor. Built an encoder wheel with mini neodymium disc magnets from Amazon.
Data sheet said that the sensor would be driven 'hard-off' by the presence of a S pole from a magnetic field.
So the encoder wheel has 20 magnets with opposing poles around the edge of the disc. Giving 10ppr.
Made the hall sensor distance adjustable from the encoder wheel. Set the distance so 'scope showed a clean square waveform.
There are hall devices with a Schottky output (this one doesn't have..but will be soon) which will assure a square pulse.
I am now getting very accurate dish positioning with awesome repeatability. Adding a Ku LNBF though has given counts errors as I've scanned the arc for sats and had to bump the actuator to peak signal before storing the position in the ASC-1.
It's when doing the bump or suddenly changing motor direction is where counts errors show up now.
A week after getting all usable satellites stored in that stinky ASC-1 and I get no errors.
Blaming the controller added up to just being able to give it a good signal and eliminating dish slop.
Industrial background and knowing that a bad encoder/drive and coupling and other mechanical tolerance defects will make a CNC operator go nuts.
I'm all out of ideas. Except for some wire brushing and Rustoleum.
Thank you Brian, Arlo, Magic Static, and Michael for your observations of dish position drifting; your comments of what you did to try to fix the problem; how well they worked; and suggestions of what I might try.
As Magic Static suggested, I will try the shift function on my ASC-1. I don’t recall seeing it in the ASC manual, but I’ll look again.
A common thread was that drift was due to slop, caused by aging dish pivots and jacks. Makes sense, but I had drifting right from the start when my dishes and installations were new.
I’m thinking that it might be in the nature of reed switches and plastic actuator jack parts to not hit the mark precisely every time.
So maybe I can minimize drift but not eliminate it entirely. Reprogramming has been a fact of my life and might just have to continue.
In my case, reprogramming is not easy because the receiver boxes are scattered all throughout my house while the ASC-1 mover controls are concentrated at a central point in the basement where all of the dish cables enter the house.
Again, thanks for sharing your experiences. I appreciate it.
As Magic Static suggested, I will try the shift function on my ASC-1. I don’t recall seeing it in the ASC manual, but I’ll look again.
A common thread was that drift was due to slop, caused by aging dish pivots and jacks. Makes sense, but I had drifting right from the start when my dishes and installations were new.
I’m thinking that it might be in the nature of reed switches and plastic actuator jack parts to not hit the mark precisely every time.
So maybe I can minimize drift but not eliminate it entirely. Reprogramming has been a fact of my life and might just have to continue.
In my case, reprogramming is not easy because the receiver boxes are scattered all throughout my house while the ASC-1 mover controls are concentrated at a central point in the basement where all of the dish cables enter the house.
Again, thanks for sharing your experiences. I appreciate it.
Here's a few tips for programing. Maybe will help, maybe not.
Forget about pulses per inch of actuator travel. You want to know pulses per degree (or 2 degrees actually).
All sat positions are stored in the receiver as a DISEqC Stored Position. And I used the serial cable and software to program sat names per satellite and position.
I determined that 127W was at one position on the ASC-1. Then hunted for 125W.
Turned out to be 30 counts. So 30/2 degrees lower arc. Pretty darned close.
Found 131 almost too easy. Sat charts. Signal finder and a good strong tp.
Stored every sat in its slot in the ASC-1.
Getting up past 101W 2 degree counts went down to 26 pulses. Knowing that got me in the ballpark all the way out to 55W. Then going back to the lowest sat on the horizon and peaking signal, storing.
Pretty rock solid. Just added a Ku lnbf. The 30 pulse, 26 pulse rule worked out slick.
In my experience it is quite usual the miscounts.
For this reason many positioners implement "goto reference".
Where the actuator is moved to one of the mechanical limits and the count is reset.
The miscounts can be for various reasons, the most common is the same noise induced by the motor.
I solved them using UTP cable for the sensor and an optocoupler with a separate power supply from the positioner.
Every many days I have 1 bad pulse or 2 at most
For this reason many positioners implement "goto reference".
Where the actuator is moved to one of the mechanical limits and the count is reset.
The miscounts can be for various reasons, the most common is the same noise induced by the motor.
I solved them using UTP cable for the sensor and an optocoupler with a separate power supply from the positioner.
Every many days I have 1 bad pulse or 2 at most
This is the method that I use as well. In my case it works out to about 10 clicks per degree.
It's quite simple.
Suppose you've found 121W and your ASC1 says the count is 0320.
Now, bump the dish to the west until you find a peak signal from 123W. Check the ASC1 and the count is, say, 0335.
So 2 degrees was covered in 15 pulses on your system.
To find 125W - another 2 degrees west - you don't need to hunt around. Just move the dish by another 15 pulses to 0350 and you'll have it.
This is especially useful when there is a big gap between satellites and you have no idea where your dish is pointing in the sky.
You'll notice it's not uniform across the sky. At the center of the arc, you'll see a smaller number of pulses per degree than at the edge, but it's not a huge difference.
To test it out, write down your dish positions and the actuator counts and compare the gaps.
Similar threads
- Locked
