C Band Positioner Question

[Robz]

Hi,


Is there a DisEqc 1.2 36 volt actuator positioner which will support the DisEqc 1.2 command '6E GoTo x.x'? This is a five byte data command used to position a motor. Typically the motor is an HH type where the data represents degrees. In my case however, I am using an actuator to position a 6 foot FortecStar prime focus dish. The polar mount will be modified to add a rotary encoder located at the axis of rotation. The encoder will be scaled to represent degrees, which is compatible with HH motors. The micro switches will no longer be used. The 5 byte 6E command should not to be confused with the satellite position index value,which is the 6B command GoTo nn; where nn is a 4 byte value. The command bytes can be found in ETELSAT “bus_spec.pdf page 17.”


Any comments are welcome.

 

[SatelliteAV]

Prototypes? Yes

In Production? No

Good luck with your project! Keep us updated.

 

[Anole]

I wish !

Are we talking about a linear actuator, also known as a jack, to move the dish?
The geometry of the mount does not appear to offer a linear translation between motor movement and degrees.

There are however, some actual H-H mounts (AJAK and Birdview come to mind) where the relationship is linear.
Depending on a lot of variables, the motor controller would have to permit calibration of degrees per count.
That is to say, a stock AJAK has so many counts per degree, and a Birdivew has a different number, depending on if it's equipped with an 8-magnet wheel, a 12, 24, or 32.
So, a controller would need to accept a user-supplied calibration factor to handle the different motors.

I've brought this up in the past, and been quietly ignored.
Either because it's assumed to translate into low sales, it is not understood, or they just think I'm crazy. -
Search the FTA Department (and sub-forums) for old threads.

 

[pendragon]

I've also pointed out that the mechanical vagaries of BUD HH motors and polar mounts with an encoder may be such that more than a simple calibration factor could be required for this to work to any level of precision. I have two HH motors on BUDs and some time ago I fit the motor angles (not orbital positions) to the GBOX pulse counts. The slopes weren't particularly constant over the range. I didn't keep the spreadsheet but it's pretty easy to for anyone with a HH motor to do this.

 

[Robz]

Are we talking about a linear actuator, also known as a jack, to move the dish?
The geometry of the mount does not appear to offer a linear translation between motor movement and degrees.

There are however, some actual H-H mounts (AJAK and Birdview come to mind) where the relationship is linear.
Depending on a lot of variables, the motor controller would have to permit calibration of degrees per count.
That is to say, a stock AJAK has so many counts per degree, and a Birdivew has a different number, depending on if it's equipped with an 8-magnet wheel, a 12, 24, or 32.
So, a controller would need to accept a user-supplied calibration factor to handle the different motors.

I've brought this up in the past, and been quietly ignored.
Either because it's assumed to translate into low sales, it is not understood, or they just think I'm crazy. -
Search the FTA Department (and sub-forums) for old threads.

Thanks for your input.
Yes, I plan on using the linear actuator or jack screw and replace the reed switch for an optical rotary encoder placed at the east/west axis of rotation. The optical rotary encoder provides the advantage of a linear relationship. Only the encoder scaling task remains.
As you indicated, the relationship between the jack screw and east/west angular degrees is not linear and a curve fit method needs to be applied. The Lagrange interpolating polynomial provides a good curve fit to convert east/west axis degrees to jack screw pulse count. I have a program which calculates this for each satellite. Just move the dish until the positioner pulse count matches the program count. The problem however, is several satellites need to be located and identified before the curve fit will work. The Lagrange method will work from two known points, up to n points. Four known points, spaced uniformly east to west, seems to work accurately.
The optical encoder on the other hand can be calibrated without locating satellites and aligning the dish to track the Clark belt simplifies. A side benefit is part of the polar mount will be replaced. Two tapered roller bearings, a machined bearing housing, the optical rotary encoder and a single 0.5 inch stainless steel axle will replace the brass bushings and M12 bolts; providing far more stability and accuracy.
The optical encoder I plan on using provides 10000 pulses/revolution, providing high resolution and enough counts such that a PID controller will have enough error count to maintain position. A Silicon Laboratories C8051F35x series cpu will make a nice custom controller to achieve PID motion control and user input requirements.

 

[Robz]

I've also pointed out that the mechanical vagaries of BUD HH motors and polar mounts with an encoder may be such that more than a simple calibration factor could be required for this to work to any level of precision. I have two HH motors on BUDs and some time ago I fit the motor angles (not orbital positions) to the GBOX pulse counts. The slopes weren't particularly constant over the range. I didn't keep the spreadsheet but it's pretty easy to for anyone with a HH motor to do this.

Were you able to use the GBOX to pass through USALS commands?

 

[Lak7]

Yes, the GBox will pass USALS commands.

 

[Anole]

Couple of years back, we discussed using this as an inexpensive encoder.

 

[pendragon]

Yes, the GBox will pass USALS commands.

Unfortunately not. It will only accept commands to go to a particular user defined position number (1-99). This is sometimes referred to as DiSEqC 1.2 compliant. USALS requires commands that specify a motor angle. The GBOX does not recognize such commands.