This long discussion seeks feedback on an application that requires precision tracking of a two axis telescope:
I’m working on the design of a pilot balloon (piball) theodolite. This is an optical device used to track a small helium balloon so as to determine wind speeds and headings at various altitudes above the ground.
Here is a classic theodolite design:
https://www.pilotballoon.com/ml-474gm.htm
The device consists of a gimbaled right angle telescope with horizontal and vertical protractors. The horizontal protractor reads the compass heading, while the vertical protractor reads the angle of elevation to the piball. Serious instruments employ a vernier or worm gear micrometer screw for fine adjustment of the telescope, with resolution to a tenth of a degree. This allows the operator to precisely keep the piball centered in the telescope crosshairs. All this precision comes at price. Classic mechanical instruments start at about $8,000USA. Newer electronic readers run up to $20,000.
While a lower cost surveyor’s transit can perform many of these functions, limits on its elevation angle and its straight-through telescope limit utility. This is especially true when light winds cause the piball to rise directly above the theodolite.
My proposal: I would like to build an instrument that employs a motorized drive operated by a joystick. For my application 0.5 degree resolution is adequate.
One approach is to drive each axis with a stepper motor. Surplus vendors are currently marketing gear reduction stepper motors that provide up to 3600 steps for a complete revolution. This would provide 1/10 of a degree resolution. Each telescope axis could be directly coupled to such a motor.
Rather than depending on a mechanical protractor, each axis position would be read with an optical encoder. Here is one possible encoder. It breaks the circle into a 10 bit (1024) segments with a resulting resolution of about 0.35 degrees:
**broken link removed**
Tracking the piball generally requires small incremental angle changes. But occasionally, the horizontal axis has to be rapidly shifted by about 180 degrees. This can occur if, shortly after release, the piball moves in an opposite heading from that expected. To accomodate this requirement traditional instruments can be physically rotated, by hand, to the new heading.
Classic theodolites treat the protractors as worm driven gears. The micrometer drive is a worm gear that can be snapped into position to lock it against the protractor, to allow very precise motion. This method results in expensive, machined gear systems. A less expensive drive mechanism is needed.
Now to my questions:
Can readers suggest other drive alternatives?
Are there other options for measuring the angles other than using encoders?
I’ve considered the following alternative: Imagine a rifle stock with a telescopic sight. The unit is aimed at the piball and the trigger is pulled each time good aim is obtained. The rifle stock would contain a multi-axis accelerometer to read heading and elevation. Readings are taken by a computer that generates a table of output values.
I’ve looked at electronic compasses and electronic inclinometers for this application. There are a number of incline units which would work well for elevation measurements. The compass is a more difficult nut to crack. I’m not convinced that electronic sensing compasses have the repeatability and linearity this application requires. Most compass units also require the electronics to remain parallel to the horizon.
Comments are appreciated.
I’m working on the design of a pilot balloon (piball) theodolite. This is an optical device used to track a small helium balloon so as to determine wind speeds and headings at various altitudes above the ground.
Here is a classic theodolite design:
https://www.pilotballoon.com/ml-474gm.htm
The device consists of a gimbaled right angle telescope with horizontal and vertical protractors. The horizontal protractor reads the compass heading, while the vertical protractor reads the angle of elevation to the piball. Serious instruments employ a vernier or worm gear micrometer screw for fine adjustment of the telescope, with resolution to a tenth of a degree. This allows the operator to precisely keep the piball centered in the telescope crosshairs. All this precision comes at price. Classic mechanical instruments start at about $8,000USA. Newer electronic readers run up to $20,000.
While a lower cost surveyor’s transit can perform many of these functions, limits on its elevation angle and its straight-through telescope limit utility. This is especially true when light winds cause the piball to rise directly above the theodolite.
My proposal: I would like to build an instrument that employs a motorized drive operated by a joystick. For my application 0.5 degree resolution is adequate.
One approach is to drive each axis with a stepper motor. Surplus vendors are currently marketing gear reduction stepper motors that provide up to 3600 steps for a complete revolution. This would provide 1/10 of a degree resolution. Each telescope axis could be directly coupled to such a motor.
Rather than depending on a mechanical protractor, each axis position would be read with an optical encoder. Here is one possible encoder. It breaks the circle into a 10 bit (1024) segments with a resulting resolution of about 0.35 degrees:
**broken link removed**
Tracking the piball generally requires small incremental angle changes. But occasionally, the horizontal axis has to be rapidly shifted by about 180 degrees. This can occur if, shortly after release, the piball moves in an opposite heading from that expected. To accomodate this requirement traditional instruments can be physically rotated, by hand, to the new heading.
Classic theodolites treat the protractors as worm driven gears. The micrometer drive is a worm gear that can be snapped into position to lock it against the protractor, to allow very precise motion. This method results in expensive, machined gear systems. A less expensive drive mechanism is needed.
Now to my questions:
Can readers suggest other drive alternatives?
Are there other options for measuring the angles other than using encoders?
I’ve considered the following alternative: Imagine a rifle stock with a telescopic sight. The unit is aimed at the piball and the trigger is pulled each time good aim is obtained. The rifle stock would contain a multi-axis accelerometer to read heading and elevation. Readings are taken by a computer that generates a table of output values.
I’ve looked at electronic compasses and electronic inclinometers for this application. There are a number of incline units which would work well for elevation measurements. The compass is a more difficult nut to crack. I’m not convinced that electronic sensing compasses have the repeatability and linearity this application requires. Most compass units also require the electronics to remain parallel to the horizon.
Comments are appreciated.
