After using the focuser for a time I found
the small stepper motor had insufficient torque to turn the threaded
shafts on the coldest nights.
I have been planning the filter wheel for a while and the focuser
would have to cope with the added weight too.
The backlash in the gearbox was akward to manually compensate
for by always moving the focuser in the same final direction.
A dc motor and encoder to provide feedback was chosen.
![]() |

The Mechanics'
I have changed the reduction
of the gearbox to 250:1 and replaced the stepper motor with a
dc motor.
The Electronics'
A miniture rotary encoder provides
500 pulses per rev.
An Atmel 80c2051 micro-controller is instructed via a rs232 link
to a pc running a simple focus application.
Focus can be changed by either clicking on the in/out buttons
or by typing in a position in the blank window next to the target
position.
The control buttons set the amount of focus change for each mouse
click.
The bottom blank window shows the raw rs232 comms to / from the
focuser. Helped debugging only.
Focusing with the modified head.
Focusing is much the same, only
slightly faster, and my desk is less cluttered having removed
the hand controller.
The encoder gives 0.0025mm resolution and is directly mounted
on the shaft so that the gearbox backlash is automatically removed.
This has been a major improvement.
When imaging at f6.3 on most nights I can detect correct focus
to with in about +- 30 counts which equates to +-0.07mm.
Further changes to the focus application will be to add RGB focus
offset buttons for each filter and also to emulate one of the
auto-focuser protocols on the market to allow standard software
to provide auto-focus ability with my hardware. Maybe I get around
to adding temperature compensation too.