In post #18 when you say one revolution is that of the motor shaft or the output shaft of the gearbox ? Is the 11 seconds one revolution, part of a revolution or many revolutions ? You can get some idea of the state of the commutator and brushes by connecting a DVM on a low ohms range to the motor terminals (The brushes.) with the motor disconnected from the controller. Note the resistance reading. Rotate the actual motor shaft (NOT the gearbox output.) a few degrees at a time and note the reading again when it the reading has settled. (The process of rotating the motor generates a voltage which changes the reading.) Continue doing this for a full revolution. The readings should all be about the same resistance. If they vary quite a lot then the motor probably needs replacing. You could try repacing the brushes if they are worn and cleaning the commutator before replacing the motor. Another test you can do is to power the motor from a stable voltage (of about half it's rated voltage.) and look at the output of the two encoder outputs with a dual channel oscilloscope. If the output is a square wave both channels should have a duty cycle of about 50% and have a phase relation of about 90 degrees. If the output is a sine wave it should be a reasonable shape sine wave again with a phase shift of about 90 degrees between channels. If the frequency is varying a lot try removing the motor from the machine in case there is something tight in the machine causing the speed variation. Note you will need to have the encoder part connected to the machine for these tests so the encoder part is powered.

Les.

 

Sounds like a mechanical issue. Could a shaft have got bent or misaligned? Has grease in the gearbox gone hard?

I don't think so to the first question; and no to the 2nd.

However one thing I noticed is that the pinion is just hammered onto the end of a grooved shaft and it does run a bit eccentrically when looking from end-on. Both motors are the same.

Another thing was the optical wheel was quite grubby with some dirt. Now clean.

The variation in motor speed is about 0.6%. The presume the encoder's job is to keep it absolutely constant.

 

In post #18 when you say one revolution is that of the motor shaft or the output shaft of the gearbox ? Is the 11 seconds one revolution, part of a revolution or many revolutions ?

11 sec/rev for the motor shaft, not the gearbox output shaft. The gearbox reduction ratio is 54:1.