PWM offset voltage?
Posted: Tue Aug 30, 2016 3:57 am
As noted elsewhere, I'm trying to use the FG085 as a servo testing tool. It actually works very well - or at least has a great set of features for that. It will do everything a high end servo tester will do, and more, for not a lot more money.
However, I have a problem with a winch servo, which rotates 4 full rotations from end to end. This servo came off my bench, and isn't very good - but it works fine when attached to a receiver. However, when plugged into the FG085 it goes into continuous rotation mode. This normally means the servo is shot - except that this one works on an Rx.
The only difference between the two signals is that the FG085 base signal is slightly negative (- about 0.15 volts). The PWM modes don't seem to have any way to set the offset, and changing the offset in CW mode doesn't change it. That may be the issue, but I can't really tell without being able to offset it.
This could be an issue with my build - I've noticed that the amplitude and offset voltages tend to be inaccurate, off by as much as 10%, but not in a predictable way. I couldn't find any documentation on expected accuracy, and I can deal with it if it really matters, so I haven't worried about it. But if so, a suggested fix would be nice.
However, I have a problem with a winch servo, which rotates 4 full rotations from end to end. This servo came off my bench, and isn't very good - but it works fine when attached to a receiver. However, when plugged into the FG085 it goes into continuous rotation mode. This normally means the servo is shot - except that this one works on an Rx.
The only difference between the two signals is that the FG085 base signal is slightly negative (- about 0.15 volts). The PWM modes don't seem to have any way to set the offset, and changing the offset in CW mode doesn't change it. That may be the issue, but I can't really tell without being able to offset it.
This could be an issue with my build - I've noticed that the amplitude and offset voltages tend to be inaccurate, off by as much as 10%, but not in a predictable way. I couldn't find any documentation on expected accuracy, and I can deal with it if it really matters, so I haven't worried about it. But if so, a suggested fix would be nice.