Re: Re: Servo Hum (Was: Two Q's...)

I believe that the gecko drive will always hunt between adjacent encoder
pulses. This can be minimized by the amp tuning that you are working on.
But hunting while a motor is idled can also be caused by the EMC. The
trick is to minimize both of these.

The EMC could be driving the motor one step each way if you have a
commanded position that is part way between two encoder pulses. The way
to fix the EMC side is to set DEADBAND to about 1/2 of the distance
between two pulses. Say for example that you are using the stock ini
values of 1000 pulses per unit and that units is set for inches. This
would mean that you can not get finer resolution that 0.001. I'd set
deadband to 0.0005 or a bit more so that when the command said go to
1.0013 it would find 1.001 and be satisfied. At that point it would no
longer spit out one step pulses in each direction.

If the display of the last digit not matching commanded position, bothers
you set the display to commanded rather than actual.

Hope this helps.

Ray

On Friday 09 August 2002 06:58 pm, you wrote:
> Message: 23
>    Date: Fri, 09 Aug 2002 23:43:55 -0000
>    From: "mariss92705" <mariss92705@...>
> Subject: Re: Servo Hum (Was: Two Q's...)
>
> Hi,
>
> A badly tuned servo can break out into oscillation and rapidly
> overheat the motor. Just short of that, it may "hum" for a second or
> so after any direction or load change.
>
> I don't think that is what's happening here though. This could be the
> result of poorly spaced step pulses. The motor faithfully tries to
> follow them; the resultant acceleration / deceleration causes the hum
> you hear. The tipoff is the abrupt change from no hum to hum at a
> certain speed.
>
> I don't know your CPU speed. If it is slow, you may wish to try it
> again with a faster PC if you have one.
>
> Mariss