Re: [CAD_CAM_EDM_DRO] Limit and Homing switches

Hi Drew,

Encoder parts (PCB's) are often built so they "could have" an index, but
the extra parts may not be there. Hang a 'scope on it, and look! If
TTL, you could probably get by using an LED and 220 resistor to +5V to
"see" the index (very narrow). Move slow, and see if you can see find
the index.

I wouldn't think the 5X of the Gecko would come into play, but rather
the resolution of the encoder. I know, let's ask Mariss!

Sorry I can't be of more help. I'm anxious to get a pair of G340's and
try out some servo's stuff myself!

Alan KM6VV


Drew Rogge wrote:

>
> Hi Alan,
>
> I was wondering if you could check me out on something. I'm currently
> using servo motors with Gecko G340's. The encoders have an index pulse
> (or at least a pin for one, should check that out I guess) and I was
> hoping to use some sort of an AND gate to generate the home signal. I
> think I may have a problem though because I'm using the G340's at a
> X5 multiplier. For some reason I think this means the index pulse MAY
> occur between "step pulses" from the computer. In other words it may
> be the case that the index pulse happens "between" step pulses. Has
> anyone else tried to use an index pulse with G340s in something other
> than X1 mode?
>
> Drew
>
> Alan Marconett KM6VV wrote:
> >
> > Hi Carlos,
> >
> > I have used a "phase" of the motor signal before in an RDF project. The
> > Optical switch must be narrow enough to only "gate" one phase signal or
> > index signal. So if the optical or mechanical switch was 3 steps wide,
> > one could position the switch so that the combined signal (sw AND index)
> > would now only be as wide as the index.
> >
> > The index only occurs once every rev of the motor, so in a simple case,
> > it's easy to "fit" it in 8000 steps! If you use a phase of the motor
> > (full step), then the switch must be a little smaller then 4 steps.
> > Half step? it would seem that you'd have 8 steps, but the resultant
> > width would be two steps.
> >
> > Encoders will give you 4 phases, if you gate: (A AND B AND sw). If the
> > switch is TOO WIDE (more then 4 phases), then you'd get multiple
> > "homes", which may or may not work for you. It appears it WOULD be
> > useable/helpful, if you always approach the home switch in the same
> > direction).
> >
> > I have some timing belt 9:1 reductions, and there is a disk with a
> > narrow slot on the output shaft, AND the motor shaft (two 3:1
> > reductions). They turn in opposite directions, and when they align,
> > there is a very short "index" pulse as a result (optical interrupter
> > sw).
> >
> > For a limit switch, I don't think it would be important (or even a good
> > idea) to use this. The limit doesn't have to be accurate, and the
> > additional logic would degrade the "fail safe" features of just switches
> > in series (N/C). For home switches, the extra TTL logic would not be as
> > much of a problem, IMO.
> >
> > HTH
> >
> > Alan KM6VV
> >
> > Carlos Guillermo wrote:
> > >
> > > Hi Alan -
> > >
> > > The idea of AND'ing a low-repeatability switch with the encoder
> > > index pulse sounds like a great low-cost method of finding a
> > > precise home. Is it really that simple? You just look for the
> > > AND'ed signal and set that as home? Wouldn't you have to
> > > calibrate the alignment of the index pulse to the homing/limit
> > > switch? Or, can you just have, a short ON-time of the coarse
> > > switch, which would necessarily be less than the axial travel of
> > > one revolution of the encoder? Could you elaborate? I'd like to
> > > leave room for something like this on my interface PCB, if it's
> > > simple enough.
> > >
> > > TIA,
> > >
> > > Carlos Guillermo
> > > VERVE Engineering & Design
>
> --
> Drew Rogge
> drew@...