Re: [CAD_CAM_EDM_DRO] Interfacing EMC with Bandit drives

Jack Ensor wrote:

> The Bandit in discussion here is on a Bridgeport mill and I would
> like to hopefully replace the control circuitry with a PC while
> retaining the present Bandit drivers and thought that EMC may be the
> appropriate software for this.
>
> The Bandit driver uses a sense resistor for sensing current and
> controls it by varying the duty factor of the voltage pulse applied
> to the motor windings. The control board does put out a command
> coupled into the driver board via an opto-isolator to cut the holding
> current in half.
>
> I found that the resolution is .0002 in. per step and the rapid
> traverse is 60 ipm. This calculates out to a 5 Khz for the step
> input. It would seem to me that the chop frequency should be several
> times the max step frequency. I don't know why they used 5 Khz, it
> is very noisy.

Ah ha! The trick is, that when the machine is moving fast, the current
never builds to the limit value, therefore, the switching is disabled a
the higher feed rates. That's how they get away with the low freq.

> There are 5 input signals to the Bandit driver. The position signal
> input consists of 4 opto-isolators. A move signal is coupled to the
> 5th opto-isolator which terminates the hold mode. After the move
> signal is goes to zero, there is a 5 second delay via a 555 timer
> before it reduces the current to the hold level.
>
> The step signal sequence into the two channels is as follows:
> A B
> 01 01
> 01 10
> 00 10
> 10 10
> 10 00
> 10 01
> 00 01
> repeating beginning with 01 01
>
> Direction depends on which way you go through the above sequence.
>
> With this information, can EMC handle this task? If not, what other
> software might be appropriate?

Yes, for the most part, EMC can do it. EMC has been modified to generate
phase signals like the Bridgeport BOSS drivers use. I think it is the same
scheme, but it looks like the Bandit half-steps, while I think the BOSS
is full-steps, only. The code could be changed for that. But, it is VERY
easy to convert step/direction to phase signals. There are many chips
that can do this. I'd swear I had a drawing on my web pages of a circuit to
convert step/dir signals to full steps that only requires 2 chips, but I can't
find it now. To do half steps, it is just a tiny bit more complicated.
Controlling the hold mode logic with a timer chip would be trivial.

Jon