Re: [CAD_CAM_EDM_DRO] Another EDM Approach

Ian Wright wrote:

> ----- Original Message -----
> From: "Hugh Prescott" <hugh@...>
> >
> > Grab any blown PC powersupply and you can salvage at least one toroid,
> somes
> > times three from it that will make a nice current transformer.
> >
>
> Hi,
>
> I'm a little bit puzzled how this could work on an EDM for current sensing
> as it seems normal to use a DC voltage. Granted, when the thing is actually
> sparking, the voltage/current will be chopped but, at times other than this,
> surely, the coil method won't sense anything.

You are absolutely right, that this will not work in an R-C EMC setup. it
will fail to detect a short. I was suggesting this for the EDM system with
an electronic pulser. This will allow the current transformer to reset between

pulses, and it will sense current when the pulser turns on the current, even if
the
electrode is shorted.

> As someone else mentioned, the
> circuit which was published in the Model Engineer magazine some time ago,
> and which formed the basis of my first EDM experiments, used a resistive
> divider across the main spark current lines to feed current to the base of a
> transistor, the output of which was then used to flip the direction input of
> an SAA1024 stepper controller chip feeding the electrode positioning
> stepper. So, when the macine was first switched on, the stepper drove the
> electrode into the work (a simple 555 oscillator circuit fed the 'step'
> input of the SAA1024) until the electrode made contact. This then shorted
> across the resistive divider and grounded the base of the transistor which
> reversed the direction of the drive motor so raising the electrode again. As
> the contact broke, sparks were produced which ate away at the workpiece
> until the current level in the resistive divider fell enough to allow the
> transistor base to be pulled high again. The circuit was a bit crude as it
> didn't actually monitor the current in the electrode/work and so there was
> no real control over the quality of the cut. It also suffered from the
> problem of blowing the transistor regularly as the sparking produced a lot
> of RF spikes which seemed to easily get past the zener diode voltage limiter
> across the base/emitter junction.

This is why i was suggesting the current transformer, as a serious pulsed EDM
system might be a lot tougher on a sense transistor!

> When the circuit was working, it did a
> fairly good job of tap removal etc. but just wasn't refined enough to do any
> really useful work.

This mUst be EXTREMELY slow. The pulsed EDM systems run 10,000 sparks
a second and more! backing up the electrode after every spark must take
hours to make even a thousandth of an inch progress.

>
> This system of hitting the work and backing up the electrode would be no use
> whatever on a wire EDM machine where the reversal of direction would have to
> be on two axes simultaneously in the exact direcvtion they had just moved
> and so it is imperative we find a reliable method of current sensing which
> will work on both DC and chopped DC and will tolerate large, fast voltage
> spikes and which we can then use to control the speed of feed of the CNC
> part.

The current transformer seems the best approach, assuming a pulser circuit
is to be used.

Jon