Re: [CAD_CAM_EDM_DRO] Re: Building an EDM - switching

turbulatordude wrote:

>>>I figured on using a trio of photo-flash caps, 120uF so they
>>>
>>>
>could
>
>
>>>be wired for 40uF, 80uF or 120uF.
>>>
>>>
>>>
>>>
>>If you are using a pulser, you don't need to vary the capacitor
>>
>>
>banks.
>
>
>
>
>It seems that work would be done by the current in the pulse.
>joules being the energy measurement.
>
>if the cap it 1uF or 100uF doesn't the amout of work performed
>change ?
>
>

As long as the capacitors are sufficient, they don't control the energy
per pulse, when
you use a pulser. The resistance of the transistors in the pulser, and
the duration of
the pulse (set by a timing circuit in the pulser) are what sets the
energy (multiplied by
the voltage, of course).

>It seems that by charging a cap directly thru a semiconductor with
>low resistance, the time constants would be so low as to not offer
>any benefits to one size cap or the other.
>
>I mean that the 100uF will not take 100 times longer to charge, but
>will offer signifigantly more energy for work.
>
>or, am I confusing the RC (charge/drain) with a circuit that offers
>control over the entire pulse and duty cycle ?
>
>

Yes. If you use a pulser, a pretty standard capacitor bank is charged
directly by a
transformer/rectifier to some desired voltage. This could be several
thousand uF
for a small EDM, more for heavy-duty machines. The pulser delivers
currents of
hundreds to thousands of amps, with the current primarily limited by the
transistor's resistance. The current is cut off after some set time,
somewhere in
the neighborhood of 10 uS, and then an off-time is controlled by another
timer,
this might be in the range of 100 uS. After this time expires, the
transistors
are turned on again, and after some time the discharge starts again. When
current flow is sensed, the on timer starts timing.

Jon