On 15 Dec 2010, at 13:19, Simao Cardoso wrote:
>
>
>
> Simon, i looked into your descriptions. Reverse pulse can be really good
> but neither do you seem to full understand it either is a
> simple/complete answer.
>
Oh, I'll easily admit I don't fully understand everything I need to know for this :) Yet :)
> I am also building my driver, also with a avr,
> but with very low RDSon FETs (and fet drivers), using 4 psu's and 8 fets
> (multiple anodes and cathodes) for a tricky chemistry (big expectations
> in it). But for example i never figure out which voltage is appropriate
> for the reverse pulse.
>
My understanding was that you tend to vary either current or voltage in the pulses (so a constant-voltage, varying current supply would be ok). The CAT4101 has a wide range of voltage over which it will work (2-24v), so I'm covered, as long as a constant supply level at varying current is ok.
> You don't seem to realize it but it all goes
> around current densities on surface vesus hole. And the magic number is
> 3 (or higher) - 3 times shorter reverse pulse, 3 times much current. For
> _SLAPPING_ copper from surface and leave holes without touch, using same
> values for both is wasting energy.
>
Ok - I'd seen 2x rather than 3x, but ok. I was aware it ought to be different, that was why there are 15 CAT4101's in the forward direction and 30 of them in the reverse direction. I wasn't actually planning on running them at max output, and the current ought to be a function of the board area anyway, right ?
So yes, I was planning on 2x the current, 0.5x the pulse-width in the reverse direction. If you're saying 3x current and 0.33x the pulse-width, well, that's an easy scale-up :)
> You need at least 3V for plate pcb's
> nicely, but since the common setup uses additives to require voltage
> increase as current increases (without them current goes exponential
> above ~0.7V) there isn't document proof on which voltage to use (or
> known which psu to buy). Commercial systems say 12-20V capable but thats
> crazy it will electrolise water. Ohh and i am having big trouble on this
> so the avr changes voltage and current limits at cheap switched mode
> psu's.
>
Right - I was planning on using the 5v supply on an ATX PSU and see where that got me. If I get burning, I'll reduce the voltage/current. If I get nothing, I'll increase the voltage (dropping it from the 12v supply on the PSU down to 6, 8, 10 etc. using a simple voltage divider and big resistors) and see what I can do. It's a relatively small investment for me (apart from the chemistry from LPKF) because I have a lot of the stuff on-hand. I use CAT4101's in LED drivers, and I have a spare 50 or so of them lying around :)
Simon
[Non-text portions of this message have been removed]