On 27 Dec 2010, at 17:22, Simao Cardoso wrote:
>
> Simon Gornall wrote:
> > The SSR's will arrive in January next year, so perhaps I'll have a
> > chance to work on it then.
> > Apart from the chemistry, you're still looking at the high-hundreds
> > for a reverse-pulse-plating machine if you DIY it. I can't see the
> > price being much more "affordable" than that if it's sold
> > commercially.
>
> Hi, Which SSR's are you using? Can they do 0.5ms pulses? Or operate at
> 1V or less? The turn on time and the minimum voltage to operate seems
> out off necessary range. Most the datasheets i shaw for 10-50A DC SSR's
> had 10ms maximum turn on time, and 3V minimum required voltage to
> operate. There was some at sub-0.5ms turn on but for 1A or so. Just
> wondering did you checked if the SSR's you ordered fit the purpose?
>
Yes, I checked :)
The ones I ordered were
http://www.futurlec.com/Relays/SSRDC50V80A.shtml - these are "cheap" ($28 each) and can switch on/off within 0.5ms. I've never seen an SSR with a minimum load voltage - they have minimum control voltages but 3v is pretty ideal for that, considering it'll be driven by a microcontroller. Technically, I suppose it takes 0.5ms for the switch-on and another 0.5ms for the switch-off, but I'm guessing there's a wave-function on both on/off, so the time that the pulse is actually 'on' will be ~0.5ms, where 'on' is defined as > 0.5 x peak pulse voltage.
If it turns out that the pulses aren't good enough (0.5ms is pretty much at the limit of these devices), the ones I initially saw can do 20kHz (0.05ms) but these are more expensive ($63 each).
>
> Pulse plating documentation refers the value of a good square wave so
> fast turn on/off is recommended. Mine idea of a plating driver was
> started by a crazy wish of a electronically regenerated etchant, was to
> test chemistries than could fit this desire, were very high current
> densities define the efficiency. Fitting both purposes became not so
> simple.
> But any person with it's feet on the ground would just need 2 MOSFETs, 2
> PSUs, a dual driver and a MCU with PWM to archive it. STP80NF55 (one for
> direct other for reverse) and any dual 10V 2A driver gives you around
> 100ns for less money than one SSR.
>
I guess my feet are up in the air then :) I'm far more comfortable in the digital electronics world than the discrete components sphere. What I know about MOSFETs can be written down on the back of a postage stamp :)
I was using SSRs to isolate the two circuits, and make sure there was no current flowing between them (since both anode and cathode are connected to + and - on different PSU circuits).If MOSFETs can do that, cool :)
Simon
[Non-text portions of this message have been removed]