Archive of the former Yahoo!Groups mailing list: Homebrew PCBs

previous by date index next by date
previous in topic topic list next in topic

Subject: Re: EDM process for isolation milling?

From: "curt_rxr" <cwrxr@...>
Date: 2006-10-03

--- In Homebrew_PCBs@yahoogroups.com, "lcdpublishing"
<lcdpublishing@...> wrote:
>
> Hi guys,
>
> About a year ago or so there was some talk of using EDM for
> isolation milling and such. Stefan was tinkering a bit with it and
> another fellow did some heavy, indepth work in the process,
> including making a power supply to do the grunt work.
>
> I remember the initial approach was to use a bitmap image for
> processing of the machining. There was a good reason why this
> approach was used and I don't recall why - does anyone remember?
>
> Last night I was giving thought to this process. The one problem I
> see with EDM and isolation milling is keeping the PCB conductive.
> For example, if the output from the isolation milling routine
> creates and island in the middle of the board - then further
> machining needs to be done in that island area, it would no longer
> be conductive. So I suspect that is part of the problem with using
> vector data as opposed to bitmap data.
>
> Chris
>


Sure thing Chris,

I started an EDM project about the same time the list was started.

The reasons for choosing raster instead of vector were twofold.
Firstly, I was trying to adapt a printer and use the built in PCL 5
routines to "print" the layout directly to the circuit boards.
Secondarily, as you have noted it's the easier way to maintain a
ground return path.

I ended up custom building a stepper driven XY table as the printer
was incapable of the needed accuracy and the built in software was
dependant on a parallel row of jets in a cartridge. While printing
one dot after another was possible, the speed was glacial. Even with
the XY table the speed was limited by the resolution needed for fine
pitch packages. Using high speed IGBTs for the pulse circuitry and
incorperating breakdown voltage control of the electrode gap. the best
speed I was able to obtain for a an 8 inch by 11 inch board was 45
minutes.

Another system that might be faster would the vector or pen plotter
approach with a ground return incorperated into a dielectric pipe
surrounding the electrode. The return path would travel with the
electrode.

With the advent of the Volkan direct resist printing system there no
reason to continue with EDM for PCBs since you can print and etch
multiple copies of your layout in less time than you can burn one.

There are several folks in Europe continuing the EDM work so you may
want to look on other lists.

Curt Richards