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

I used to have an entry level LPKF, the machine cost US$11,500 and with all
the tooling and options it came in at US$19,200. It did a terrible job, and
was a royal pain to use. We ended up putting it in a closet as the worst
mistake ever and then sold it off a couple years later. I think we might
have done a lot better spending the money on an Excellon drill and a chem
system.

Here are the problems we had. Its been over 10 years, so they may have been
solved by now.
1. The machine was hideously slow, it took about 6-8 hours of milling time
to make a two layer 4"X6" pcb.
2. The main cutter, a triangle type of affair would wear out in about 2-3
hours of run time. At $30/bit that got really spendy.
3. The software was really difficult to use and it had a high learning
curve.
4. Tool changes were very tedious and alignment was a real challenge, as you
had to make cuts, then adjust, and repeat over and over.
5. The through hole process required such large vias that it was totally
impractical.
6. The syringe and vacuum for the silver paste was so combersome, we stopped
using it after a month.
7. The vacuum was so noisy, that we could not work in an adjacent lab due to
the noise.
8. The board alignment backer was again very marginal.
9. A tech had to be around to monitor is almost contiuously. As such the
pcb's we made with it were roughly double the cost of having them done in 3
days by a vendor.
10. The software had tool monitoring. However, you had to program in the run
time parameters by hand. If a tool ran too long, one could figure that the
resulting pcb would be full of shorts and opens.

It is possible we had a very early design of the LPKF. It had a lot of good
ideas coupled with poor implementation. It was almost as if someone designed
it in a garage, and then did zero testing before putting it on the market.

If the LPKF had been $3000-$5000 instead of close to $20,000, a lot of the
above issues might not be so bad. Once the novelty wore off, there was only
1 tech who felt it worthwhile to run, and then only on through hole boards
with large traces.

LPKF did have a more capable machine which sold for around $60,000. Looking
back, thats probably what we should have went for. It probably would have
worked much better.

Getting back to your question, if you solve the LPKF problems, $10K would
seem to make it a winner. If not, in order to get some decent sales, figure
on $3K-$5K with a bunch of advertising.

Thanks
Ron