Re: USB CNC (was Interface. Gauging interest.)

> This topic has a life of its own.

And it's coming around full circle. This one got started May 12 (msg
87191), about the time the "wood duck" thing was winding down.

Apart from two early replies, it didn't take hold until the 15th.

The first few days of re: exchanges are good; anything of real
significance gets said then. After that the subjects get divergent
and unrelated and the thread winds down. Start to finish, good for
about a week.

This thread is a good case in point. The emphasis has been on:

1) USB connection.
2) Make a simpler, cheaper, better box at the end of the cable.
3) The parallel port is the devil's spawn.
4) The devil is Mother Theresa when compared to Windows.
5) I have some good 386 PCs, why can't "it" work with them?

Let me take them in turn:

1) Like the parallel port cable and firewire, USB is a short-haul
connection. Don't fixate on it. The long-haul choice is ethernet when
using generic PCs.

2) This one brings out the best and the worst in people. The best is
the person that originated the thread by saying "I see what's out
there but I think I can do it better." That is what innovation is all
about. He may very well do it if he has the talent and persistance.

The worst in people is "I want it to have $5,000 performance but I
want it to cost $29.99". Can't be done; you build stuff, it costs
money. You build complicated stuff it costs even more money.

Don't envy the guy that builds stuff and charges you $$$ for it. He's
not making a killing, believe me. You don't want his problems.

3) I agree and thank goodness for me. In it's twisted form today it
is no longer a real-time "window" into the CPU. It takes some
very talented individuals like Art to eek out even modest performance
from it in Windows.

It may be and it is slowly being exterminated. Darn shame in my
opinion because it is the last remaining non-serial PC port. Anything
serial trashes timing related outputs; it's the only port that had
even the pretense of being a real-time "window" into the PC. That and
it is the perfect $0 solution.

To see it being used right and see what it really can do, try CNCpro
or other good DOS-based programs; no comforting GUI though. Were DOS
still king I never would have bothered with developing the G-Rex; a
real-time 4-GHz 64-bit CPU beats an 8-bit 22MHz CPU everytime.

4) When I play poker, I play the cards I'm dealt. The cards I hold
have little MS logos on them and I do not argue with the dealer.
Philosophical arguments aside it's reality. Final market-share score:
MS 95, Apple 4.99, Linux 0.01.

5) The world moves on. Learn to deal with it; no one will ever write
an "it" for you.


Mariss



--- In CAD_CAM_EDM_DRO@yahoogroups.com, "skullworks" <skullworks@...>
wrote:

>
> This topic has a life of its own.
>
> May I be so bold as to ask that someone start a new group for this
> topic.
>
> I am an end user type - a kit builder, not electronics designer. I
> would love to see a sort of open source design come togather from

all

> your great ideas.
>
> As stated in a prior post PCI card options exist, but start around

$550

> for the Basic I/O PCI card. In addition you will need a breakout

board

> with powersupply and have to design the rest of the system, and
> drivers/software if the hardware you choose is not supported by

your G-

> code interpreter of choice.
>
> A simple 4 (5?)axis USB interface that can also monitor limits and

E-

> stop would be a great boon to this hobby. (4+ spindle control

outputs)

>
> Thanks to all.
>