Re: Interface. Gauging interest.

Dennis,

Answers inserted below...

>In CCED, "Dennis Schmitz" <denschmitz@...> wrote:
>1. How much would you pay for a USB I/O box to control servos
>(requiring Gecko or equiv to drive the motors)?

Don't need ot want 6 axis, 3 or 4 is fine. For that, with some
reasonable IO, say 16 bits; 50 bucks, maybe 75...(no enclosutre
desired, just a bare board with a USB conn and .1 headers for IO and
step/dirs.)
See www.deskCNC.com, www.cncdudez.com. And do a yahoo seach on
CNCplayer for an "embedded DOS" version. See gloomy andy's website
for another. Visit any of dozens of CNC engraver suppliers; vision,
roland, xenetech, suregrave, etc. for even more examples...

The problem is that these are all using a proprietary CNC interface.

If you make it work with Mach3 or preferably Mach4, I'll still want
to pay less than 80 bucks, but will buy at least 1000 in the next 6
months. And likely a LOT more, if you can get the price/features
balance right.

>2. How much would you pay for this USB I/O box if it had all the
>drive electronics?

Not interested in this option. Too many drives already out there,
and we're already a sys integrator...
(Though we would probably end up putting your design into an
integrated single board solution ourselves. Your licensing would
need to allow that.)

>3. How much would you pay for a box slightly larger with a built in
>computer running EMC2 and communicating with a shop Ethernet
>network?

Not interested in this one either. Have absolutely NO interest in
EMC at this point. Though once I was pretty "taken" by it. The
extensive configurability of M3(by OEM's OR individual users) just
blows it--EMC--out of consideration, IMO.

The idea of an embedded PC is okay, but then you're in the "PC"
support biz. Do you wanna be there? (would depend on OS and its
visibility in the final app.)

I think you'll find that it IS "easy" to get something onto the end
of a USB cable that will do "motor control". I think you'll find it
FAR harder to give it the capability it needs to be truly useful for
CNC.

IMO the "hole" in the market is something more like the DeskCNC
controller than the Gecko G100. Give me something CHEAP, that is NOT
tied to certain software--unless it's Mach3 or 4--, and has
reasonable IO, as mentioned above. Do that for 50-75 bucks and
you'll have a winner. (BTW, 50-75 bucks is surely possible from the
hardware cost factor; you just have to believe in the market
viability to sell enough to re-cover your software development and
support costs)

If a person wants more capability the G100 already fills that need
at a cost and with proven developers and products behind it that
will be hard to beat. (I'd love to have a simplified G100! While it
is PERFECTLY designed for its intended market, IMO; it has FAR too
many bells/whistles for 3-4axis cnc control that can be sold to the
masses.)

If a person wants "less" than what the g100 offers, the links and
companies mentioned already fulfill that.

What we truly need is a usb EQUIVALENT of the parallel port.
Ideally, a USB equivalent of TWO p-ports. Just do that, and
the "problem" of no pports in PC's and protection OF those ports by
MS Windows architecture goes away. Then folks can use any of the
MANY pc-based pport solutions which already exist...

Do have a look at the DeskCNC API, and rest assured that there is
far more than meets the eye in getting GOOD RELIABLE motion control
across a USB cable. Especially if you're going to be performing at
the current level of sophistication; where constant contouring is
happening at many hundreds of inches per minute.

Personally, I'd be pretty interested in a DeskCNC type unit married
to Mach3 or 4. That's what my beehives were intended to be, but the
software and hardware development changed direction. Even 25K SPS
would be "enough" to go after a HUGE market that is not currently
being addressed. If you're SERIOUS about doing this, please contact
me offlist.

Ballendo

P.S. Having been in this DIY-CNC thing so long it's funny to see the
once powerful solution of using your PC to replace the embedded
proprietary electronics of the machines; come full circle back to
the point of needing to re-address an external "smart" device/black
box. It sucks really. As Art F said to me once, There's all this
power in the PC, but I can't get it out of the box! ---to a
machine