Re: Alternative IO cards

["Peter Cook" <kf2qd-at-terraworld.net>]

Printer sharing devices are not the way to go - REAL TIME - the farther the
device is from the computer (network connections and such) the lees real
time the control gets. This is why there are a number of PCI and ISA cards
out there for motion control. Real time control requires that the software
be garranteed access to the processor and that the real time program has
direct control over its hardware.

They don't design those motion control cards just to be expensive and out of
reach to many of us who are just doing this on a hobby level. The more
"drivers" and non-processor busses you try to access the less likely you are
able to have real time control.

Pete

Subject: Re: Alternative IO cards


>
> Hi,
>
> I'm fairly new to this list, so I don't understand everything that goes
> past, but I'm learning.  On the other hand, I understand computer
> technology quite well.
>
> ISA is dying fast.  New motherboards supporting ISA are getting harder and
> harder to find.
>
> Parallel ports are dying slowly.  They will probably be available on
> motherboards for the next 5 years or so.
>
> What's left?  Custom boards, USB interfaces and Network interfaces.
> Custom boards tend to be expensive, by their very nature.  USB interfaces
> tend to have latency issues (protocol not designed with 'real-time'
> applications in mind).  Network interfaces may be what is needed.
>
> Has anyone tried any of the network devices for sharing parallel printers?
> These are available with 1-3 parallel ports, from HP and Intel, among
> others.  They are fairly inexpensive (about $100-$250), quite small, and
> VERY reliable.  You plug in a 10Base-T network, power, and the parallel
> device.  What I don't know is if they will work with stepper motor
> controllers.  I do know that some of them have speed settings for the
> parallel ports, so that they don't outrun the printer.  This would be
> quite helpful for CNC controllers.  Will it work?
>
> The other thing that is needed is a 'parallel protector' that protects the
> parallel interface from shorts, overvoltage and undervoltage, in the
> controlled device.  Under normal conditions, it wouldn't do much, but
> controller boards fail, and we don't need a voltage spike taking out
> hard-to-find (in a few years) parallel interfaces.  Has anyone seen or
> tried this?
>
> Peter Ashford
>
>
>