With a proto board and some skills, you could build a serial system with
a total cost around US$30, small enough to not even need a rail support.
You could also try to hang on the I2C iface of your mainboard and add
you own devices, but if you're not so much into electronics... Go the
Arduino way; readily available, cheap as chips and infinite expansion
boards.
Ralph Becker-Szendy escribis:
For one of my OpenBSD machines, I need to be able to measure a few
analog voltages, and act on them in a control process. The requirements
are quite simple compared to typical data acquisition: I absolutely
need two voltage inputs, either 0-20V or 0-100mV; doesn't have to be
differential, acquisition can be slow (1s is fine), and resolution can
be as small as 10-12 bits (1% accuracy is more than good enough). A few
extra input channels, more accuracy/resolution, and a few digital IOs
wouldn't hurt, but are not necessary. DIN rail mounting and connection
breakout would be nice, but can be improvised.
On the software side, there will be OpenBSD, with ad-hoc monitoring and
control scripts. With a little programming and script-writing, I can
adapt anything that the OS can reasonably access.
Now come the issues: I can't use PCI cards, only external units, most
likely connected via USB (as Ethernet and serial are expensive or rare).
And it needs to have some software support under OpenBSD - a Windows-
or Linux-only solution doesn't work. And this application is not worth
spending thousands of $$$. For Windows and LabView, solutions are easy
to find (for example EMant300, DAQPodMX, a variety of Omega products).
Does anyone now of a solution that would work with OpenBSD?