[time-nuts] Book: Phase Noise in Signal Sources by Robins.
wa1raj at arrl.net
Mon Apr 25 21:38:04 EDT 2005
> I do like the idea of using the RS-232 dongles whenever/wherever they are
> adequate for the job.
While I've found the ISA (~$20) and PCI (>$100) cards from NI a good
value and supported quite well under both Linux and Windows, I'm also a
fan of the serial-to-488 converters. If you don't need the raw bus
bandwidth (as has been mentioned for doing binary mode transfers from a
counter at measurement rate), you can go this serial route.
I've been using a IOTech Micro-488A converter with success. These are
old and only support a maximum baud rate of 57K, but I was able to pick
up my unit for ~$20 on EBay. You can download the full manual for these
beasts from the IOTech web site. Be careful with what unit you buy -
there's a Serial-488 adapter that was intended for interfacing a serial
printer to 488 devices, and it only supports a subset of the bus
commands the Micro-488A controller handles.
Even the serial speed restrictions are not that big of a deal if the
instruments you are controlling are doing a little math, as is normally
the case when I'm taking data from the 5370A.
Perhaps my favorite thing about using a serial-to-488 converter is that
you can place a terminal server in the oscillator rack and use that over
the network for all the instrument control. Whether it's a serial or 488
based instrument, this solves the interconnect problem and reduces the
heat load by a couple computers worth of power in the area I'm trying to
keep at constant temperature. It also means there's finally enough
serial ports available to a single computer to be useful in logging and
I'll make available some trivial Perl scripts for using this IOTech
controller if anyone has interest. They really are quite easy to use.
More information about the time-nuts