19 Jun
2013
19 Jun
'13
11:30 a.m.
FWIW, the choice of "R" *may* be short for "radix" (point) which is
what
that dot means to mathematicians. Often called the "decimal point" when
we're speaking about base 10. See: https://en.wikipedia.org/wiki/Radix_point
On Tue, Jun 18, 2013 at 4:15 PM, John Wilson <wilson at dbit.com> wrote:
On Tue, Jun 18, 2013 at 07:33:51PM +0100, Tony Duell
wrote:
The EIA one has a pair of 50 pin filters mounted
ona sub-panel that fits
behind the normal distribution panel. There's a whort ribbon cable that
links one side of the device to the header on the distribution panel PCB,
the cable from teh Unibus card goes on the other side.
Same deal on the 20 mA, except that they're only 40 pins (which figures --
it really is possible to have *less* modem control than a DZ11A, i.e. none
whatever, and no "reader run" either for that matter).
Bew warned that the short cable from fiter to
distribution panel has a
strange conencotr on one end that swaps the rows of pins.
I wish you'd warned me that in 1985!!! I spent weeks tearing my hair out
(and bugging the guys at Heffrons for different cables) before I gave up
and just bypassed the filter.
That way, the
filer can be straight through without swaping anything (think aobut it).
Yeah I thought about it a lot. For a brief second (in 1985 I mean) I
convinced myself that it mattered which side of the cable you crimped the
connector on (OK, I was young) but my experiments failed.
AFAIK it's just 50 feethrough capacitors (one
for each pin) to the metal
casing and thus to mains ground. The valuse of each capcitor is goign to
be fairly low -- a few hundred pF? If I can find one I wil ltry to
measure it.
If this happened to be convenient I'd *really* appreciate it please!
No sweat if not ... a few hundred pF sounds like a decent guess, and so
far the circuit is all experimentation anyway (seemingly non-critical
part values have to be just right to work past a few hundred BPS, and I
still haven't quite hit 19.2K).
With that sort of value, stray capcitance will
affect the measurement,
and most DMM capcitance ranges aren't that good anyway.
And this thing is fed through unshielded ribbon cable so there's tons of
stray capacitance in real life too.
I believe the only reason for the filters was to
remove RF noise from the
interface cables to meet FCC requirements (part 15?)
I'm not equipped to measure the RFI but I like to design things so that
they *should* keep it minimal. Hopefully. I just read in QST about a guy
in Florida getting busted by the FCC because his well pump was putting out
a ton of MF, somehow. So this may be unwinnable!
John Wilson
D Bit