On 9/17/2013 9:49 PM, John Wilson wrote:
On Tue, Sep 17, 2013 at 09:09:46PM -0700, Josh Dersch
wrote:
Sorry to reply to my own post here, I had a bit
more time to screw
with this this evening. Turns out it was a programming error
(actually a program entry error :)) that was (apparently) causing the
40mA current instead of 20mA -- the little "echo" program I coded up
wasn't properly waiting for data to come in before echoing it back,
so the 8/L was constantly sending 377 back across the loop, which I'm
assuming accounts for the difference?
With this corrected, the loop is registering ~23mA idle, dropping to
19-20mA when the PC attempts to send data. Can anyone comment on
whether this sounds reasonable? Anyone have any advice? (Getting
desperate here :)).
I may be misunderstanding you, but are you saying that
you're using one
combined loop for both TX and RX? That's the right way to do RTTY
(with the old high-voltage 60 mA loop and Baudot machines) but the wrong
way to connect an ASR33 to a DEC machine (unless the 8/L is an oddball).
Separate loops for each -- and the ASR33's RX is just a solenoid, and
the ASR33's keyboard is just a switch.
I have it wired as in Fig. 2 in the PDF here:
http://www.bb-elec.com/Products/Datasheets/r2_Website-Only_0812DS.pdf. I
had thought that this was two separate loops, but given the behavior I
described earlier (with the 40ma while sending) perhaps I'm misreading
things...
Anyway you definitely want to do early testing by sending BREAK.
~20 mA idle (yes 23 is close enough), 0 mA breaking (or anyway vastly
lower) since you're supposed to be opening the (normally closed) circuit.
Thanks -- I'll give that a shot tomorrow when I have some time.
FWIW, here's the relevant snippet of a multi-line RS232/20mA converter
I've been gradually slapping together in my copious spare time (for
emulating DZ11Cs with RocketPorts etc.):
http://www.dbit.com/wilson/convert1.pdf
It's pretty standard stuff (DEC and BB both use similar circuits). This
shows two lines' worth, minus the RS232/TTL stuff (which is obvious, and
connects to the optoisolators), and it's jumperable for active vs. passive.
The only thing (as I'm discovering) is that it takes some fiddling to get
pullup values that work well over a wide range of bit rates (I've had to
slightly tweak the values from this schematic, I forget to exactly what, to
get up to 9600 but still not quite to 19.2K).
No guarantees that it would work with the 8/L, but it should be within a
mile. Optoisolation is probably overkill for connecting through a short
cable to a known circuit (with known ground reference). Even the
current-limiting (the 33-ohm resistors) is probably a bit much since no
one's trying to feed you infinite current. I went the other way (ASR33
attached to RS232 printer port on a PDT-11/150) eons ago with neither.
Cool! I'm hoping that I can manage to get the adapter I already have to
work, but it's good to see other options.
Thanks,
Josh
John Wilson
D Bit