From: ard_at_p850ug1.demon.co.uk (Tony Duell)
>May I ask a very simple question. What machine are you using this with
(from the description below, I assume a PDP8). If it _is_ a PDP8, do you
have the reader control relay mod fitted correctly? If not, then you
_will_ drop characters I think.
My computer is an 8/L. I do not have the relay fitted at all. After a
careful examination of the schematic and manuals, this relay only forces TTY
power on, in LINE mode (this is redundant if the TTY is already on and
prepared to read a tape), and also forces 115V to the distributor trip coil
so that a tape will immediately start reading if present. Without the relay
a manual start is required by simply flipping the reader lever to "Start".
Everything gets up to speed after one or two characters of leader no matter
how the reader is initiated. No characters will be dropped since the program
will loop indefinitely at the usual KSF, JMP.-1 until a character is
assembled and sent to the TTY receiver card.
>That's very odd. The most obvious mechanical problem would be wear on the
ratchet wheel teeth in the reader. But that would cause problems once (or
n times) for every revolution of that wheel (and the drive sproket) --
Quite. That's why I'm scratching my head!
>So what you're saying is that it sends the correct number of chracters,
but some of them are corrupted to 0's?
Yes, when an error occurs while printing the test tape I can "hear" the
nonprinting character being sent to the typing unit (it "jumps" but nothing
appears on the paper) and the next consecutive char is then printed
correctly.
>Either your read pins are misaligned (they should come through the middle
of the tape holes --
check this), or the contacts are sometimes not making properly.
Agreed. They look centered to me (but what IS the definition of "centered"
i.e. at what point in the read cycle? The tape and pins do not seem
perfectly stationary with respect to one another during the read cycle.)
The manual shows that the sprocket wheel should be rotated forward to take
up the slack against the detent at which point the pins are centered in the
holes. IF I'm reading this right. Actually they show a "Gate Flash" that
looks like a hole on the side of the sprocket, which does not exist in my
reader.
>Try running the reader without tape, with the EOT sensor held down. It
should then send all 1's. Check that it does (use the PDP8 for this with
a similar program to yhe one you've written already, just get it to halt
if the character is not all 1's). Let it run for many minutes.
Excellent suggestion, thank you. I will try this right away.
>A really Heath Robbinson (Rube Goldberg?) trick
Who's Heath Robbinson? I always liked Rube Goldberg's inventions...
>3) focus a camera on the reader mechanism. Set the aperture appropriately
for the flashgun. Make the room dark, hold the shutter open on B, let it run
4) When the flash fires, close the shutter, turn everything off
5) Develop the film and hope that it shows you what the problem is.
-tony
Sounds awfully complicated. Perhaps I can do this with my Sony digital
camera if it has a sync input!
thanks for the tips
-Charles
Received on Sun Apr 13 2003 - 11:38:00 BST
This archive was generated by hypermail 2.3.0
: Fri Oct 10 2014 - 23:35:43 BST