Hi Gregory,
It's all good. Everything's working now. My exacto knife fix for the
solder apparently took care of things. I rebuilt the pi's image and got
it working great.
1. Configure cups to your printer (mine's a Brother DCP-L2550DW, using
cups-filter driver) and set the defaults to your liking (mine are):
US Letter
2-Sided On (Portrait)
2. Print a test page (better to know than to wonder, spoken from recent
experience).
3. Change two of the settings (ssh into the pi, cd /root/config):
echo -n "CR" > auto_linefeed
echo -n "ANSI-4" > page_size
4. reboot the pi (some setting changes don't require them, others seem
to, better safe than sorry)...
5. Test by llisting a simple listing in BASIC:
LLIST
6. press F8 to get back to the menu and press enter to test screen print
(not sure if any graphics will print, but the text works fine).
See the attached image (or look at the scrubbed attachment in the
mailing list archive).
Absolutely uh-mazing :).
Will
On 3/17/24 10:48 PM, Gregory McGill wrote:
Adapter was made because there was a board error on the original
board. So this adapter fixes whatever was wrong. I'll see if I can get
the details for you.. If it turns out to be the adapter I could
probably get it replaced for you if need be or repaired.
On Sun, Mar 17, 2024, 6:47 PM Will Senn <[email protected]> wrote:
All,
Looks like it was a problem with the adapter. My theory that bits
2 and 3 were shorted (oh, that's right, not my theory, but
Steve's) was confirmed with a few more character test. So, I
posted in the retro printer board, sent several 'contact us' posts
to the retro folks (their site sucks), and then...
Full disclaimer, I am NOT a hardware guy!! But, I madly took that
adapter apart and then after doing both sides, saw where some
solder was bleeding over from one pin to another on the female
side... and a bit of exacto knife wizardry later and a bit of
worry that I'd destroy it, and voila:
10 RMM ABCLMNOLMNOLMNOPQRS\]^_\]^
20 RMM abclmnolmnolmnopqrs|}~|}~
30 RMM 0123<=>?<=0 -_=/_]?>/",<.>
<0 RMM /?!@#,-^..,-
10 REM ABCDEFGHIJKLMNOPQRSTUVWXYZ
20 REM abcdefghijklmnopqrstuvwxyz
30 REM 01234567890 -_=+[];:'",<.>
40 REM /?!@#$%^&*()
Yay!
Now, I can go about setting this goofy retroprinter up properly.
Thanks for the forbearance, assistance and sympathy.
Will
On 3/17/24 5:46 PM, Stephen Adolph wrote:
Bit 2 shorted to bit 3?
On Sunday, March 17, 2024, Will Senn <[email protected]> wrote:
I tried every config setting... twice or more :).
It's super consistent in that bits 2, and 3, counting from
zero, from the least significant bits, are having some kind
of issue (they aren't always, zero, or one, but they are
always the bits that are wrong and they are consistently
wrong (of the 40 chars or so I tested):
<0x7f> vs w
01111111
01110111
00001000 (b3 1->0)
\ vs T
01011100
01010100
00001000 (b3 1->0)
4 vs <
00110100
00111100
00001000 (b3 0->1)
5 vs =
00110101
00111101
00001000 (b3 0->1)
6 vs >
00110110
00111110
00001000 (b3 0->1)
7 vs ?
00110111
00111111
00001000 (b3 0->1)
8 vs <
00111000
00111100
00000100 (b3 0->1)
9 vs =
00111001 (b3 0->1)
00111101
00000100
Weird, huh? Anybody seen anything like it? Can I troubleshoot
it with a multimeter?
Will
On 3/17/24 2:13 PM, Will Senn wrote:
Yep, it's consistent. It took me a while to make some
progress on this. I tried redoing the Centronics side of the
cable, and here's my source vs what the pi sees:
10 PRINT "Hello, world!"
20 GOTO 10
10 PRMN\ "Lmllo, orll!"
20 OO\O 10
I'm not sure how to troubleshoot...
I found this in the retroprinter handbook:
Missing Characters or Repeated Characters:
This is generally because the equipment sending the printout
is using
a specific timing mechanism and not necessarily adopting the
correct
Centronics signal methods for acknowledgement of data.
We have added the following configuration options to help
address
this:
/root/config/handshaking
This allows you to specify how the handshaking is handled
between
the computer and the Retro-Printer. This can help overcome
issues
with lost characters or repeated characters when the equipment
misses the busy / acknowledge signals.
The parameter takes a value between 0 and 4.
0 = Busy On (for 5ms), Busy Off, Ack On (for signal time),
Ack Off
1 = Ack On (for signal time), Busy On, Ack Off, Busy Off
2 = Busy On (for 5ms), Ack On (for signal time), Busy Off,
Ack Off
3 = Ack On (for signal time), Ack Off, Busy On (for 5ms),
Busy Off
4 = Busy On and Ack On (for signal time), Ack Off and Busy Off
Default is 0
Any idea how the M100 handshakes?
Will
On 3/17/24 7:18 AM, Mike Stein wrote:
Is it consistent, i.e. do you always get the same garbled
output for a given file?
At a fast glance it looks like bits 2 and/or 3 are being
dropped; have you checked the computer to Pi cable and
connectors?
m
On Sun, Mar 17, 2024 at 2:14 AM Will Senn
<[email protected]> wrote:
I am finally coming back around to this. I bought a
retroprinter a year and half ago or so and shelved it
out of frustration. Now, I know a lot more about this
sorta stuff and so I pulled it out, updated the
software to latest and tried to get it working.
The PI prints a test page fine, but it won't print
anything I send it from the M100. After hours of
troubleshooting, it appears that whatever codes the pi
is sending aren't DMP-15, EPSON ESC/P or Plain Text
codes... When I do llist, I see the data coming across
to the retroprinter and have set up a file to capture,
but I can't find anything that will make sense of the
data.
Here's a sample:
10 PRMN\ "lmllo"
1= RMS\ORM
=0>NORM=0\O1>RMALR,M-,Q,,M-,C,,M,0-,C,,M,1->CL,M-=0>NM\\M
20 M=0
2= O=0>OOS]B<0
30 OOS]B<=
It looks like reasonably valid data and not complete
gibberish, but who am I to judge. Is it one of:
Epson ESC/P 9 Pin - didn't work, when I tried it
Epson ESC/P 24/48 Pin
HP Printer (PCL3 or PCL5)
HP Plotter (HP-GL)
IBM ProPrinter
Plain Text- didn't work, when I tried it
Postscript
Printronix-P Series
Printronix-S Series
Seiko QT-2100P
Siemens PT-88
Apple Image Writer II
Seiko STP
Star Micronics SP700
Tandy DMP-105- didn't work, when I tried it
Help and thank you.
