About two decades ago, I was working on POS software for a company. At that time, most receipt printers used RS232 communication, and I remember a project where we created virtual printer based on communication data generated by the PC's RS232 port. The purpose was to test whether software outputs were accurate without actually using a printer.
That's was a cool idea, when I had that problem, I wrote a table of command translation between the matrix printer and ESC/P so I could just send the thing to a fast laser printer, It was very fast to make tests, but wasted paper.
I and my friend worked together to replace an old (1-pin with 9 colors ribbon) dot printer which was a part of scintigraphy thyroid scanner. this scanner was built as a tank and it prints as it scanning, the printer and the sensor mounted on one big moving arm and it prints directly on a paper laid down and held on a metal plate using adhesive tape , you really have to see it 😁🤦♂ The reason to do the app is that the stock ran out from compatible ink ribbons The problem was it doesn't output any digital signals to rely on and al what we got is analog sound generator producing a wave depending on the strength of the radiation detected by the sensor. My friend designed an electronic interface circuit to convert these waves to something our program can read via RS232, and I built the app to mimic the output of the printer in order to print it on a modern inkjet printer. Thanks Jim, you bring back so many memories
One way to slow it down to a fraction of a millisecond would be to use a counter, and then you add the delay for every 2nd or 3rd byte. A live human might not spot the difference. Basically this: int count = 0; ... if ((++count % 3) == 0) { delay(1); count=0; }
I would have thought converting the character set would be much more involved, this is very cool! I had a moment to think about it, and I think it will be 30 years this year since I last used a dot matrix printer. The sound as it tracks across the sheet and advances rows is very nostalgic.
In amongst my stuff I have the manual for the star NX1000 mono dot matrix printer, The manual includes some BASIC listings for testing the capabilities of the printer, including user defined characters. Maybe include support for sending “escape” codes (char 27) for enabling underline, sending data for user defined characters, etc
My printer at the time was Epson, and Epson printers had that ESC/P print control. The name actually comes from Epson Standard Code for Printers because Epson made it. I hope to update this in future to add ESC/P print control like underline and even graphics mode.
Depends if you already have a robust TSR framework to build on, there's just a few int 21 calls to intercept and you'd have to parse pathnames for 3Dh. Making a TSR that knows when it can take over the system is the tricky part, and it would need to be quite big for the graphics buffer.
I think an easier way would be to redirect printing to a file (many DOS word processors can already do that) and then use this to "print" it to the screen.
I need a better sound setup. I'm recording on the same system where I'm running FreeDOS, and I think that causes issues sometimes. Pros use a second system to record.
I always liked that Epson 9-pin printer - it saw me through university. But to get high quality output, like when printing physics reports from LaTeX, I had to print in graphics mode which was quite slow. (Although for most class papers in the 1990s, regular 9-pin text output was acceptable.)
Note, that he didn't supply any font (he mentioned that twice, but with other words), so the program uses the random bitmaps for characters. That's why it is unreadable. If you modify the program to load correct shapes for characters everything will be readable enough.
About two decades ago, I was working on POS software for a company. At that time, most receipt printers used RS232 communication, and I remember a project where we created virtual printer based on communication data generated by the PC's RS232 port. The purpose was to test whether software outputs were accurate without actually using a printer.
That's a great project - and practical!
That's was a cool idea, when I had that problem, I wrote a table of command translation between the matrix printer and ESC/P so I could just send the thing to a fast laser printer, It was very fast to make tests, but wasted paper.
Thanks for this Jim. I always find your videos very interesting. Keep up the great work. Cheers from Canada!
Thanks! I'm glad you liked it.
I and my friend worked together to replace an old (1-pin with 9 colors ribbon) dot printer which was a part of scintigraphy thyroid scanner.
this scanner was built as a tank and it prints as it scanning, the printer and the sensor mounted on one big moving arm and it prints directly on a paper laid down and held on a metal plate using adhesive tape , you really have to see it 😁🤦♂
The reason to do the app is that the stock ran out from compatible ink ribbons
The problem was it doesn't output any digital signals to rely on and al what we got is analog sound generator producing a wave depending on the strength of the radiation detected by the sensor.
My friend designed an electronic interface circuit to convert these waves to something our program can read via RS232, and I built the app to mimic the output of the printer in order to print it on a modern inkjet printer.
Thanks Jim, you bring back so many memories
That project sounds pretty awesome!
@@freedosproject It was indeed, we enjoy doing it to revive an old machine.
Thanks for the great reply Jim
Too bad that delay() requires an integer argument. Otherwise you could adjust the delay to something closer to, say, an Epson LX.
One way to slow it down to a fraction of a millisecond would be to use a counter, and then you add the delay for every 2nd or 3rd byte. A live human might not spot the difference. Basically this:
int count = 0;
...
if ((++count % 3) == 0) {
delay(1);
count=0;
}
Also needs sound emulation.
I would have thought converting the character set would be much more involved, this is very cool!
I had a moment to think about it, and I think it will be 30 years this year since I last used a dot matrix printer. The sound as it tracks across the sheet and advances rows is very nostalgic.
The old dot matrix sound was kind of soothing, in a way. 😄
@@freedosproject it helps that now it doesn't come with a parent shouting "stop that screeching!" from the next room 😃
@@freedosprojectI still remember the IBM pro printer. Such a nice sound.
@@skilletpan5674 : Legendary!
In amongst my stuff I have the manual for the star NX1000 mono dot matrix printer, The manual includes some BASIC listings for testing the capabilities of the printer, including user defined characters. Maybe include support for sending “escape” codes (char 27) for enabling underline, sending data for user defined characters, etc
My printer at the time was Epson, and Epson printers had that ESC/P print control. The name actually comes from Epson Standard Code for Printers because Epson made it.
I hope to update this in future to add ESC/P print control like underline and even graphics mode.
This is awesome. I wonder how difficult it would be to redirect output to LPT1 to this as a TSR?
Depends if you already have a robust TSR framework to build on, there's just a few int 21 calls to intercept and you'd have to parse pathnames for 3Dh. Making a TSR that knows when it can take over the system is the tricky part, and it would need to be quite big for the graphics buffer.
@@ssl3546 It's actually not int21, but int17 (AH=0 lpt_putch, AH=1 lpt_init, AH=2 lpt_status).
I think an easier way would be to redirect printing to a file (many DOS word processors can already do that) and then use this to "print" it to the screen.
again, you had some issues with sound, just in one place
I need a better sound setup. I'm recording on the same system where I'm running FreeDOS, and I think that causes issues sometimes. Pros use a second system to record.
9 pins SUCKED! 24 ftw.
I always liked that Epson 9-pin printer - it saw me through university. But to get high quality output, like when printing physics reports from LaTeX, I had to print in graphics mode which was quite slow. (Although for most class papers in the 1990s, regular 9-pin text output was acceptable.)
I still have a dot matrix printer dunno if works.
Epson FX-80? That was the one I had from high school through my university days. Solid printer.
@@freedosprojectEpson LX 810L actually.
Very cool
most excellent
🤍
Great video!
Glad you enjoyed it
What is purpose of this program, because result is clearly unreadable...
15:23 it's mostly a demonstration of the way dot matrix printers work without wasting paper on several demonstrations
Note, that he didn't supply any font (he mentioned that twice, but with other words), so the program uses the random bitmaps for characters. That's why it is unreadable. If you modify the program to load correct shapes for characters everything will be readable enough.