Recently, I was contacted by [hpsoftek65] asking for a printer interface to enable printing from a Swiss Micros DM42 calculator.  It does support IR printing as with the original, but as printers are hard-to-find [hpsoftek65] reached out asking for an assembled unit.

It took a little while for the printer core to arrive from China, but I did test it all out and sent the un-cased unit off to Italy.  It's a pity that shipping cost nearly as much as the parts! I guess that's just the way it is.  [hpsoftek65] is a better photographer than me, and I was granted  permission to share some clips of the functioning unit.

What I sent was the wired components -- no case.  Out-of-box it was tested it successfully:

Then had to wait for a simple box to arrive to produce a case.  Semi-assembled:

And all closed up ready to test:

And here it is in action:

Noticeable things:

Printing speed is pretty slow.  This is partially because that's the way the sender (the calculator) is sending it out.  The link is one-way, and it is advised to send lines no faster than the printer can print lest the buffer overflow.  The original printer could take 1.2 - 1.8 sec to print a line (depending on battery state), so the calculators sent them out slowly.  I haven't measured the speed of this particular printer.

The IR data transfer rate is 78 by/sec, and for text there is 25 char/line, so maximally 3 line/sec.  However in this case the lines are always converted to bitmaps in order to support the distinctive HP82440B character set, and a 'line' of graphics takes about 1,000 bytes.  These printers usually operate at 9600 bps, so that means each line is again about 1 sec to transfer the data to print.  So it winds up being about the same as the original printer.  (The printers can often be configured to transfer faster than 9600 bps, though how to do that is model-specific.)

The range is short as well.  The interface board can support a much greater range -- the receiver module is designed for TV remote control applications.  It's more a function of how bright the emitter is.  Some suggest that the range is deliberately limited to thwart cheating in classroom environment, though this seems less likely than the more straightforward explanation of saving battery.  The DM42 which I have seen has a 100 ohm series dropping resistor, which is pretty high for pulsed IR.  15 ohm would not be unusual.  But the calculator does run off a single CR2032, so a lot of printing would be a lot of draining.  A bold owner could parallel a low value resistor and possibly eek out a little more range.

Anyway, it's always satisfying to see the project in the field working for someone else!