TashTalk USB

[Bolacore]

I am admittedly biased as I adore my Saleae, especially because I can have the analogue view side-by-side with the decoded digital side to track down quirks. But your capture I think is exactly what I needed -

From looking at the Sigrok capture there's pretty significant pulse width distortion on the LC II's packets, and where the ImageWriter should be responding I think the signal is so weak the RS422 transceiver is just clipping it (I see like 30 microseconds sometimes between pulses), so no wonder the PIC isn't outputting anything.

Any chance you have an oscilloscope to confirm, or are up for some surface mount soldering? I have an idea on how to fix it, but it does require some 0603 parts being added hah. If not I'll be picking up a really long phone cable tomorrow and can hopefully repro it too.

 

[iigs123]

I don't have an oscilloscope, but I can do SMD soldering. Just let me know what you have in mind. Of course I have no issue just going with a shorter cable too if that is the solution. I have no barbed wire fence in the suburbs!

 

[Bolacore]

Ah ok if you are fine with a shorter cable then it is no worries. My current thought is the signal is being attenuated enough that the transceiver is going in to fail safe mode and just going high causing the frames to be lost. I think if we add a little DC bias network on the RX lines to shift the neutral point we can pick up these weaker signals. I am only an amateur hour electrical engineer so i could be totally wrong here of course too.

If I can reproduce this and the bodge makes it better I’ll do a respin of the boards and get one out to you as thanks

 

[iigs123]

If you are able to reproduce and are interested in a change to address it, I'd happily test things out. I do enjoy learning more about all of this stuff. Its vintage tech, so I have no expectations that any of this stuff works perfectly at all times.

I'll do another test soon, putting my G3 at the end of the long cord to see how it might "see" the tashtalk serving files.

 

[iigs123]

Well, more somewhat goofy results from today!

I put my G3 at the end of the long phone cord, to see if it would be able to connect to the tashtalk. And, it did not, until I removed the terminator from the phone net connector next to the tashtalk!

I then tried the ImageWriter with the terminator removed. It too was found using nbp-lookup.

 

[adespoton]

Well, more somewhat goofy results from today!

I put my G3 at the end of the long phone cord, to see if it would be able to connect to the tashtalk. And, it did not, until I removed the terminator from the phone net connector next to the tashtalk!

I then tried the ImageWriter with the terminator removed. It too was found using nbp-lookup.

That kind of makes sense; the terminator, after all, is essentially a resistor. With the long cable, you've got built-in resistance, so the cable is likely self-terminating

 

[iigs123]

@Bolacore do you have a special pap-print you used with your ImageWriter? I took your example and paired it down by getting rid of the postscript stuff but the connection fails.

It does see and connect to the ImageWriter it seems. After a minute or so something times out and the ImageWriter does a form feed. I'll admit, the first time that happened the noise startled me!

 

[Bolacore]

I do yes - it’s still a work in progress but as @NJRoadfan mentioned the card is super quirky hah so I had to fix some other stuff. I found the card doesn’t set EOM on ATP responses so we just sat there waiting forever and is why it times out. I tweaked our logic so it’s either “we get EOM or get the expected number of packets” and that seems to make things work.

My codes pretty messy still but I’ll push up what I have so far in a little bit and you can give it a try.

 

[Bolacore]

Here's my current WIP stuff that I used to get it all working: https://github.com/FeralFirmware/TailTalk/blob/imagewriter-printing/examples/iw-print/main.rs. It's largely the same as pap-print, but routes the data through ghostscript's ImageWriter driver to convert it to the format needed. It's a hard coded page of course, but you can modify it.

I also added some really basic integration support in the IPP bridge that I used to print Ferris in the other picture, but it's still unstable and not really in a state to be merged. Will be reworking a good bit of this before I add it in later this week.