Proposal: Multiple button signaling on PTT line?

[hcarter333]

Not so much an issue as a proposal. Could the standard also include a pinout for a CW keyer rather than a mic? Folks might like, (I suspect very much), to bring their own keyer as well.

I've built up a prototype of this idea at hcarter333/rockmite#6

Thanks very much for putting together the standard! It inspired the work I did above. Also, if this isn't appropriate place for this question/suggestoin, please let me know, and I'll close or delete the issue. Thanks again!
73 de KD0FNR

[SmittyHalibut]

I literally just responded to your email to the OHIS mailing list and came here to enter an Issue in the backlog. You beat me to it. :-)

This is absolutely an appropriate place for this. Thank you for submitting your idea. It's a great one.

Unfortunately, the pinouts you listed above are not OHIS compliant. OHIS is the definition of the signals, and what pins they go on. Repurposing the Mic+ and Mic GND pins for contact closures breaks OHIS compatibility.

Instead, I propose the PTT line be extended from "Contact closure to ground for PTT" to a more complex analog signal:

• 1.00 * V_cc: Idle.
  • This is the "pulled up" state, same as before.
• 0.75 * V_cc: Button 3, Dit, Up
• 0.50 * V_cc: Button 2, Function
• 0.25 * V_cc: Button 1, Dah, Down
• 0.00 * V_cc (aka GND): PTT.
  • This is the "contact closed to ground" state, same as before.

This should probably be an optional extension that not every device has to support. eg: some radios that don't have or use extra buttons like a keyer, won't bother implementing these.

Encoding this would be pretty easy, just a voltage divider, so long as the pull-up impedance is known, or higher than the voltage dividers.

Decoding is more tricky. Probably a pair of LM393 comparators, a voltage divider ladder, and some logic gates on the output to select only one output at a time. Unless someone knows a better way to do this.

If I help with the circuits for this, would you be willing to prototype and test this @hcarter333 ?

[hcarter333]

@SmittyHalibut I'd be wilingl to prototype. The circuits as described above seem straightforward enough. Cool!

[SmittyHalibut]

Updating summary to reflect how I'm thinking about this now.

[SmittyHalibut]

Also worth considering that CTIA has something similar, but on the Mic line instead of PTT (which CTIA doesn't have.)

https://source.android.com/docs/core/interaction/accessories/headset/plug-headset-spec

I can think of several reasons NOT to go this route with OHIS, but its worth thinking about.

[SmittyHalibut]

This came up on the OHIS mailing list again: https://groups.io/g/ohis/topic/classic_radio_integration/102146840

Rather than define it by voltages as in https://groups.io/g/ohis/topic/classic_radio_integration/102146840, I think I like defining it by impedance instead. Something like:

• Hi-Z is idle.
• 0R to ground is PTT. (These two make it backward compatible with the current design.)
• 10k to ground is Dit.
• 33k to ground is Dah.
• 7.7k to ground is Dit and Dah together (spoiler: 10k||33k=7.7k)