Building a VersaTerm - A Retro Computing Serial Terminal

[Editor]

But next week I will try your proposal to ground the CTS input and I will let you know !

I look forward to hearing how you go. Hopefully you can get to the bottom of this.

Perhaps others will also chime in with their experience / thoughts.

[djrm]

Probably not relevent here but a couple of things which have caused me grief with serial interfaces are, 1) dodgy fake max3232 devices not working very well and sometimes getting hot. 2) bad choice of 'protection' resistor value placed in series with signal lines. 3) devices nearly always needs cts asserted in order to receive data unless explicitly programmed to ignore it. 4) When dealing with rs232 signals always check with an inline led tester gadget to know what the signals are doing.

Hth, David.

[PE1KEL]

Hello everyone.

After a long time of inactivity, I would like to share the current state.

I have tested the grounded CTS but that didn't solve the problem.
But, 3 weeks ago I've got a reply on the same question I asked on github from @ Cam-analogue and that was the solution.
R29 and R30 has to be changed to 10kOhm. Now I don't have to use the 74HC14 and my Z80 board is directly connected to the Versaterm and it works great.
This is the link where @ Cam-analogue explains the reason about it:
https://github.com/dhansel/VersaTerm/is ... 3067105643

Maybe someone else with this problem will benefit from the solution.

Thank you all for your thoughts.
Best regards.

[Editor]

I have tested the grounded CTS but that didn't solve the problem.
But, 3 weeks ago I've got a reply on the same question I asked on github from @ Cam-analogue and that was the solution.
R29 and R30 has to be changed to 10kOhm. Now I don't have to use the 74HC14 and my Z80 board is directly connected to the Versaterm and it works great.
This is the link where @ Cam-analogue explains the reason about it:
https://github.com/dhansel/VersaTerm/is ... 3067105643

Nice find. On looking at the schematic, the 1K values for R29 & R30 do seem to be an excessively strong pull-up. Particularly for retro 5V TTL or NMOS based interfaces.

For earlier 5V TTL level digital circuits, I've always traditionally used 4K7 as the default pull-up resistor value. With more modern, lower powered CMOS based devices, 10K is generally the go-to for an appropriate pull-up value.

On reading the link, I see the first post mentions: "an inverter between the TX from the Z80 and the RX of the Versaterm".
This will not work, as you are then inverting the serial data line (corrupting the serial signal).

Unless, they were meaning the use of 2 invertors back-to-back (to return the signal to the correct polarity)?
So, effectively just buffering the Tx signal from the Z80 board?
In this case, simply buffering the signal (with a suitably capable driver), would've also resolved the excessively strong input signal pull-up issue.

However, changing the VersaTerm's R29 & R30 input signal strong 1K pull-ups, to either 4K7 or 10K instead, is indeed a better general solution.

Maybe someone else with this problem will benefit from the solution.

Thanks again for posting this solution. I'm sure it will be helpful to others experiencing the same issues!