In the course of troubleshooting a perceived serial port problem on my M100
I found that when the DTR/DSR was looped back (and asserted) the voltage at
the DTR/DSR junction was 2.7V and the input voltage at M35 was only +1V.
There was 2V dropped across the 5.6K series resistor on the DTR output. 

 

Testing DTR/DSR loopback on the T102 and there is no issue. The voltage at
the DTR/DSR junction was about 4.5V as one would expect. 

 

Comparing the schematics of the two revealed the reason. The T102 has a much
beefier output drive. Not only did they double up the output drivers they
dropped the series resistor to 330ohms! That us a HUGE change which makes me
suspect that there was reliability issues they were trying to fix. 

 

I know folks have had different success with various baud rates, for some
higher rates seem to work OK but not for others and I wonder if this 'weak
tea' drive on the three serial output lines is part of the issue. I think
changing the 5.6K resistor with something like a 1.27K would take care of
the issue and still be half the current that each of the drivers in the T102
are handling. (One could also take a suitable resistor in parallel with the
existing one of the PCB as a quick fix without removing the PCB.)

 

It would be interesting to find a test case where this can be tried with the
TX line only, ignoring the handshaking signals.

Jeff Birt

Reply via email to