IBM 6094-020 Lighted Program Function Keys

[Michael B. Brutman]

I'm still trying to make this thing talk ...

Here's the quick recap.  It is driven by a derivative of an Intel 
MCS-51.  It is supposed to speak 9600 N81 over a serial port to a PC. 
The protocol is supposed to be simple.  But I haven't been able to make 
it talk.

I borrowed a friend's multi-meter that does frequency counting, hoping 
to use it as a poor man's oscilloscope.  I verified it with the PC 
serial port - sending characters definitely makes it wiggle.

Connected to the device (which is powered by an external adapter), it 
shows +10VDC on its pin 2 (which is 'receive' from the PC perspective), 
but no wiggling with the frequency counter.  This confirms what I saw 
with my meter - a straight 10VDC signal coming out.  The voltage level 
is fine for serial, but it is not a data stream if it is continuous.

Opening the device up shows the MCS 51 feeds a Maxim chip, which is used 
to get the TTL voltages up to proper RS232 voltages.  The input and 
output to the Maxim chip are behaving the same way - a continuous 
voltage.  Kind of like the MCS 51 is fried or stuck in a loop waiting 
for initialization.  (And of course, I don't know what it expects.)

The unit has a small switch on the backside that is either in a 'normal' 
or 'loopback' state.  It appears to function in loopback - you can press 
a button and the light for the button turns on, turning off any other 
previously lit button.  But nothing comes out of the serial port in that 
state.  In normal state it does absolutely nothing, and has the steady 
voltage on the output pin.

Is it possible thing thing is waiting for the serial port to set one of 
the other pins for flow control?  Any other ideas?  (I need an 
oscilloscope ...)


Regards,
Mike