Malfunctioning VT240 - help please

[Charles]

On 6/10/20 4:31 PM, Jon Elson wrote:
> On 06/10/2020 12:48 PM, Charles via cctalk wrote:
>>
>> That leaves the unlikely possibility that one of the octal TTL 
>> devices, or ROMs. has developed a weird internal pathway that only 
>> interferes with DAL3 & 1 on some bit patterns, but not all the time. 
>> Seems like a zebra rather than a horse. The only part that drives 
>> multiple low-order DAL lines at once besides the E19-22 ROMs is the 
>> E55 LS245.
>>
> Quite possible that this could happen when a specific device is 
> driving the bus -- or that NOBODY is driving the bus in that state. 
> When it is stuck at the ~1V level, try a resistor of about 1 K to 
> ground on one of those lines.  If it moves several hundred mV lower, 
> it is a TTL open circuit.  If it doesn't change at all, it is a bus 
> contention (TWO drivers driving at once).
>
> Jon

Yes, I've been experimenting with this. It's not 1 volt, it's 4 to 3 
volts and back again at the rate the lines should be switching :)

If it were contention caused by the LS245, the short circuit current 
would be far higher. I've also tried strapping the OE\ on the '245 high 
with no change. I removed all four ROMs and if there's bus contention it 
is not coming from them. Unfortunately the USART and UART are not in 
sockets, but no change when their chip selects are forced invalid.

Even more interestingly, I have discovered that when the T11 is crashed 
completely (e.g. after I slip with a scope probe on the DAL or other 
lines), if I connect DAL3 when at a steady high, through a 1K to ground, 
results in a 0-3 volt output switching at the instruction cycle rate 
with a slow risetime and a rapid fall! That is not possible if the bus 
were tristated or in contention...

There has to be something in the T11 internal drivers that is latching 
up somehow. It would really help to have another T11 aka 310 aka 
21-17311 ALSO aka KR1807VM1 (Russian clone) to try - but apparently the 
Atari enthusiasts have snapped them all up and I refuse to pay over $100 
for a tested working one on Ebay!