help - 11/34 console problem - first measurement results

[Tony Duell]

> > What does it do if you press 4 repeatedly? Does it flash the bottom 3 or
> > bottom 2 digits? If it's only updating the bottom 8 bits, you see, then
> > only 2 digits will flash from 0.
>  =
> 
> Indeed, pressing the "4" only flashes the last two digits!
> How did you know that ?!

Call it a very lucky guess. Since the 8008 is an 8-bit device, and we 
know it's not updating the full 16 or 18 bit word correctly, it would be 
logical to guess that the carry had failed between the low 8 bits and the 
next 8 bits. Now thin what happens when you try to enter 444444...

> Next, the response on the display from the keypad. I always
> started with pressing "CLR", then repeatedly the same key.
> The rightmost (lsb) display is called "A", the one next to it
> is called "B", the leftmost (msb) display is called "F".
>  =
> 
>    button   display reaction (1st time, 2nd, 3rd, etc)
> --------------------------------------------------------
>       0    nothing to see
>       1    nothing to see

Interesting, the LSB is being lost somewhere (that would explain why 1 
does nothing, why 5 behaves as 4, etc).

What sort of RAM is on this board (I can't be bothered to find the prints 
tonight ;-)), do you know if it's good. I've found old SRAM to be 
notoriously unreliable.

[...]

> Last 3 signals are the display values (3-bit) to the 7447 input
> A, B, and C. On connector J1 pins 12, 11, and 10 respectively.
> As always, the last thing to check should have been the first!
> These 3 signals are continously 0 V. That explains why the display
> always shows "000000". I can see a short positive pulse when a key
> of the keypad is pressed.
>  =
> 
> This leads me to the following (early) conclusions.
> 1) the keypad is scanned. All "X" signals of the matrix (READ x H)
>    are OK, and all "Y" signals of the matrix (the 8881 outputs) are
>    all valid, and give the correct information when a key is pressed.
> 2) When a key is pressed it is momentarily shown on the display, but
>    then the value returns to all zero. I can not see the value, so
>    it could be any garbage ...

Since I haev a logic analyser 'to hand', I would start looking at those 3 
inputs to the '47. Where do they come from on the M7859? An output port 
on the 8008? Do we know that port is latching correctly? Or more likely, 
the 8008 is writing the wrong values to it


-tony