• Please review our updated Terms and Rules here

SCELBI Data Bus Oddities and Debug

I probed Z11 and can confirm that D is in fact toggling on it:
1750259310996.png

So that's 0101 or 5. Which is pin 6 of Z11:
1750259388040.png
Which goes to to OS15 on pin AX.
1750259436159.png
I will next probe this pin. If it works, I might swap one of the strobe pins on an existing socket for it and see if serial works that way. I still think something is fishy, but if it gets me serial then it gets me serial.
 
AX is held low.
BU (OPSN) is held high. This is the other input to that NOR gate. So that makes sense that AX is low then. Hmmm
 
If I toggle in 135 though, which is output to oct 16 ("OUT6")
1750260512275.png
Top trace is pin A21 (OS6) and bottom trace is BU (OPSN)
So it seems like the hardware is working. We know the chips are programmed properly, but maybe I need to try and read addresses in ROM as you suggested yesterday and double check it's reading right.
 
I have just worked my way from the 8008 CPU to the buffer/output board.

When an OUT instruction is used RR is 01 (binary) and MMM can be 0 through 7 (MMM being binary and 0 through 7 being the octal equivalent).

Z12 is the decode for these output ports driving OS0 through OS7.

I can't see what the value of Z11 actually is at this point in time. It seems to decode for an INP instruction with RR=00 (binary)?

The OUT instruction to the Z12 IC to the OS0-OS7 through to the sockets all seem to align.

Dave
 
I have chased down the enable logic for the input board now. The some of the signals are (for example) S0, but these do not necessarily go to S0 on the CPU card!

S0 XA04 BU is actually wired to XA02 A11 (D4N).

This makes sense:

Code: 
D7 D6 D5 D4 D3 D2 D1 D0
 X  X  R  R  M  M  M  X
 X  X  X  0 = INPUT port (decodes for INPUT card based upon D3-D1).
 X  X  X  1 = OUTPUT port (decodes for the buffer/OUTPUT card based upon D3-D1).

Not still sure what Z11 on the buffer/output card actually does for us. It is almost as though it should decode for another range of output ports, but I can't see (yet) how it does...

Dave
 
It looks to me like D4 is selecting either Z11 or Z12 via this:
1750262663114.png
Specifically, Z16 pin 9 / Z13 Pin 2 originates ultimately from D4N

Also to clarify, yes the DBB has more output port capability than the SCELBI chassis actually physically has:
1750263941886.png
 
Last edited:
I was just thinking that whilst driving home from work. Great minds think alike!

I shall check the backplane connector wiring later to see what 'Harry Potter magic' is done there! I have the schematic sketched out in my notebook.

Dave
 
OUT 16 = Octal 135 = Hex 5D

1750265184868.png

This shows up four times in the TTY ROM at addresses
HexOctal
3E0CP: 76 A: 014
3E2AP: 76 A: 052
3EA6P: 76 A: 246
3EC6P: 76 A: 306

I can check these to see if they actually read 5D.

EDIT: Yes, we are getting 135 at all four of those addresses.
 

Attachments

  • 1750265073452.png
    1750265073452.png
    58.9 KB · Views: 0
Last edited:
I just found that data bit 0 is stuck high -_-

One of my RAM boards oddly. I'll just take it out for now and figure that out later.
Still no serial though. So that wasn't the issue.
 
Last edited:
Hold on, so we determined we were writing to port 13, or so the page lights said. That's the tape interface. So probably what I'm seeing is an attempt to write to the tape interface for some reason. But at least that address makes sense now.

So maybe I'm just missing when it tries to interface with the TTY.

Maybe I'll try and play with triggering on the scope to see if I can capture anything.
 
I think I have worked out the 'magic' for the second output decode IC (Z11) - it is done in the CPU card via the OPSN signal driving the buffer/output card via the BU pin. This aligns with your oscilloscope trace in post #84. OPSN goes LOW (to enable the decoded strobe outputs) and the selected strobe (pin A12 OS6) goes HIGH.

D4H is still used to decode whether Z11 or Z12 are actually enabled (via their 'D' input). But the outputs are only enabled (via the OPSN signal and the following NOR gates) to signal the outside world.

Post #48 states that you used instruction 135 and it triggered OUT 6. I think we were getting confused with the 'port 13' issue...

Incidentally, whilst looking at the buffer/output card I see there are two (2) versions of it. There is an extra IC (Z16A) on some boards and not others. This is to do with the 'D' decoding for Z14 and Z17. On scelbi.com we appear to have the schematic and layout for the version of the buffer/output board without Z16A, but a photograph of the top and bottom of a board that is equipped with Z16A.

Which version do you have (out of interest)?

Dave
 
I used 135 to trigger OUT6 (Output 16) to ensure that it was working at all. Output 13 seems to be the tape drive. I was actually able to trigger the scope on strobe OS6 whenever I start MEA. This is consistent with sending the '>' char to begin with. However, I cannot get a good enough view to really verify that.
I think next I should type keys and see if I get more triggers on the scope at least.

I did notice that the revision on SCELBI.com of the DBB is slightly different, which has been a mild pain. Mine does have Z16A. I haven't seen any schematic for that myself.

EDIT: IT DOES TRIGGER ON A KEYPRESS. This tells me that MEA is probably actually running fine. I suspect highly that I just have some RS232 issue. Either my settings are not set right or my RS232 adapters both don't support such a logic level. I guess the serial output to the PC would be the next thing to probe then.
 
>>> I did notice that the revision on SCELBI.com of the DBB is slightly different, which has been a mild pain. Mine does have Z16A. I haven't seen any schematic for that myself.

The modified DBB (with Z16A) is in the back of the SCELBI-8B Assembly Instructions manual...

You should see some activity from one of the Q outputs of the 7474 on your RS232 board.

If you don't, you should have activity on the clock input pin of the 7474 (this is the strobe) and the D input of the 7474 (this is the data).

You should also be able to observe the same pulse train on the input to (and output from) the RS232 driver.

Are you able to post the front and back side of your RS232 board? I should then be able to tell you which pins to look at with your oscilloscope. Ignore this request. I have managed to print out Mike's files from home. The IT system at work wouldn't let me...

You are getting there!

Dave
 
Last edited:
IMG20250516081645.jpg
IMG20250516081651.jpg

Also design files are on my github. I believe this design uses the 2nd input of the 7474 rather than the first. I was probing the card edge connector pins because it was easy. Unfortunately, I think my scope may have too much lag after triggering to see much. That or there isn't much there. I'm trying to figure out how to address that.
But also I have a meeting for work, so it'll be a bit before I can do anything more. I appreciate the help!
 
Yes, the second gate (number 2) of the 7474 is used.

Check pins 10 and 13 of the 7474. These should be pulled up to +5V (HIGH) via the resistor on the board.

7474 pin 11 is the clock input = SCELBI strobe output (from whatever OS we decided it was in the end!). You should observe one (1) pulse here for every serial data bit transmitted - including start, data and stop bits.

7474 pin 12 is the data input. This is the data itself that will be latched (every time the output strobe is activated).

7474 pin 9 is the serial data output from the 7474 to the RS232 chip. This should be 'waggling' up and down as the serial bitstream is transmitted.

EDIT: I see this warning "PLEASE NOTE I HAVE NOT TESTED THIS CARD YET!!!". Perhaps we are testing it now!

>>> I appreciate the help!

Absolutely no problem. I am enjoying myself working on something that I always fancied building, but never got around to it...

Especially if you can get MEA working for the exhibition... SCELBAL may be too much though...

I wrote a software emulator for an 8008 machine. I enjoyed doing that - I gained an awful lot of knowledge about the actual operation of the 8008 CPU. We were also disassembling the assembler package for the SIM08 at the time.

Dave
 
Last edited:
I have data and strobe swapped :o
One moment. Quick solder job between meetings. Brb

EDIT: Not my whole issue (but definitely an issue)

Lemme check other pins quickly.
 
Last edited:
My data wire broke off. Gotta meeting then I'll fix that. Also I was able to get my scope to trigger how I want fast enough. So I should be able to get better data.
 
You must log in or register to reply here.
Back
Top