Wrong start-up message

[Derry UK]

Just built my Linc80 but getting the wrong start-up message. It's the correct structure so seems to be reading the correct message but I am getting the wrong text. I have checked the output at the Z50 bus (with a PicoScope that will read and decode serial data) and the SIOB port output using PuTTY. I get the following in hex and what it should be in ascii.

OD OA OD OA AA A8 AE 28 28 20 96 18 A8 00 BB B2 D8 94 20

                          S   m    a   l    l         C   o   m  p   u    t    e    r

32 18 58 1F B2 18 94 20 2D 20 20 1F 46 96 38 30 0D 0A

M   o   n   i     t    o    r        -          L  i     N  C   8  0

As you can see some characters are correct CR, LF, Space 8 and 0.

One other comment, initially I got no message and investigation found that D3 was being held low. I found that it was my CF interface at fault so swapped it for another one.

Is it just a flaky SIO? Any help gratefully received.

Derry.

[Steve Cousins]

Hi Derry

Well that's an interesting problem.

I've looked for a pattern in the values you have listed. 

It appears that characters with bit 6 set are 'corrupted' while those with bit 6 clear are not.

The corruptions are repeatable, but no obvious relationship to what they should be.

No characters are lost or inserted, suggesting the clock is ok.

To get as far as outputting the message, albeit corrupted, the processor, bus, RAM, ROM are probably ok

This seems an odd fault, so it could have an equally odd cause. Thus I wouldn't rule out anything. However, on the face of it I'd agree it looks like the SIO.

Here are a thoughts / things to try.

Obviously if you have another SIO, try it.

Remove unnecessary hardware, such as the CF adapter. If you get desperate, other optional chips too.

Try booting from Grant's loader by changing the ROM jumper A14.

Try changing the serial port B clock source so it runs at 9600 baud instead of 115200.

Try typing a command on PuTTY and looking to see if the LiNC80 responds correctly. Try RESET or HELP. Reset should give you the corrupt startup message again. If this appears to work it would suggest the LiCN80's receive is ok, but the transmit is not.

Running out of ideas now...

Try a different terminal program. Long shot, but you never know.

As always, double check you have a good clean 5v supply.

Do you have a different system, such as the RC2014, so you can verify all other components (terminal program, serial cable, etc) are ok. If so can you swap suspect components from the working system??

Steve

[Derry UK]

Thanks for that input Steve and the pointer to bit 6.

I believe the fault is on board as I get the same error with my scope on Zbus50 so it looks like my cable and ftdi are ok but I have another ftdi cable coming tomorrow to try.

I do have spare SIOs in my Kaypros but sadly they are /0 not /2 and I am unsure of whether they would be OK. I don't want to wreck anything else unnecessarily.

I'll work through your other suggestions and feed back with the outcome.

Derry.

[Derry UK]

It now starts up.

One of the problems I had with the build was bent pins on some ICs after insertion. On investigation some of the IC sockets were stuck in a closed position so I persuaded the stuck ones to open with my dentists probe. One of my problem solving checks was to review all sockets and IC pins again, all were OK but some 16 pin ICs seemed loose in their socket.

I removed the IC and closed the socket terminals with my probe, reinserted the IC and checked it now felt snug. That seemingly did the trick.

I would add that to me the IC sockets seemed "lower quality" than previous kits I have built where some 4 or 5 IC sockets had connectors that were stuck together, hence the bent pins. Yes I accept that if I was more adept I would have investigated sooner rather than just bend a pin but hey ho. Anyway when it happened a third time it then became normal for me to run my probe into the sockets to makes sure none were stuck before I attempted to fit an IC.

But now, back to Z80 life.

Derry.

[Steve Cousins]

Good news.

Very strange symptoms though. I really couldn't figure out how a fault could cause the ROM to run ok and output repeatable characters that were corrupt in the way you documented.

Interesting puzzle to try and solve. I like that kind of thing.

Steve