IEEE488 and 1541 drive, the IEEE interface...

[fachat]

I've monitored the ATN, NDAC and NRFD lines and some others. This shows the states during a reset and then 5 dir commands. The first two returned full listings. The third and fifth returned nothing, and the fourth returned a partial listing. This was all within the space of a minute. There are a few more detailed sections of these, but they seem to show that ATN doesn't stay low on the two fails.

It would be good to see the more detailled traces on the fails. The PET sets ATN lo when it wants to start an action, but then checks NRFD and/or NDAC low to see if there actually is a device attached. If it does not detect that, it sets ATN high again without further ado.

You should set the time resolution on the scope such that you see the whole time where ATN is low (which is the full command transfer, and should be two bytes, LISTEN(device) + OPEN(secaddr) when ok, or ...??? when not.

André

 

[TynemouthSoftware]

It would be good to see the more detailled traces on the fails. The PET sets ATN lo when it wants to start an action, but then checks NRFD and/or NDAC low to see if there actually is a device attached. If it does not detect that, it sets ATN high again without further ado.

You should set the time resolution on the scope such that you see the whole time where ATN is low (which is the full command transfer, and should be two bytes, LISTEN(device) + OPEN(secaddr) when ok, or ...??? when not.

André

Here is the whole of the failed directory trace.

Here is a close up of the start and end of this, with just NRFD pulses in between

And even closer, here is the whole time ATN is low on the first pulse at the start

Thanks,

Dave

 

[fachat]

Here is the whole of the failed directory trace.
View attachment 22242
Here is a close up of the start and end of this, with just NRFD pulses in between
View attachment 22243View attachment 22244
And even closer, here is the whole time ATN is low on the first pulse at the start
View attachment 22245
Thanks,

Dave

Hi,
I'm sorry, I'm not sure I understand which one is which. AFAIU http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=22244&d=1421137716 is the end of the bus traffic, right? At least it shows some activity I just can't make any sense of. (I'm a bit busy right now also...)

Maybe you can try yourself to find out what's happening by reading the PET ROM, and IEEE488 protocol info like http://www.hp9845.net/9845/tutorials/hpib/
Given the (missing) documentation I don't find on the web, I should probably write something up, but - as I said - I'm rather busy right now...

I may have a look at some later time, but just wanted to give you some notice.

André

 

[TynemouthSoftware]

The first is the entire trace of a failed 'cbmctrl dir 8' command. The second and third are zoomed into the start and end of this sequence. The final one is showing only the whole time that ATN was low at the very start of this sequence.

I've been trying various things. It's a bit difficult as it seems to fail most of the time initially on the pet, then after about 5 minutes starts working, and works most of the time. On the USB adapter, it works about 2 times in 5, even if that is 5 commands in a row, so never long enough to be useful.

I appreciate the help so far, it is almost there, so it would be good to get it fully working, but it's by no means urgent.

I did fin some 2031 test programs here:

http://www.vintagecomputer.net/commodore/2031/
and
http://www.vintagecomputer.net/commodore/IEEE_drives/

I can't seem to get those to load. They also fail in the Vice. The same thing happened with some of the other memory test program, but all the other .prg files were fie. There appears to be some assembler after BASIC program, no sure if they have been transferred correctly? Anyone have these as a d64 or d82 image or working .prg files?

Thanks,

Dave

 

[fachat]

I did fin some 2031 test programs here:

http://www.vintagecomputer.net/commodore/2031/
and
http://www.vintagecomputer.net/commodore/IEEE_drives/

I can't seem to get those to load. They also fail in the Vice. The same thing happened with some of the other memory test program, but all the other .prg files were fie. There appears to be some assembler after BASIC program, no sure if they have been transferred correctly? Anyone have these as a d64 or d82 image or working .prg files?

I don't know in what format those files are, but they are very weird Hexdump starts with "03 00 33 00 01 00 8f 20 32 30 31 20 42 55 52"
So it COULD be that just the load address has been replaced by "03 00" - unfortunately the PET starts loading at the address given in the file (the LOAD "*",8,1 behaviour from the c64) so it crashes while loading.

Loading it into the C64 crashes with LISTing though ...

Sorry, can't help further right now.
André