Good luck with it Dave. I had not tried out the IDE routines in the 68K monitor. I got it as fear as hand cross coding the 8086 routines and to assemble with no errors. Glad you are catching these subtle issues.
If we are to shorten pWR* there are a couple of flip flops available on U37, but I have to confess I don’t really know what Wilcox had going on there.
From: n8vem...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of yoda
I plan to tonight - btw still have not gotten IDE board to work yet with 68K (I think the problem is with writing registers - read seems to be OK so the pWR* maybe the issue) - though I did get the disk select to work with the mods I have made so far. Also on you monitor code you will have problems in your IDEWr routine as you do an ori with a memory location which is a port - the 8255 will not be able to do the read modify write so you need to do an ori with register and move the register to the port - I fell into same trap - I think a long session with Mr. Logic Analyzer and I will figure out what is going on.
Hi Dave, just wondering if you had a chance to look into this pWR* signal issue on the 68K board.
From: n8ve...@googlegroups.com [mailto:n8ve...@googlegroups.com] On Behalf Of yoda
z80 has plenty of margin - 250ns data valid after pWR* goes hi. I am seeing about 10 - 20 ns but not consistent. Need to figure out how to shorten pWR* or get data to be valid longer. I guess I will be getting intimate with Wilcox book on his S100 state machine. I believe Pontius experienced that is was very marginal. Guess you have been lucky John.
Yes I think we need to get this resolved quickly before we do another round of boards. I did find in Wilcox's book on page 331 that the chips should be schottky devices which would be U38 and U42 which I have replaced with 74F series - the prop console works but I would not say reliably yet. I am putting it back on the logic analyzer now. I also notice compared to that diagram the pWR* path has an extra inverter in it which would cause and extra 7-10 ns delay which may be stretching pWR* out maybe too far. I want to compare to Z80 board after I take another snapshot on the analyzer - I think pWR* is getting delayed too long. The circuit you are thinking of in the next post is the one that generates A0 from LDS and uses an RC circuit on one of the legs - that seems to be working ok as the address lines look solid. I will continue digging and let you know - think we should delay the 68K board until after you get back. Have a fun vacation
Dave I have not had a problem with that board with the Propeller Console IO board at all. Have not run it with the IDE board. That said I could only get about 6MHz out of a 10Hz chip. I know Todd is about to order another batch, we should resolve this ASAP
John Monahan (mon...@vitasoft.org)
On Jun 28, 2014 3:32 PM, "yoda" <yo...@r2d2.org> wrote:
Wondering if people have been using the 68K board. I think I have timing issue because pWR* is going HI at the same time the data is being released. You can see this in attached LA capture. I think that is why I am not getting characters from the prop console and probably explains why I have not gotten reliable data from the IDE card either. The dual serial card and memory work just fine. I have tried different clocks as well - this one is at 2MHz so that should not be the issue. Was wondering if someone had seen this and resolved. I will dig through the schematics and the Wilcox book as well. I think once I have this resolved, I can get CP/M 68K going.
Thanks in advance