Well
I have some good news & bad news Gary.
Quite
by accident and for other reasons, I popped the 80386 board
into my test system using the Cyix 80486-40 rather than the
normal AMD 80486-40. To my surprise I found I could not
get the SMB V2 with K4 (2-3) i.e. output a bit 0 to a 1 to
switch in the 80486.
If
you remember the Cyrix is a poor man’s 80486. It’s an 80486
(without the math processor) in a 80386 socket. It runs
80386 code about 10% faster.
What
is interesting is that the old SMB V2 K4 1-2 position
switched fine and back to the Z80. Removing the K4 and
using the Cyrix P32 1-2 & 3-4 also worked fine back
& forth.
As
Shakespeare said: Something is rotten in the state of
Denmark here. I suspect the problem may be on the SMB
V2 port and not the 80386 board.
There
were some 74Fxx chips on the Cyrix board that were not on
the AMD board (have two BTW, both behave the same), I
switched these – no difference.
It’s
late, so more tomorrow – the goes my weekend!
On
yours BTW, remember the SMB V2 LEDs are not monitoring the
actual S100 bus lines, they are looking only at the output
of the FF, U33. Also remember you are seeing Q* and not Q.
The front panel “TMA” LED however does show the S100 bus
status of TMA0.
In
any event your TMA3* should never light up. This indicates
that pin 3 of U33 is going low. My BIOS for MSDOS for the
8086 family uses TMA0*
From
the MASTER.Z80 Monitor…
;
;THIS
ROUTINE ACTIVATES THE S100 TMA0* LINE USING THE SMB
(SWITCHES COMTROL OVER TO THE 8088, 8086 or 80286....)
;Inputting
(any) data from Port SW_DMA0 lowers the S-100 TMA0* PIN #55
line.
;THIS
WILL CAUSE THE 8086/80286 BOARD TO BECOME ACTIVE AND TAKE
OVER THE BUS.
;THE
Z80 WILL BE IN A PERMANANT HOLD STATE UNTIL THE S100 PIN #55
IS AGAIN RAISED.
;NOTE
FOR THE NEW V2 SMB WE RAISING BIT 0 OF PORT 0EEH (AND ALSO
LOWERS TMA0* FOR THE SMB V1 Board).
;WE
WILL INCLUDE BOTH OPTIONS
SWITCH_8086:
LD HL,MSG14
CALL PRINT_STRING
IN A,(SW_TMA0) ;THIS SWITCHES CPU'S with no block
Move
NOP ;Z80 WILL BE HELD HERE
NOP
LD A,01 ;Utilize the more specific circuit
on the V2-SMB
OUT (SW_TMAX),A ;Make sure its bit 0
NOP
NOP
JP BEGIN ;WILL DROP BACK TO REBOOT MONITOR
Hope
this helps. Will dig some more tomorrow.
John
From:
n8vem...@googlegroups.com
[mailto:n8vem...@googlegroups.com] On Behalf Of Gary
Kaufman
Sent: Friday, February 6, 2015 12:10 PM
To: n8vem...@googlegroups.com
Cc: mon...@vitasoft.org
Subject: Re: [N8VEM-S100:6218] Re: 80386 CPU Board -
Correction
Thanks David for your "eagle eyes",
unfortunately correcting the .1uf -->47pf didn't fix my
issue. I've also swapped every IC except the 386, checked
every resistor and resoldered very joint.
Slowing the system way down (2mhz oscillator) makes it
switch back and forth properly.
I've tried swapping the 74LS74's for 74F74's without any
effect. Unfortunately I lack the expertise or equipment to
see what's really going on.
Using QIED swaps back and forth perfectly, but once the
QOEE,01 is used the system won't reset to Z80 mode using
either QIED or QOEE,00 and the LED's stay fixed at the TMA0
setting.
I'll try Johns suggestion this weekend and see if it works.
Once more curiosity is that when I boot MSDOS the TMA3 led
comes on. If I've activated the 80386 by QOEE,01 booting
MSDOS will switch to TMA3 and drop back to the Z80, but if
I've activated the 80386 with QIED it boots fine.
On Friday, February 6, 2015 at 12:39:37 PM UTC-5, monahanz
wrote:
Good
point Dave, C67/R18 is to pull out noise/false
triggering from pHLDA transferring control to the
slave. (came from an old Godbout board, if I remember
correctly). Could be an issue getting back to Z80.
One
test you could do is use a completely separate port in
your system to flip between CPU’s…
Find
a board that has a port in which you can output a bit
high or low. Our Parallel Port I/I board is ideal for
this. It must be high on reset. We need a spare OC
gate. Use U49 pins 11 & 10 on the 80386 board.
Disconnect the K4 jumper on the SMB V2 and all
jumpers on P36. First check the “O” command then
does nothing. Next run wire jumper from a bit on your
(new) output port to the bent out pin of U40 pin 11 .
Connect the bent out pin 10 of U40 to pin 2 of P36.
Also jumper P36 3-4. Lower the bit on your new output
port, the Z80 should switch out and the 80386 in.
With a logic probe check the levels of pins 11 &
10 of U40.
Next
within the 80386 RAISE the bit. The Z80
should switch back in. Again with a logic probe check
the levels of pins 11 & 10 of U40.
Report
back!
John
From:
n8ve...@googlegroups.com
[mailto:n8ve...@googlegroups.com] On
Behalf Of David Fry
Sent: Friday, February 6, 2015 12:13 AM
To: n8ve...@googlegroups.com
Subject: Re: [N8VEM-S100:6212] Re: 80386 CPU
Board
I
compared your picture to my board and jumpers are
the same BUT one thing I did notice,
Check
C67 directly beneath the 80386 chip, have you
mistakenly fitted a 0.1UF here as it looks the same
type as the cap next to it (C18 which is a 0.1uf)
C67
should be a 47pf and is in the master/slave area of
the circuitry.
I
only noticed it because on my board the 47pf is blue
and the 0.1 next to it is yellow :-)
On Friday, February 6, 2015 at 4:42:55 AM UTC, Gary
Kaufman wrote:
John
-
Thanks, that helps narrow things down!
- Gary
On 2/5/2015 10:53 PM, John Monahan wrote:
Jumper
looks exactly as I have Gary.
John
--
You received this message because you are subscribed to
the Google Groups "N8VEM-S100" group.
To unsubscribe from this group and stop receiving emails
from it, send an email to n8vem-s...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to
the Google Groups "N8VEM-S100" group.
To unsubscribe from this group and stop receiving emails
from it, send an email to n8vem-s...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the
Google Groups "N8VEM-S100" group.
To unsubscribe from this group and stop receiving emails from
it, send an email to n8vem-s100+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--