Re: Restoring a NorthStar Horizon, problems with SRAM board

On May 21, 2:41 pm, Herb Johnson <herbrjohn...@xxxxxxxxx> wrote:
On May 20, 2:01 pm, lync...@xxxxxxxxx wrote:

OK, I hear you on the oscilloscope. I have bought one on Ebay and it
will be a few days for it to arrive. It is a dual trace 40MHz model
which should be sufficient for this task.

Andrew Lynch

Using an oscilloscope is not so bad. In brief, you can use it as a
logic probe (trigger on the incoming signal) and just check those
various chip pins for activity. Unlike a volt meter or LED based logic
probe, you can look at changing logic levels (or lack of change). A
"level" of about 2.5V is an open TTL input: if it's supposed to be
connected - and all TTL inputs SHOULD be connected to something - then
you know you have a bad socket or broken trace. Those "barely driving"
outputs may also suggest you have some bad INPUT somewhere on the
line; but it's possible the drivers are at fault instead. If they are
bus drivers, the bus may have had some "fault" on it, a -5V across
those drivers for instance, to damage them.

You can also tie the second channel of the 'scope to some reference -
the Z80 clock, the Z80 M1, whatever - and trigger off that refernece.
That's how you track down timing problems, in circuits with marginal
designs. The Northstar I believe does not have marginal designs. And
you can trigger off the approprate "chip select" of a given IC and
watch the activities of the outputs. That's useful for tracking down
program actions versus desired activities, such as that floppy
controller board. Since the N* controller is all discrete chips, it's
harder to track than a FDC chip.

Otherwise your posts look through. Keep up the good work, you'll have
this Northstar up in due course!

Herb JOhnson

PS thanks for the reference link on 2708's. I'll add that to my Web
site, with the linked page owner's permission!

Herb

Herbert R. Johnson, New Jersey USAhttp://www.retrotechnology.com/herbs_stuff/ web sitehttp://www.retrotechnology.net/herbs_stuff/ domain mirror
my email address: hjohnson AAT retrotechnology DOTT com
if no reply, try in a few days: herbjohnson ATT comcast DOTT net
"Herb's Stuff": old Mac, SGI, 8-inch floppy drives
S-100 IMSAI Altair computers, docs, by "Dr. S-100"

[ed note: this is what I tried to post on Monday but Google Groups
kept having errors]

Thanks Herb, I am glad you can use the 2708 link. I have had to deal
with 1702 family EPROMs in the past and they are also a pain to deal
with.

I am looking forward to getting the oscilloscope and more deeply
checking the operation of this computer. My plan is to get the CPU
board working first since that is where I think the majority of the
problem is. So far, I have identified about 6 chips with bad outputs
by using the wait state procedure and some liberal swapping of like
chips from unknown states into the known state locations. I have not
replaced all the buffers and bus transceivers yet but suspect before
it is all said and done they will be.

Certainly, when the manual calls out there should be $E800 on the
address lines and one bus transceiver chip gives rock solid .20v for
LOW and 3.4 for HIGH but the same chip right next to it is giving
values ranging from .20 to 1.5v for LOW something is clearly wrong,
especially when the suspect chips gives values like 2.0v for a HIGH
level when its swapped. It is for certain a bad chip when the problem
moves with it. I replace the suspect chip with a new one and recheck
all the values until they behave in a consistent and logical way. I
am using a bus extender and a plain S-100 backplane with no other
cards in it as the test chassis for the CPU board.

Unfortunately, there are limits to the static analysis and the
oscilloscope is going to shed a lot of light on the inner workings on
the board especially since it can do triggering on events, edges and
the like. I really look forward to it.

Right now, I believe the computer is almost booting. When it starts
up, I can see there is inconsistent activity on the floppy drives,
like it is trying to home the heads and spin the motor. Sometimes the
drive motors start and spin for 12 seconds or so, sometimes the drive
lights come on briefly. Other times there is rattling in the drive
and the head is jittering back and forth. The behavior is
inconsistent and is "choppy" in it rattles the drives for a moment
then quits. I suspect there is another bad part in the CPU board
which is causing all the inconsistent behavior but I am almost certain
that the CPU is trying to execute the FDC boot ROM code. I need to
confirm that however.

I followed the advice in this newsgroup on cleaning the floppy drives
out, cleaning off the head and head rails with alcohol and swabs. The
I put a little sewing machine oil on the head rails so the head can
move smoothly assuming the controller decided to tell it to do so.

So the way ahead is to just put the FDC and floppy drives aside for a
while and build a EPROM replacement circuit. I found a couple
DS1220's at the surplus store which the data*** claims are
compatible with the 2716 EPROM. It does have a similar pinout
configuration to the 2716 so my plan is to get a 24 pin socket and
build a "shim" socket following the instructions at the previous page
where the 2708 used to be. Basically, keeping the CPU boards pins 19
and 21 away from the DS1220 and making it work with the rest of the
pins in the socket. Then I can program the DS1220 on my EPROM
programmer, which supports NVRAM programming, and just plug it into
the shim socket. Presuming the DS1220's are still any good which is
pretty dicey after all this time.

http://www.maxim-ic.com/getds.cfm?qv_pk=2643

http://www.xs4all.nl/~ganswijk/chipdir/pin/2708.txt

http://www.xs4all.nl/~ganswijk/chipdir/pin/2716.txt

Once the 2708 is replaced, I will write a tight loop starting at $F800
so that the CPU just jumps back to that location repeatedly. I can
replace the auto-jump DIP header so that CPU board jumps to $F800
instead of $E800 for the FDC boot code PROM. Put this all together
and I should be able to boot the CPU board into a tight infinite loop
so I can use the oscilloscope to verify the control, address, and data
lines are behaving correctly on the CPU bus and move out from there.

Hopefully, after that shakes out the bugs in the CPU board, I'll put
it back into the NorthStar Horizon bus and start trying to get it to
talk to the serial port hardware. Dave Dunfield mentioned he had a
nice small ROM only monitor which maybe I could use or just modify the
one I used on my other Z80 home built computer to not use RAM
locations. However, I need to focus on one step at a time and getting
the CPU board working properly is top priority.

Dave, if you are reading this, would you email me that monitor please
so I can study the assembler source?

Unfortunately, I have to go on business travel for the next few days
so I will just bring the manuals along for some more re-reading and
maybe some new thoughts will come to me.

This weekend I will go to another surplus electronics store and get
some more replacement chips. The one I went to last weekend had most
of what I needed but only had 74F241's instead of 74LS241's as the
original design called for. I am using the 74F241's to get the
voltage settings right but am almost certain they are not going to be
properly compatible with the 74LS timing requirements. 74F TTL is
great stuff and really fast but I think it is way TOO fast for this
CPU and motherboard. 74LS is a lot slower and gentler timing wise. I
certainly would like to keep all the chips in the same TTL family if
possible although the original design did mix in a couple plain 74 TTL
and rest is 74LS.

Have a great week and hopefully I can post some more news this
weekend. Thanks!

Andrew Lynch

.