Re: [sebhc] Finally got a Z-90 computer

10 views
Skip to first unread message
Message has been deleted

Carroll Waddell

unread,
Jan 20, 2008, 8:27:32 AM1/20/08
to se...@googlegroups.com
Andrew Lynch wrote:
> Hi,
>
> I finally got a Heath/Zenith Z-90 computer. I meant to get an H-89 and I
> did buy one on Ebay but it was damaged in shipping and UPS returned it to
> the sender as part of the damage claim process.
>
> To help fix the H-89, I bought a Z-90 for spare parts. Since the H-89 is
> gone for good, I've decided to repair the Z-90 instead. It is actually in
> pretty good shape.
>
> I am starting to assess for repair the Z-90 and thought I'd update the SEBHC
> group and provide some details of what I am doing. Also ask for some advice
> and any pointers before I do something dumb.
>
> There is no structural damage on the Z-90 and when I power the system up,
> the terminal appears to work in "offline" mode. I can type letters on the
> keyboard and they appear on the screen. The video is reasonably stable and
> clear. The computer does NOT give any sort of power up message.
>
> The main CPU board appears to be dead but that is no surprise. I would like
> to start fault isolating. I believe the problem is in the main CPU board
> but I would like to confirm that theory.
>
> My plan is to first take a lot of digital photos which I did this evening.
> Then remove all the cards from the CPU board and see if that helps.
>
> The Z-90 has three peripherals attached; a serial card with three serial
> ports, a FDC controller based on a WDC 1797 attached to a Tandon TM100-2A as
> best I can tell, and a 16K RAM card with a socket cable jumper to the CPU
> board.
>
> It appears that the only connection between the main CPU board and the
> terminal board is through a serial link. Is that correct?
>
> Since the main CPU board is the primary suspect, what I'd like to do is take
> it out of the circuit and see if the rest works as it should. The plan is
> to confirm the "Z-19 terminal" portion still works by attaching the serial
> port from the terminal board directly to one of the serial ports in the back
> and communicating with a laptop.
>
> Are the serial port plugs compatible with the terminal board serial line
> plug on its back? They look similar but I notice the serial line plug on
> the back of the terminal board has an extra black line.
>
> I have fixed other vintage computers (Horizons, Vector Graphic, homebrew,
> etc) before and I have an oscilloscope, logic probe, and VOM. I have
> downloaded some of the Z-89 and Z-90 documentation and will be reviewing the
> schematics.
>
> The Zenith documentation looks really good. I wish all vintage computers
> were so well documented!
>
> Thanks!
>
> Andrew Lynch
>
>
> >
>
You're right about the serial connection being the only connection. I've
repaired 2 H-89's that I bought. Both had the same problem. Supply
voltages. Both CPU boards had blown tantalum caps, and one had melted
the land pattern coming in from the power supply connector. That's where
I'd start.
Carroll

dwight elvey

unread,
Jan 20, 2008, 12:26:20 PM1/20/08
to se...@googlegroups.com


> From: lyn...@yahoo.com
>
> Hi,
>
> I finally got a Heath/Zenith Z-90 computer. I meant to get an H-89 and I
> did buy one on Ebay but it was damaged in shipping and UPS returned it to
> the sender as part of the damage claim process.
>
> To help fix the H-89, I bought a Z-90 for spare parts. Since the H-89 is
> gone for good, I've decided to repair the Z-90 instead. It is actually in
> pretty good shape.
>
> I am starting to assess for repair the Z-90 and thought I'd update the SEBHC
> group and provide some details of what I am doing. Also ask for some advice
> and any pointers before I do something dumb.
>
> There is no structural damage on the Z-90 and when I power the system up,
> the terminal appears to work in "offline" mode. I can type letters on the
> keyboard and they appear on the screen. The video is reasonably stable and
> clear. The computer does NOT give any sort of power up message.
>
> The main CPU board appears to be dead but that is no surprise. I would like
> to start fault isolating. I believe the problem is in the main CPU board
> but I would like to confirm that theory.
>
> My plan is to first take a lot of digital photos which I did this evening.
> Then remove all the cards from the CPU board and see if that helps.
>
> The Z-90 has three peripherals attached; a serial card with three serial
> ports, a FDC controller based on a WDC 1797 attached to a Tandon TM100-2A as
> best I can tell, and a 16K RAM card with a socket cable jumper to the CPU
> board.
 
Hi
 The Z-90 is mostly identical to the H-89. There are slight differences in
the firmware. It seems it has the softsectored controller. Good and
bad. Most of the library we have for the 89 is hard sectored. Still,
CP/M is easier on the 90 with softsectored.
 
 As mentioned by Carroll, get supplies up first.
 
I have a method for finding blown tantalums but most don't understand
it but it is quite effective at finding short. I can describe it but until
one does it a few times, it is different than most expect.
 
 The connection between the two is just the serial.
Dwight


Helping your favorite cause is as easy as instant messaging. You IM, we give. Learn more.
Message has been deleted

Carroll Waddell

unread,
Jan 20, 2008, 1:38:56 PM1/20/08
to se...@googlegroups.com
Andrew Lynch wrote:
>
>
>> -----Original Message-----
>> From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
>> Carroll Waddell
>> Sent: Sunday, January 20, 2008 8:28 AM
>> To: se...@googlegroups.com
>> Subject: [sebhc] Re: Finally got a Z-90 computer
>>
>>
>> Andrew Lynch wrote:
>>
>>> Hi,
>>>
> [snip]

>
>> You're right about the serial connection being the only connection. I've
>> repaired 2 H-89's that I bought. Both had the same problem. Supply
>> voltages. Both CPU boards had blown tantalum caps, and one had melted
>> the land pattern coming in from the power supply connector. That's where
>> I'd start.
>> Carroll
>>
>>
> [snip]
> [AJL>]
> Hi Carroll,
>
> OK, tantalum capacitors are frequent suspects for failures. So far, I
> haven't had any luck in finding any culprits though.
>
> I've been checking the capacitors for shorts, burns, etc and also the
> regulators for proper voltages. They seem to be checking out OK so far.
>
> I pulled the 16K RAM card, serial card, and FDC card out so the CPU board is
> "bare". It didn't have any effect on operations though. I have been
> pressing RESET + RIGHT SHIFT to try to reset the CPU board but no effect. I
> am looking for that "H:" monitor prompt on the screen.
>
> After reading the service manual, it had a neat trick for determining if the
> terminal board serial port is working or not. I put a jumper between pins 3
> and 5 on the J404 of the terminal board to form a serial port "loop back"
> test.
>
> Sure enough, when reassembled and J404 not attached to the CPU board but
> with the jumper installed, I powered the system up and the terminal prints
> characters in either "offline" or "online" mode which confirms the terminal
> serial port is sending and receiving properly.
>
> The loop back test pretty much clears the terminal board as having the fault
> at least in my mind. I am now almost certain the problem is on the CPU
> board.
>
> I am going to continue checking the tantalum and electrolytic capacitors
> looking for a shorted unit. On other machines the bad tantalum capacitors
> were easy to find since they conveniently burst into flames or burned up.
> There was no question as to which one was bad. I had one capacitor on a
> Tandon TM100-2A floppy drive literally explode like miniature firecracker!
>
> Another quick technique is to remove all the ICs and wire brush the pins and
> return to their sockets. Sometimes when PCBs sit for extended periods they
> accumulate oxide enough to remove them from the circuit. I will try that
> next.
>
> I am bit concerned though on how to service this machine. I am really
> anxious working around the back end of the CRT and the flyback transformer.
> Those high voltage units are notorious for "reaching out and biting".
>
> How do you debug the CPU board with an oscilloscope so close to those high
> voltage sources? The power supply cables don't seem long enough to remove
> the CPU board entirely from the rear of the unit. Do you use extender
> cables?
>
> Any ideas greatly appreciated. Thanks in advance!
>
>
> Andrew Lynch
>
> PS, if anyone has a known good CPU board they'd let me buy or borrow, please
> let me know. I plan to fix this unit but some known good parts would surely
> help fault isolate with some good old fashioned "swaptronics". :-)
>
>
> >
>
Someone told me how to diagnose the CPU board. You take the screws out
that hold it in the card frame, then lay the bottom of the CPU board on
top of the CRT, and put the screws back in the top of the CPU board
through the card frame. That lays the CPU board flat on the top of the
CRT. The cable connecting the CPU board to the terminal board in short,
but it will reach (barely). Once on top, you can use a scope to check
signals on the board. When I diagnosed the ones I had, I removed all but
16K of memory chips, and got that working first, then kept adding 16K
until I got to 64K. You will have to use the configuration guide to know
how to set the jumpers on the board.
Carroll

Carroll Waddell

unread,
Jan 20, 2008, 1:40:16 PM1/20/08
to se...@googlegroups.com
Andrew Lynch wrote:
>
>
>> -----Original Message-----
>> From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
>> Carroll Waddell
>> Sent: Sunday, January 20, 2008 8:28 AM
>> To: se...@googlegroups.com
>> Subject: [sebhc] Re: Finally got a Z-90 computer
>>
>>
>> Andrew Lynch wrote:
>>
>>> Hi,
>>>
> [snip]
>
>> You're right about the serial connection being the only connection. I've
>> repaired 2 H-89's that I bought. Both had the same problem. Supply
>> voltages. Both CPU boards had blown tantalum caps, and one had melted
>> the land pattern coming in from the power supply connector. That's where
>> I'd start.
>> Carroll
>>
>>
PS. Swaptronics is a good diagnostic tool. When I used to diagnose large
IBM mainframes, I used "swaptronics" a lot.
Carroll
Message has been deleted

Peter

unread,
Jan 21, 2008, 12:52:18 PM1/21/08
to SEBHC
Andrew,

I have read all. I will recommend to remove the serial, floppy and 16K
memory board. YOU HAVE TO SET THE VARIOUS DIP SWITCHES FOR THIS
CASE !!! (write down at first the settings) - Keep in mind that you
have to select the correct memory size - in case you are removing the
chips - you need for the initial setup (H: prompt) ONLY 16K! In case
you don't get the H: with this minimum, you should check the addresss-
and data bus with the scope, if there is any activity. May be that one
or more of the lines are pulled low or high because of a faulty bus
buffer (74ls245, 74ls244, 74ls373 ???). There also two or three PALs
(programmed logic arrays) which differ with the part number according
to the part-number of the monitor ROM ! Read in the manual the various
chip numbers (customer specific - PALs and EPROMs) with the various
jumpers on the board. The first H88/H89 had eproms with 3 supply
voltages (+5,+12,-12V) and the later issues had single supply (+5V).
Check the wire jumpers on board in accordance to the EPROM types.
Concerning the 16K memory board for having total 64K you should notice
that the cable connection uses one of the ram sockets on the cpu
board!!! If you want ONLY 32K you have to remove one of the ram-chips
on the 16k board and you have to put it in the socket on the cpu
board.

REDUCE THE HARDWARE TO A MINIMUM! The system MUST come up to the H:
prompt with only 16K ram and without all intercaces (serial, floppy,
16k upgrade to 64k). In my time as technician in the Austrian Heath/
Zenith representative I had made all repairs on h8, h/z89/h90, z100
and I managed ALL faults. Only once I had a faulty CPU (Z80), but
PALs, memory chips and standard logic devices can become fault. The
supply voltages +5, +12 and -12V can easily be checked and then you
should check the reset on the cpu, the conditions on the address and
data-bus.

Good luck ... will watch your comments here
peter
On 21 Jan., 06:02, "Andrew Lynch" <lync...@yahoo.com> wrote:
> > -----Original Message-----
> > From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
> > Carroll Waddell
> > Sent: Sunday, January 20, 2008 1:40 PM
> > To: se...@googlegroups.com
> > Subject: [sebhc] Re: Finally got a Z-90 computer
>
> [snip]
> > > Andrew Lynch
>
> > > PS, if anyone has a known good CPU board they'd let me buy or borrow,
> > please
> > > let me know.  I plan to fix this unit but some known good parts would
> > surely
> > > help fault isolate with some good old fashioned "swaptronics".  :-)
>
> > PS. Swaptronics is a good diagnostic tool. When I used to diagnose large
> > IBM mainframes, I used "swaptronics" a lot.
> > Carroll
>
> [snip]
> [AJL>]
>
> Hi Carroll,
> Well, I worked on the Z-90 today and found out a few more things.  First, I
> have seen the reset circuit is working.  It outputs a brief pulse to reset
> the CPU rather than holding the CPU /RESET line down as long as the keys are
> pressed as I was expecting.  I was able to catch occasional glimpses of the
> reset pulse on my oscilloscope but the logic probe detected it reliably.
>
> I have been focusing on the power supply and tantalum capacitors but so far
> haven't found any bad ones.  Have also reseated and cleaned all the IC's but
> no change.  Rechecked all the jumpers and noticed the Z-90 CPU board was
> configured to boot from an 8" drive that I don't have so I changed that
> around but otherwise they were OK.
>
> The system EPROM is a 4K 2732 at U518 which made for some interesting jumper
> selections.  It also taps A11 off of the bus using a jumper wire which took
> quite a while to figure out.  I removed the EPROM and verified it was OK in
> my EPROM programmer.  The copy seemed to work the same as the original but
> neither has brought up the MTR yet.  Sometimes the EPROM chip select has a
> signal, sometimes not.  The CPU board does not seem to start up
> consistently.  There is chip at U520 which I presume is a 2K Floppy ROM from
> the documentation.  I didn't recognize the chip number so I presume it is a
> mask ROM of some sort.
>
> One thing I have noticed though about this computer I don't really like is
> that the peripherals modify the operation of the main board more
> significantly than I would have expected.  For instance, the floppy
> controller inserts itself into the interrupt circuit and if you remove the
> FDC, the /INT line goes low and stays there indefinitely.  I have to replace
> the FDC in order to get the CPU to work properly.
>
> I'm using the classic technique for finding CPU board problems by just going
> around the CPU, memories, etc examining states looking for anomalies.  There
> certainly is one on the Data #2 (D2) line.  It seems to be operating in some
> weird tristate mode or something is dragging it down.  The maximum voltage I
> am seeing on that line is less than 2 volts which makes me think it may be
> coupled to something.  I have removed all the parts I could find searching
> for a broken input or output gate but no luck yet.  Definitely something is
> weird on D2 but I don't know what.  The address lines seem to be operating
> normally.
>
> I suspect someone spilled something on the CPU board since all the chips
> were stuck in the sockets much more than normal.  Also there was a residue
> towards the top of the board.  I wiped it off but who knows where it seeped
> into.  The liquid probably went under the sockets and so finding the sneak
> circuit is going to be a challenge.
>
> Next, I am going to try swapping out some of the chips on the data bus to
> see if I can find the one pulling D2 down.
>
> Thanks!
>
> Andrew Lynch- Zitierten Text ausblenden -
>
> - Zitierten Text anzeigen -
Message has been deleted

Peter

unread,
Jan 21, 2008, 2:55:38 PM1/21/08
to SEBHC
Andrew,

Congratulations! A faulty CPU is rare, but in your case okay ... May
be that its a 4 MHz CPU instaed of a 2 Mhz ? ... anyway ... you found
it ... now you can proceed ...

peter

On 21 Jan., 20:27, "Andrew Lynch" <lync...@yahoo.com> wrote:
> > -----Original Message-----
> > From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of
> > Andrew Lynch
> > Sent: Monday, January 21, 2008 12:03 AM
> > To: se...@googlegroups.com
> > Subject: [sebhc] Re: Finally got a Z-90 computer
>
> [snip]
>
> > Next, I am going to try swapping out some of the chips on the data bus to
> > see if I can find the one pulling D2 down.
>
> > Thanks!
>
> > Andrew Lynch
>
> [AJL>]
>
> Hi, Sorry to reply to my own message but I have an important update!  I have
> gotten the MTR to run and now I can see and execute commands!  Yahoo!
>
> I knew there was something very odd about the data bus and I was searching
> and searching for the culprit.  I have literally traced out every single
> thing attached to the data bus looking broken components, bent leads, failed
> parts, burned traces, whatever would cause all the weird signals present
> there especially on D2.
>
> Alas, I finally found the real problem when I swapped the Z80 CPU out simply
> because it was the last unswapped part left on the data bus.  Yes, the CPU
> itself had failed!  That is a first for me.  Normally it is stuff like
> latches (74LS273) or buffers (74LS245, 74LS240, etc) or some bent trace or
> something.  
>
> I have literally never seen a CPU go bad before and it was the last thing I
> was expecting.  Maybe I have lead a sheltered life but on all my other stuff
> the Z80's have been practically indestructible.  It is literally the last
> place I would have expected to look for a failed part.
>
> I guess you learn something everyday!
>
> Thanks!
>
> Andrew Lynch
>
> PS, I am attaching a photo of the failed part for your viewing enjoyment.  
>
> :-)
>
>  Z90-08-sm.jpg
> 267KAnzeigenHerunterladen- Zitierten Text ausblenden -
Reply all
Reply to author
Forward
Message has been deleted
0 new messages