Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi Dwight

The machine is in fact equipped with three connectors, a DB25 for RS-232 
communication, a DB9 male for the keyboard and a third connector, 
another DB9, female this time, not sure but I think it was to connect a 
card reader.


Do you think the serial I/O channel B can share its resource for the 
RS-232 and the keyboard at the same time?
Clearly the problem is around the SIO B, not the SIO itself because I 
tested this IC by replacing it with another identical SIO with exactly 
the same result.


As said below, the "OPTIONAL RAM BANK 0.1: FAILED" is not the problem, 
the machine can start without this memory extension board. It was the 
case when the machine worked by the past, it was also the case when I 
was able to enter the page setup some weeks ago, when the failure was 
still intermittent and not permanent.


Note that once the default parameters are loaded into the non-volatile 
memory and manually completed, after a reset, some FAILEDs are replaced 
by "??", which can be considered as "assumed as absent ".


I will continue to check and repair the eventually damaged solder/tracks 
(battery leakage), this is what I have been doing for 2 weeks now, and 
I'm going to try a suggestion from Chuck : Put an external "loopback" 
connector to the RS-232 DB25 which is linked to the SIO channel B, just 
to see if this has any impact on the diagnosis.


If not there is no result, I always would like to induce to the 
self-test program that the channel B is OK by injecting something (but 
what ?) to the SIO (perhaps by acting on this pin 30 which is W/RDYB, 
something like "channel B Ready to write" right ?) in the idea to force 
a positive diagnosis in the self-test.


Dominique

On 18/10/2017 06:44, dwight via cctalk wrote:

By the way, how is the terminal and keyboard connected to the computer?

Is it by RS232?

If so, it might be through SIO B. If this is the case, it might be the problem.

Also, the RAM failure may be the issue as well. You have something that is 
intermittent there. Bad RAM can cause programs to misbehave as well. Just 
because it passes some times doesn't mean it is good enough to run programs.

Dwight



From: cctalk <cctalk-boun...@classiccmp.org> on behalf of dwight via cctalk 
<cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 4:08:46 PM
To: Dominique Carlier; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Since you know the pinout of the SIO chip, you might first look to see if where 
the rx and tx pins go. This may require some hunting with an ohm meter. I'd 
suspect they go to a RS232 level shifter.

You may also have to write some code to run the serial chip and any possible 
external loopback. As I recall these chips may have an internal loopback.

If you have a working logic analyzer, you can trigger on the select pin to the 
SIO and look to see what address the SIO is located at. That will allow you to 
create some debug code.

I'm not saying this is easy. Still, this is the way I'd attack the problem as a 
start.

You have to realize, there is not much more we can do for you.

Creating a ROM with a diagnostic looping program is about the only practical 
way to deal with a machine with no documentation.

I fixed an old mini once without schematics but it was all DTL and TTL and 
there were signal names at the card edges.

You have a Z80 computer. If you can program EPROMs you have a chance, otherwise 
it is unlikely that your current easter egg hunting method is likely to be very 
fruitful. You have already gone through most all of the likely failure items. 
From here you will likely have to begin to troubleshoot.

Dwight



From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Dominique Carlier via 
cctalk <cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 1:59:47 PM
To: Chuck Guzis; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Thanks for your response Chuck,

As described in the topic, just after the Power On, the machine runs a
self-test which is called "POC TEST" on the UTS range of Sperry Univac.
During this test, the machine checks the status of the RAM, the ROMs,
the Counter Timer, and the various communication interfaces.
Everything is OK except the ninth line which says: "SERIAL I / O CHANNEL
B: FAILED"

(ignore the FAILED next to "OPTIONAL RAM BANK 0,1", the machine can
start without this board)
http://www.zeltrax.com/classiccmp_forum/uts40_screen03.jpg

This channel B error completely freezes the machine, it does not load
the default settings, I no longer have access to the setup page, in this
situation the keyboard remains without effects.

When the POC test i

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[dwight via cctalk]

By the way, how is the terminal and keyboard connected to the computer?

Is it by RS232?

If so, it might be through SIO B. If this is the case, it might be the problem.

Also, the RAM failure may be the issue as well. You have something that is 
intermittent there. Bad RAM can cause programs to misbehave as well. Just 
because it passes some times doesn't mean it is good enough to run programs.

Dwight



From: cctalk <cctalk-boun...@classiccmp.org> on behalf of dwight via cctalk 
<cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 4:08:46 PM
To: Dominique Carlier; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Since you know the pinout of the SIO chip, you might first look to see if where 
the rx and tx pins go. This may require some hunting with an ohm meter. I'd 
suspect they go to a RS232 level shifter.

You may also have to write some code to run the serial chip and any possible 
external loopback. As I recall these chips may have an internal loopback.

If you have a working logic analyzer, you can trigger on the select pin to the 
SIO and look to see what address the SIO is located at. That will allow you to 
create some debug code.

I'm not saying this is easy. Still, this is the way I'd attack the problem as a 
start.

You have to realize, there is not much more we can do for you.

Creating a ROM with a diagnostic looping program is about the only practical 
way to deal with a machine with no documentation.

I fixed an old mini once without schematics but it was all DTL and TTL and 
there were signal names at the card edges.

You have a Z80 computer. If you can program EPROMs you have a chance, otherwise 
it is unlikely that your current easter egg hunting method is likely to be very 
fruitful. You have already gone through most all of the likely failure items. 
From here you will likely have to begin to troubleshoot.

Dwight



From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Dominique Carlier via 
cctalk <cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 1:59:47 PM
To: Chuck Guzis; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Thanks for your response Chuck,

As described in the topic, just after the Power On, the machine runs a
self-test which is called "POC TEST" on the UTS range of Sperry Univac.
During this test, the machine checks the status of the RAM, the ROMs,
the Counter Timer, and the various communication interfaces.
Everything is OK except the ninth line which says: "SERIAL I / O CHANNEL
B: FAILED"

(ignore the FAILED next to "OPTIONAL RAM BANK 0,1", the machine can
start without this board)
http://www.zeltrax.com/classiccmp_forum/uts40_screen03.jpg

This channel B error completely freezes the machine, it does not load
the default settings, I no longer have access to the setup page, in this
situation the keyboard remains without effects.

When the POC test is successful, it loads the default settings in non
volatile RAM:
http://www.zeltrax.com/classiccmp_forum/uts40_screen01.jpg

And I can access the setup page:
http://www.zeltrax.com/classiccmp_forum/uts40_screen02.jpg

The few times I was able to go into the setup page (CHANNEL B: PASSED),
I rushed to try to encode (with the dead keyboard) the data to declare
the subsystem and finally return to the CP/M mode. And I had twice the
case where without warning hop! Blank screen, automatic reset, self-test
(POC) -> SERIAL I / O CHANNEL B: FAILED (again).
Maybe we have there some interesting information about the problem, the
intermittent nature of this failure, could this give information about
the type of component in fault?

But note that since weeks now the problem has become permanent, I have
never been able to return to this famous setup page and the "serial I/O
channel B" is now always marked as 'FAILED'.

So I have no way to try anything using the terminal at this point.

It is possible that the breakdown is in fact very simple (dry solder dry
or attacked solder by the acid of the battery), but I would like to
avoiding to rework all the solders, and maybe to finally find that the
problem was at the level of an IC, I would try to locate the components
linked to this problem. I look at the SIO, try to discover what is after
just after, see how I can eventually act on the pin 30 (W/RDYB) of that IC :
http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg


And try to understand why this SIO no longer considers the channel B as
'READY'
This kind of things ...
There are probably programmable loop-back circuitry used by the POC test
program (in the ROM of the program cartridge).To ignore the negative
diagnostic I would like to induce to the self-test program that the
chann

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[dwight via cctalk]

Since you know the pinout of the SIO chip, you might first look to see if where 
the rx and tx pins go. This may require some hunting with an ohm meter. I'd 
suspect they go to a RS232 level shifter.

You may also have to write some code to run the serial chip and any possible 
external loopback. As I recall these chips may have an internal loopback.

If you have a working logic analyzer, you can trigger on the select pin to the 
SIO and look to see what address the SIO is located at. That will allow you to 
create some debug code.

I'm not saying this is easy. Still, this is the way I'd attack the problem as a 
start.

You have to realize, there is not much more we can do for you.

Creating a ROM with a diagnostic looping program is about the only practical 
way to deal with a machine with no documentation.

I fixed an old mini once without schematics but it was all DTL and TTL and 
there were signal names at the card edges.

You have a Z80 computer. If you can program EPROMs you have a chance, otherwise 
it is unlikely that your current easter egg hunting method is likely to be very 
fruitful. You have already gone through most all of the likely failure items. 
From here you will likely have to begin to troubleshoot.

Dwight



From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Dominique Carlier via 
cctalk <cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 1:59:47 PM
To: Chuck Guzis; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Thanks for your response Chuck,

As described in the topic, just after the Power On, the machine runs a
self-test which is called "POC TEST" on the UTS range of Sperry Univac.
During this test, the machine checks the status of the RAM, the ROMs,
the Counter Timer, and the various communication interfaces.
Everything is OK except the ninth line which says: "SERIAL I / O CHANNEL
B: FAILED"

(ignore the FAILED next to "OPTIONAL RAM BANK 0,1", the machine can
start without this board)
http://www.zeltrax.com/classiccmp_forum/uts40_screen03.jpg

This channel B error completely freezes the machine, it does not load
the default settings, I no longer have access to the setup page, in this
situation the keyboard remains without effects.

When the POC test is successful, it loads the default settings in non
volatile RAM:
http://www.zeltrax.com/classiccmp_forum/uts40_screen01.jpg

And I can access the setup page:
http://www.zeltrax.com/classiccmp_forum/uts40_screen02.jpg

The few times I was able to go into the setup page (CHANNEL B: PASSED),
I rushed to try to encode (with the dead keyboard) the data to declare
the subsystem and finally return to the CP/M mode. And I had twice the
case where without warning hop! Blank screen, automatic reset, self-test
(POC) -> SERIAL I / O CHANNEL B: FAILED (again).
Maybe we have there some interesting information about the problem, the
intermittent nature of this failure, could this give information about
the type of component in fault?

But note that since weeks now the problem has become permanent, I have
never been able to return to this famous setup page and the "serial I/O
channel B" is now always marked as 'FAILED'.

So I have no way to try anything using the terminal at this point.

It is possible that the breakdown is in fact very simple (dry solder dry
or attacked solder by the acid of the battery), but I would like to
avoiding to rework all the solders, and maybe to finally find that the
problem was at the level of an IC, I would try to locate the components
linked to this problem. I look at the SIO, try to discover what is after
just after, see how I can eventually act on the pin 30 (W/RDYB) of that IC :
http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg


And try to understand why this SIO no longer considers the channel B as
'READY'
This kind of things ...
There are probably programmable loop-back circuitry used by the POC test
program (in the ROM of the program cartridge).To ignore the negative
diagnostic I would like to induce to the self-test program that the
channel B is OK, what should I inject to the SIO (perhaps on this pin
30) to force a positive diagnosis?
It would be interesting, just to see if in that way I can take control
again on the machine, and check that the rest is working.

And yes, the SIO is working (as the CPU, the counter timer and the DMA
controller) because I replaced these IC by another ones with the same
faulty result.
http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg

Thanks for your help ;-)

Dominique


On 17/10/2017 20:22, Chuck Guzis via cctalk wrote:
> On 10/17/2017 10:50 AM, Dominique Carlier via cctalk wrote:
>> Hi Chuck,
>>
>> Yes I understand well. But the fact that the machines Z80 based were all
>> equipped with this famous serial I/O channel A and B,

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Thanks for your response Chuck,

As described in the topic, just after the Power On, the machine runs a 
self-test which is called "POC TEST" on the UTS range of Sperry Univac.
During this test, the machine checks the status of the RAM, the ROMs, 
the Counter Timer, and the various communication interfaces.
Everything is OK except the ninth line which says: "SERIAL I / O CHANNEL 
B: FAILED"


(ignore the FAILED next to "OPTIONAL RAM BANK 0,1", the machine can 
start without this board)

http://www.zeltrax.com/classiccmp_forum/uts40_screen03.jpg

This channel B error completely freezes the machine, it does not load 
the default settings, I no longer have access to the setup page, in this 
situation the keyboard remains without effects.


When the POC test is successful, it loads the default settings in non 
volatile RAM:

http://www.zeltrax.com/classiccmp_forum/uts40_screen01.jpg

And I can access the setup page:
http://www.zeltrax.com/classiccmp_forum/uts40_screen02.jpg

The few times I was able to go into the setup page (CHANNEL B: PASSED), 
I rushed to try to encode (with the dead keyboard) the data to declare 
the subsystem and finally return to the CP/M mode. And I had twice the 
case where without warning hop! Blank screen, automatic reset, self-test 
(POC) -> SERIAL I / O CHANNEL B: FAILED (again).
Maybe we have there some interesting information about the problem, the 
intermittent nature of this failure, could this give information about 
the type of component in fault?


But note that since weeks now the problem has become permanent, I have 
never been able to return to this famous setup page and the "serial I/O 
channel B" is now always marked as 'FAILED'.


So I have no way to try anything using the terminal at this point.

It is possible that the breakdown is in fact very simple (dry solder dry 
or attacked solder by the acid of the battery), but I would like to 
avoiding to rework all the solders, and maybe to finally find that the 
problem was at the level of an IC, I would try to locate the components 
linked to this problem. I look at the SIO, try to discover what is after 
just after, see how I can eventually act on the pin 30 (W/RDYB) of that IC :

http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg


And try to understand why this SIO no longer considers the channel B as 
'READY'

This kind of things ...
There are probably programmable loop-back circuitry used by the POC test 
program (in the ROM of the program cartridge).To ignore the negative 
diagnostic I would like to induce to the self-test program that the 
channel B is OK, what should I inject to the SIO (perhaps on this pin 
30) to force a positive diagnosis?
It would be interesting, just to see if in that way I can take control 
again on the machine, and check that the rest is working.


And yes, the SIO is working (as the CPU, the counter timer and the DMA 
controller) because I replaced these IC by another ones with the same 
faulty result.

http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg

Thanks for your help ;-)

Dominique


On 17/10/2017 20:22, Chuck Guzis via cctalk wrote:

On 10/17/2017 10:50 AM, Dominique Carlier via cctalk wrote:

Hi Chuck,

Yes I understand well. But the fact that the machines Z80 based were all
equipped with this famous serial I/O channel A and B, I therefore
thought that the principle of verification of these channels would
probably be the same on this type of architecture (Z80+PIO+CTC+SIO).
Therefore, there should be probably more people able to give me some
useful information.

Perhaps, but we don't know exactly what surrounds the Z80 SIO, or
exactly what the diagnostic is complaining about.   Does your SIO have
anything other than line drivers or receivers on its external interface?
  Some systems have programmable loop-back circuitry to enable the
terminal to function to talk to itself and verify functionality.

If you ignore the diagnostic message and feed the terminal some serial
data, do the inputs on the SIO wiggle appropriately in response?  In
other words, is the data getting from the connector to the SIO chip?

Troubleshooting is slow, methodical work.

The SIO/DART chip itself is very simple--and most likely not the cause
of the diagnostic failure.  But writing your own diagnostic software can
verify that.

At least that's what I think from a few thousand km away.

--Chuck

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[dwight via cctalk]

You can not fix anything without knowing what it is suppose to do. Chuck's idea 
is sound. You will not likely get much with the logic analyzer unless the 
processor is actually running some code.

It doesn't sound like it is. You need to check that it is.

Dwight




From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Dominique Carlier via 
cctalk <cctalk@classiccmp.org>
Sent: Tuesday, October 17, 2017 10:50:29 AM
To: Chuck Guzis via cctalk
Subject: Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette 
subsystem : Restoration

Hi Chuck,

Yes I understand well. But the fact that the machines Z80 based were all
equipped with this famous serial I/O channel A and B, I therefore
thought that the principle of verification of these channels would
probably be the same on this type of architecture (Z80+PIO+CTC+SIO).
Therefore, there should be probably more people able to give me some
useful information.

Dominique


On 17/10/2017 19:26, Chuck Guzis via cctalk wrote:
> On 10/17/2017 04:56 AM, Dominique Carlier via cctalk wrote:
>> Hi guys!
>>
>> Nobody here has any information to help me to solve my problem?
>> Do you think I should talk about this breakdown on another forum? If
>> yes, have you an address to recommend me?
>>
> Dominique, the probably reason that many of us don't jump in is that
> you've got a tough job, given the house-numbered ICs and lack of visible
> traces.
>
> If this were my system and I was determined to get it working, I'd
> probably start by dumping and disassembling the ROMs to find out exactly
> what set of program steps occur to produce the error message.
>
> I'd probably then write a diagnostic ROM that would allow me to probe in
> detail to characterize the fault.  The happy circumstance is that the
> display (and possibly the keyboard) is functional.)
>
> Then I'd jump in with a logic analyzer or ICE to determine the exact
> nature of the failure and its source.
>
> You can see that this is an arduous process that few are equipped to
> help you diagnose remotely.
>
> My two cents' worth,
> Chuck
>
>

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Chuck Guzis via cctalk]

On 10/17/2017 10:50 AM, Dominique Carlier via cctalk wrote:
> Hi Chuck,
> 
> Yes I understand well. But the fact that the machines Z80 based were all
> equipped with this famous serial I/O channel A and B, I therefore
> thought that the principle of verification of these channels would
> probably be the same on this type of architecture (Z80+PIO+CTC+SIO).
> Therefore, there should be probably more people able to give me some
> useful information.

Perhaps, but we don't know exactly what surrounds the Z80 SIO, or
exactly what the diagnostic is complaining about.   Does your SIO have
anything other than line drivers or receivers on its external interface?
 Some systems have programmable loop-back circuitry to enable the
terminal to function to talk to itself and verify functionality.

If you ignore the diagnostic message and feed the terminal some serial
data, do the inputs on the SIO wiggle appropriately in response?  In
other words, is the data getting from the connector to the SIO chip?

Troubleshooting is slow, methodical work.

The SIO/DART chip itself is very simple--and most likely not the cause
of the diagnostic failure.  But writing your own diagnostic software can
verify that.

At least that's what I think from a few thousand km away.

--Chuck

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi Chuck,

Yes I understand well. But the fact that the machines Z80 based were all 
equipped with this famous serial I/O channel A and B, I therefore 
thought that the principle of verification of these channels would 
probably be the same on this type of architecture (Z80+PIO+CTC+SIO). 
Therefore, there should be probably more people able to give me some 
useful information.


Dominique


On 17/10/2017 19:26, Chuck Guzis via cctalk wrote:

On 10/17/2017 04:56 AM, Dominique Carlier via cctalk wrote:

Hi guys!

Nobody here has any information to help me to solve my problem?
Do you think I should talk about this breakdown on another forum? If
yes, have you an address to recommend me?


Dominique, the probably reason that many of us don't jump in is that
you've got a tough job, given the house-numbered ICs and lack of visible
traces.

If this were my system and I was determined to get it working, I'd
probably start by dumping and disassembling the ROMs to find out exactly
what set of program steps occur to produce the error message.

I'd probably then write a diagnostic ROM that would allow me to probe in
detail to characterize the fault.  The happy circumstance is that the
display (and possibly the keyboard) is functional.)

Then I'd jump in with a logic analyzer or ICE to determine the exact
nature of the failure and its source.

You can see that this is an arduous process that few are equipped to
help you diagnose remotely.

My two cents' worth,
Chuck

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi Mattis !

I'm afraid I do not have the required skills :-/

I have never used this kind of tools. However I recovered this recently:
http://actingmachines.com/img/photos/package/europacorp/tektronix_1240_b.jpg 



But in the present state of my knowledge, I could not use it. Is it 
difficult for a novice to learn how to use it?


If I could just know what is the principle of the verification system of 
the PIO's channels A & B, and also what are the integrated circuits 
concerned? That would be valuable information.


Dominique

On 17/10/2017 17:09, Mattis Lind wrote:



2017-10-17 13:56 GMT+02:00 Dominique Carlier via cctalk 
>:


Hi guys!

Nobody here has any information to help me to solve my problem?
Do you think I should talk about this breakdown on another forum?
If yes, have you an address to recommend me?

Thanks

Dominique



I don't know anything about the system you are about to repair but it 
make use of a standard CPU which is kind of useful when the schematic 
is missing. If I were you I would attach my trusty logic analyzer over 
the CPU to try to trace what it is doing. Which ROM does it access 
when it is doing the power up self test? Use a disassmbler on that 
particular ROM and work out what flow the CPU takes in the ROM 
contents and try to understand what causes it to take the decision 
that something is broken.


Since my logic analyzer has fairly limited memory depth I would have 
it to trigger on for example accessing the SIO.


/Mattis




On 15/10/2017 23:12, Dominique Carlier via cctalk wrote:

Hi everybody,

As I have no documentation / schematics I started to look for
computer schematics where it is about Z80 and “serial I/O
channel B”, I found a lot of data about this subject that I do
not know at all.
These information would allow me to better target the
breakdown, if any of you would consent to pass on some of his
knowledge.
In all documentation on Z80 based machines, I find the same
elements, similar architectures: CPU, PIO, CTC, SIO.

http://www.z80.info/gfx/nascom1main.gif

http://www.ep128.hu/Sp_Konyv/Pic/SIO_11.gif


I interchanged with the IC of my other CPU board which are of
the same model.
http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg


But the error persists, so the problem is around.
I'm trying to find out why the SIO detects a problem on the
channel B. How does this detection system work? What are the
solicited ICs? Do they communicate through other ICs like
multiplexers or others?

I would also like to simulate the absence of this problem in
order to see if this allows my UTS to have a successful POC
test, to have again access to the setup page of the machine.
Is it possible to do something at the SIO pin 30 (W/RDYB) to
pretend that the B channel is OK?
http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg


Can we imagine that the problem would be at the level of what
checks channel B and not channel B itself?
I also read a lot of obscure thing about memory, and I still
wonder if a memory problem could also play a part in this
failure? You must understand that I continue to try to
troubleshoot this machine like the others, comparative logic
without advanced academic knowledge, a challenge every time.

It's even harder because I have trouble identifying the
components. I do not find any results concerning the 3/4 of
the ICs, as for example these:
http://www.zeltrax.com/classiccmp_forum/sio_issue_02.jpg


And even for components like this, "AVX 224Z 8238"
http://www.zeltrax.com/classiccmp_forum/sio_issue_03.jpg


(ceramic capacitor of 220nf 50v? not sure)

There is also the fact that the tracks are not visible on the
CPU board which I try to repair, so I try to locate myself by
comparing component location to the original CPU board, in
short a beautiful mess.
Meanwhile, I restore this beautiful Key Tronics keyboard which
was in a very sad condition. Here are some pictures of this
restoration (work in progress)

http://www.zeltrax.com/classiccmp_forum/keyboard2_01.jpg


http://www.zeltrax.com/classiccmp_forum/keyboard2_02.jpg

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Chuck Guzis via cctalk]

On 10/17/2017 04:56 AM, Dominique Carlier via cctalk wrote:
> Hi guys!
> 
> Nobody here has any information to help me to solve my problem?
> Do you think I should talk about this breakdown on another forum? If
> yes, have you an address to recommend me?
> 

Dominique, the probably reason that many of us don't jump in is that
you've got a tough job, given the house-numbered ICs and lack of visible
traces.

If this were my system and I was determined to get it working, I'd
probably start by dumping and disassembling the ROMs to find out exactly
what set of program steps occur to produce the error message.

I'd probably then write a diagnostic ROM that would allow me to probe in
detail to characterize the fault.  The happy circumstance is that the
display (and possibly the keyboard) is functional.)

Then I'd jump in with a logic analyzer or ICE to determine the exact
nature of the failure and its source.

You can see that this is an arduous process that few are equipped to
help you diagnose remotely.

My two cents' worth,
Chuck

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Mattis Lind via cctalk]

2017-10-17 13:56 GMT+02:00 Dominique Carlier via cctalk <
cctalk@classiccmp.org>:

> Hi guys!
>
> Nobody here has any information to help me to solve my problem?
> Do you think I should talk about this breakdown on another forum? If yes,
> have you an address to recommend me?
>
> Thanks
>
> Dominique



I don't know anything about the system you are about to repair but it make
use of a standard CPU which is kind of useful when the schematic is
missing. If I were you I would attach my trusty logic analyzer over the CPU
to try to trace what it is doing. Which ROM does it access when it is doing
the power up self test? Use a disassmbler on that particular ROM and work
out what flow the CPU takes in the ROM contents and try to understand what
causes it to take the decision that something is broken.

Since my logic analyzer has fairly limited memory depth I would have it to
trigger on for example accessing the SIO.

/Mattis

>
>
>
> On 15/10/2017 23:12, Dominique Carlier via cctalk wrote:
>
>> Hi everybody,
>>
>> As I have no documentation / schematics I started to look for computer
>> schematics where it is about Z80 and “serial I/O channel B”, I found a lot
>> of data about this subject that I do not know at all.
>> These information would allow me to better target the breakdown, if any
>> of you would consent to pass on some of his knowledge.
>> In all documentation on Z80 based machines, I find the same elements,
>> similar architectures: CPU, PIO, CTC, SIO.
>>
>> http://www.z80.info/gfx/nascom1main.gif
>> http://www.ep128.hu/Sp_Konyv/Pic/SIO_11.gif
>>
>> I interchanged with the IC of my other CPU board which are of the same
>> model.
>> http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg
>>
>> But the error persists, so the problem is around.
>> I'm trying to find out why the SIO detects a problem on the channel B.
>> How does this detection system work? What are the solicited ICs? Do they
>> communicate through other ICs like multiplexers or others?
>>
>> I would also like to simulate the absence of this problem in order to see
>> if this allows my UTS to have a successful POC test, to have again access
>> to the setup page of the machine.
>> Is it possible to do something at the SIO pin 30 (W/RDYB) to pretend that
>> the B channel is OK?
>> http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg
>>
>> Can we imagine that the problem would be at the level of what checks
>> channel B and not channel B itself?
>> I also read a lot of obscure thing about memory, and I still wonder if a
>> memory problem could also play a part in this failure? You must understand
>> that I continue to try to troubleshoot this machine like the others,
>> comparative logic without advanced academic knowledge, a challenge every
>> time.
>>
>> It's even harder because I have trouble identifying the components. I do
>> not find any results concerning the 3/4 of the ICs, as for example these:
>> http://www.zeltrax.com/classiccmp_forum/sio_issue_02.jpg
>>
>> And even for components like this, "AVX 224Z 8238"
>> http://www.zeltrax.com/classiccmp_forum/sio_issue_03.jpg
>>
>> (ceramic capacitor of 220nf 50v? not sure)
>>
>> There is also the fact that the tracks are not visible on the CPU board
>> which I try to repair, so I try to locate myself by comparing component
>> location to the original CPU board, in short a beautiful mess.
>> Meanwhile, I restore this beautiful Key Tronics keyboard which was in a
>> very sad condition. Here are some pictures of this restoration (work in
>> progress)
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_01.jpg
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_02.jpg
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_03.jpg
>> 
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_04.jpg
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_05.jpg <
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_05.jpg>
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_06.jpg <
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_06.jpg>
>>
>> http://www.zeltrax.com/classiccmp_forum/keyboard2_07.jpg
>>
>> In any case I hold on, restarting this machine has become an obsession,
>> but without help I will still be on it for the next 10 years, thus  HELP!
>> ;-)
>>
>>
>> Everything would be perfectly fine if most of the time I did not have at
>>> startup an error at line 9. of the POC test:
>>>

> SERIAL I / O CHANNEL B: FAILED
>


>>
>>
>

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi guys!

Nobody here has any information to help me to solve my problem?
Do you think I should talk about this breakdown on another forum? If 
yes, have you an address to recommend me?


Thanks

Dominique


On 15/10/2017 23:12, Dominique Carlier via cctalk wrote:

Hi everybody,

As I have no documentation / schematics I started to look for computer 
schematics where it is about Z80 and “serial I/O channel B”, I found a 
lot of data about this subject that I do not know at all.
These information would allow me to better target the breakdown, if 
any of you would consent to pass on some of his knowledge.
In all documentation on Z80 based machines, I find the same elements, 
similar architectures: CPU, PIO, CTC, SIO.


http://www.z80.info/gfx/nascom1main.gif
http://www.ep128.hu/Sp_Konyv/Pic/SIO_11.gif

I interchanged with the IC of my other CPU board which are of the same 
model.

http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg

But the error persists, so the problem is around.
I'm trying to find out why the SIO detects a problem on the channel B. 
How does this detection system work? What are the solicited ICs? Do 
they communicate through other ICs like multiplexers or others?


I would also like to simulate the absence of this problem in order to 
see if this allows my UTS to have a successful POC test, to have again 
access to the setup page of the machine.
Is it possible to do something at the SIO pin 30 (W/RDYB) to pretend 
that the B channel is OK?

http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg

Can we imagine that the problem would be at the level of what checks 
channel B and not channel B itself?
I also read a lot of obscure thing about memory, and I still wonder if 
a memory problem could also play a part in this failure? You must 
understand that I continue to try to troubleshoot this machine like 
the others, comparative logic without advanced academic knowledge, a 
challenge every time.


It's even harder because I have trouble identifying the components. I 
do not find any results concerning the 3/4 of the ICs, as for example 
these:

http://www.zeltrax.com/classiccmp_forum/sio_issue_02.jpg

And even for components like this, "AVX 224Z 8238"
http://www.zeltrax.com/classiccmp_forum/sio_issue_03.jpg

(ceramic capacitor of 220nf 50v? not sure)

There is also the fact that the tracks are not visible on the CPU 
board which I try to repair, so I try to locate myself by comparing 
component location to the original CPU board, in short a beautiful mess.
Meanwhile, I restore this beautiful Key Tronics keyboard which was in 
a very sad condition. Here are some pictures of this restoration (work 
in progress)


http://www.zeltrax.com/classiccmp_forum/keyboard2_01.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_02.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_03.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_04.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_05.jpg 



http://www.zeltrax.com/classiccmp_forum/keyboard2_06.jpg 



http://www.zeltrax.com/classiccmp_forum/keyboard2_07.jpg

In any case I hold on, restarting this machine has become an 
obsession, but without help I will still be on it for the next 10 
years, thus  HELP! ;-)



Everything would be perfectly fine if most of the time I did not have 
at startup an error at line 9. of the POC test:


SERIAL I / O CHANNEL B: FAILED 

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi everybody,

As I have no documentation / schematics I started to look for computer 
schematics where it is about Z80 and “serial I/O channel B”, I found a 
lot of data about this subject that I do not know at all.
These information would allow me to better target the breakdown, if any 
of you would consent to pass on some of his knowledge.
In all documentation on Z80 based machines, I find the same elements, 
similar architectures: CPU, PIO, CTC, SIO.


http://www.z80.info/gfx/nascom1main.gif
http://www.ep128.hu/Sp_Konyv/Pic/SIO_11.gif

I interchanged with the IC of my other CPU board which are of the same 
model.

http://www.zeltrax.com/classiccmp_forum/sio_issue_01.jpg

But the error persists, so the problem is around.
I'm trying to find out why the SIO detects a problem on the channel B. 
How does this detection system work? What are the solicited ICs? Do they 
communicate through other ICs like multiplexers or others?


I would also like to simulate the absence of this problem in order to 
see if this allows my UTS to have a successful POC test, to have again 
access to the setup page of the machine.
Is it possible to do something at the SIO pin 30 (W/RDYB) to pretend 
that the B channel is OK?

http://www.sbprojects.net/projects/izabella/assets/images/sio1.jpg

Can we imagine that the problem would be at the level of what checks 
channel B and not channel B itself?
I also read a lot of obscure thing about memory, and I still wonder if a 
memory problem could also play a part in this failure? You must 
understand that I continue to try to troubleshoot this machine like the 
others, comparative logic without advanced academic knowledge, a 
challenge every time.


It's even harder because I have trouble identifying the components. I do 
not find any results concerning the 3/4 of the ICs, as for example these:

http://www.zeltrax.com/classiccmp_forum/sio_issue_02.jpg

And even for components like this, "AVX 224Z 8238"
http://www.zeltrax.com/classiccmp_forum/sio_issue_03.jpg

(ceramic capacitor of 220nf 50v? not sure)

There is also the fact that the tracks are not visible on the CPU board 
which I try to repair, so I try to locate myself by comparing component 
location to the original CPU board, in short a beautiful mess.
Meanwhile, I restore this beautiful Key Tronics keyboard which was in a 
very sad condition. Here are some pictures of this restoration (work in 
progress)


http://www.zeltrax.com/classiccmp_forum/keyboard2_01.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_02.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_03.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_04.jpg

http://www.zeltrax.com/classiccmp_forum/keyboard2_05.jpg 



http://www.zeltrax.com/classiccmp_forum/keyboard2_06.jpg 



http://www.zeltrax.com/classiccmp_forum/keyboard2_07.jpg

In any case I hold on, restarting this machine has become an obsession, 
but without help I will still be on it for the next 10 years, thus  
HELP! ;-)



Everything would be perfectly fine if most of the time I did not have 
at startup an error at line 9. of the POC test:


SERIAL I / O CHANNEL B: FAILED 

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi All !

I started checking / replacing all the capacitors - tantalum and 
electrolytic - on the CPU board saved from the US. Until now I don't 
found the guilty.
The CPU board of the USA is a horror to troubleshoot, because of the 
acid the battery that changed the nature of solder tin and some kind of 
resin characteristic of these PCBs, it is difficult to desolder the 
components. I wonder if I would do better to repair my own CPU board 
because this one is finally in a better state in comparison.


What is raging is that these two types of breakdownswere perfectly well 
identified by Sperry Univac. According to the UTS 20 manual :


- Blank screen and long tone (the breakdown of my CPU board) : I'm 
supposed to report this information to my Sperry Univac customer 
services, point, no explanation


- SERIAL I/O CHANNEL B failure (the breakdown of the CPU board from US) :
Their advice = Corrective action: none, and I'm supposed to transmit the 
code "3560U-0010" (still to my Sperry Univac customer services).


Zero information, nothing, it is raging. One or two clues to orient my 
research would be so precious...


I have the corresponding manual but for the UTS 40 somewhere lost in my 
attic. I will find it, I do not have the hope to find a better 
description but I must be sure.
On this subject, I would like to scan and share it because it does not 
exist on the internet.


Dominique


On 2/10/2017 18:13, Dominique Carlier via cctalk wrote:

Hi everybody,

Here is the next episode of the restoration of my UTS 40.

During the previous “season”, I was talking about buying a wreck of an 
UTS 40 from the USA.

http://www.zeltrax.com/classiccmp_forum/annonce.jpg

 I had taken the risk to buy it, after the conclusion of the sale, the 
seller had retracted, I had argued with him, I finally put a bad note 
on Ebay. However, surprise! The guy still sends me the machine. This 
one remains blocked during a full month at the Belgian custom house. I 
finally receive it, and now, if you like horror movies, you will be 
happy.


Here is the state in which I received the machine:
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_01.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_02.jpg

I never - never! - received a package in this condition !! The reason 
maybe was the hideous and ridiculously “protective” packaging ? 
Possible ... or in the plane they 400kg over my UTS ? Possible ... but 
we still have the impression that someone hammered the machine, even 
the very solid plastic pieces have also shattered.


As I am an optimistic guy, I think "the plastic ... not important, 
fortunately the chassis and the electronics of this machine is build 
like a tank".


I analyze the boards, as expected, some architectural differences but 
not so much. Here is the CPU board:

http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_03.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_06.jpg

Unsurprisingly, the battery has sunk and there are damages:
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_04.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_05.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_07.jpg

With courage I clean the board, I replace some resistors, one 
capacitor, ...

Before/After
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_08.jpg

and I replace the battery (2,4V ? A doubt persist)
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_10.jpg

And comes the time of the first test and it boot!
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_09.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_09b.jpg

The POC test at startup is not successfull but from the outset, this 
allows me to analyze and diagnose some elements of my own UTS 40, here 
is what I notice:


- The program cartridge of my UTS 40 is functional
- The power supply to my UTS 40 is OK
- The video circuit of my UTS is OK
- The keyboard of my UTS requires a restoration (aluminum discs, foam 
cylinder, classic issue of old keytronic keyboard)

http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_11.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_12.jpg

- The absence of the backup battery is not a problem, the machine can 
boot, without this it can successfully passing the POC test and the 
encoding of the parameters in the setup page is possible.


I discover other interesting things about the relationship between the 
cartridge program and the hardware configuration.


- These program cartridge are made for a specific hardware 
configuration, for example the RAM repartitions at the level of 
hardware has an impact on the result of the startup test. The number 
of kilobytes checked, the memory extension considered as PASSED or 
FAILED. Even the installed memory is working and assignable, with a 
configuration that differs from the hardware intended to run with a 
specific Program Cartridge, the POC test could show that a memory is 
missing, simply because 

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Camiel Vanderhoeven via cctalk]

On 8/13/17, 11:15 PM, "cctech on behalf of Dominique Carlier via cctech"
 wrote:


>On 12/08/2017 19:35, Al Kossow via cctalk wrote:
>> the black one in the lower left
>>
>> it is a tantalum, which are known to short, try removing them
>>
>
>Yes ! It was this one that was shorted !
>I decided in the same way to repair the second disk drive but this time
>by taking some pictures.

It would be a good idea to replace all tantalum capacitors; if two
shorted, chances are others are on their way out soon, too. Sometimes,
tantalum capacitor failure is paired with a nasty epxlosion that can make
a mess of things.

Camiel

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

You're probably right, but the work is so massive, and also this time I 
would like to understand and target this breakdown instead of working 
(as usual) in blind mode.
In addition, I do not see the same type of capacitor on the CPU board as 
the one from the subsystem you quote as being known to break down.


Do you think that the small green capacitors here
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_08.jpg

can be as fragile as this model there?
http://www.zeltrax.com/classiccmp_forum/uts40/8406repair07.jpg

Dominique

On 2/10/2017 19:05, Camiel Vanderhoeven via cctech wrote:

On 8/13/17, 11:15 PM, "cctech on behalf of Dominique Carlier via cctech"
 wrote:



On 12/08/2017 19:35, Al Kossow via cctalk wrote:

the black one in the lower left

it is a tantalum, which are known to short, try removing them


Yes ! It was this one that was shorted !
I decided in the same way to repair the second disk drive but this time
by taking some pictures.

It would be a good idea to replace all tantalum capacitors; if two
shorted, chances are others are on their way out soon, too. Sometimes,
tantalum capacitor failure is paired with a nasty epxlosion that can make
a mess of things.

Camiel

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Yes indeed, on the "Peripheral Interface" connector of the US 
configuration there was a terminator which seems indicated that nothing 
was connected there. Moreover, with the Program Cartridge US, the 
terminal does not recognize the PERIPHERAL I/O board of my original 
configuration.
This is why I use the CPU board US, its memory extension but with my 
Program Cartridge and my PERIPHERAL I/O board on which the subsystem 
8406 was connected.
It is not impossible that the PERIPHERAL I/O board US is also able to 
manage the 8406 Subsystem, the board is very elaborate, 4 or 5 Zilog 80, 
and two rows of dip switches, but without documentation ... I prefer to 
try with the board that worked with the subsystem.


It remains to be discovered what the "SERIAL I/O CHANNEL B" is, find the 
related component and the breakdown, and there is a good chance that I 
can restart this venerable machine.
But I search everywhere, I have not found any information that explains 
what it is.


Dominique


On 2/10/2017 21:28, jim stephens via cctalk wrote:



On 10/2/2017 9:13 AM, Dominique Carlier via cctalk wrote:



Everything would be perfectly fine if most of the time I did not have 
at startup an error at line 9. of the POC test:


SERIAL I / O CHANNEL B: FAILED 
I doubt the US unit you bought was used with a floppy running CPM. It 
most likely had a synchronous channel for connection to some network 
and ran just standalone.


I recall that there were modules of some sort you added to get some 
functionality, so as to not have to open the box.  I don't know if you 
set anything outside the box, but it may have some switches or such to 
indicate some other device is present that now is not.  So it would 
fail on power on test.


that's just a guess on which way I'd go to figure it out.  I don't 
have any documentation or anything other than having had one for about 
a year for a project I did to go on.


Mine had no floppies, but had a printer interface option attached 
which was the same as one that was on a Univac 1100 series mainframe I 
was making a controller for.  So I could  just run a standalone test 
and if my controller card was working it would spew on the printer.
That option was also one of these things added on.  Never dug into 
anything as the device is pretty overengineered and without a lot of 
manuals and parts, you can't do much with it.


thanks
Jim

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[jim stephens via cctalk]

On 10/2/2017 9:13 AM, Dominique Carlier via cctalk wrote:



Everything would be perfectly fine if most of the time I did not have 
at startup an error at line 9. of the POC test:


SERIAL I / O CHANNEL B: FAILED 
I doubt the US unit you bought was used with a floppy running CPM. It 
most likely had a synchronous channel for connection to some network and 
ran just standalone.


I recall that there were modules of some sort you added to get some 
functionality, so as to not have to open the box.  I don't know if you 
set anything outside the box, but it may have some switches or such to 
indicate some other device is present that now is not.  So it would fail 
on power on test.


that's just a guess on which way I'd go to figure it out.  I don't have 
any documentation or anything other than having had one for about a year 
for a project I did to go on.


Mine had no floppies, but had a printer interface option attached which 
was the same as one that was on a Univac 1100 series mainframe I was 
making a controller for.  So I could  just run a standalone test and if 
my controller card was working it would spew on the printer.
That option was also one of these things added on.  Never dug into 
anything as the device is pretty overengineered and without a lot of 
manuals and parts, you can't do much with it.


thanks
Jim

The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi everybody,

Here is the next episode of the restoration of my UTS 40.

During the previous “season”, I was talking about buying a wreck of an 
UTS 40 from the USA.

http://www.zeltrax.com/classiccmp_forum/annonce.jpg

 I had taken the risk to buy it, after the conclusion of the sale, the 
seller had retracted, I had argued with him, I finally put a bad note on 
Ebay. However, surprise! The guy still sends me the machine. This one 
remains blocked during a full month at the Belgian custom house. I 
finally receive it, and now, if you like horror movies, you will be happy.


Here is the state in which I received the machine:
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_01.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_02.jpg

I never - never! - received a package in this condition !! The reason 
maybe was the hideous and ridiculously “protective” packaging ? Possible 
... or in the plane they 400kg over my UTS ? Possible ... but we still 
have the impression that someone hammered the machine, even the very 
solid plastic pieces have also shattered.


As I am an optimistic guy, I think "the plastic ... not important, 
fortunately the chassis and the electronics of this machine is build 
like a tank".


I analyze the boards, as expected, some architectural differences but 
not so much. Here is the CPU board:

http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_03.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_06.jpg

Unsurprisingly, the battery has sunk and there are damages:
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_04.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_05.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_07.jpg

With courage I clean the board, I replace some resistors, one capacitor, ...
Before/After
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_08.jpg

and I replace the battery (2,4V ? A doubt persist)
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_10.jpg

And comes the time of the first test and it boot!
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_09.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_09b.jpg

The POC test at startup is not successfull but from the outset, this 
allows me to analyze and diagnose some elements of my own UTS 40, here 
is what I notice:


- The program cartridge of my UTS 40 is functional
- The power supply to my UTS 40 is OK
- The video circuit of my UTS is OK
- The keyboard of my UTS requires a restoration (aluminum discs, foam 
cylinder, classic issue of old keytronic keyboard)

http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_11.jpg
http://www.zeltrax.com/classiccmp_forum/uts40_from_usa_12.jpg

- The absence of the backup battery is not a problem, the machine can 
boot, without this it can successfully passing the POC test and the 
encoding of the parameters in the setup page is possible.


I discover other interesting things about the relationship between the 
cartridge program and the hardware configuration.


- These program cartridge are made for a specific hardware 
configuration, for example the RAM repartitions at the level of hardware 
has an impact on the result of the startup test. The number of kilobytes 
checked, the memory extension considered as PASSED or FAILED. Even the 
installed memory is working and assignable, with a configuration that 
differs from the hardware intended to run with a specific Program 
Cartridge, the POC test could show that a memory is missing, simply 
because the RAM is not physically in the slot or socket according to the 
program specification in the cartridge.
I drew this conclusion because the CPU board form USA has 64Kb installed 
on it, and 27 empty sockets, I filled them with RAM and now with the 
program cartridge form Ebay, 4X  64KB are detected and flagged with a 
PASSED.
However with my program cartridge, it detects a second memory extension 
as PASSED, but considers the first extension FAILED, I think that on the 
original CPU board, there was only 32kb, even 16kb and non-extensible on 
that board (hence the presence of two memory expansion boards in my 
original configuration). Briefly, according to the inserted program 
cartridge, the tests are sometimes but not always dynamically adaptable 
to a given configuration.


- A cartridge program can be programmed to operate only with a specific 
model of PERIPHERAL I/O board. The program cartridge form Ebay does not 
work with my PERIPHERAL I/O board, however my program cartridge seems to 
detect the PERIPHERAL I/O board of the US.


- I recreated the breakdown of my machine with the hardware of the other 
UTS. I'm talking about the situation of a long BEEP + blank screen, this 
happens if I remove the unique memory expansion board in the original 
configuration of the UTS from Ebay, this could mean that my own CPU 
board is maybe OK and that this could be one of two memory extensions of 
my machine that has a problem (Hypothesis).


But now, 

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

On 12/08/2017 19:35, Al Kossow via cctalk wrote:

the black one in the lower left

it is a tantalum, which are known to short, try removing them



Yes ! It was this one that was shorted !
I decided in the same way to repair the second disk drive but this time 
by taking some pictures.


http://www.zeltrax.com/classiccmp_forum/uts40/8406repair01.jpg

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair02.jpg

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair03.jpg

First, I desolder the same capacitor as on the first drive

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair04.jpg

but ...

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair05.jpg

Damn, it's not the same faulty capacitor ! Fortunately by checking the 
next one on the path of the + 24v


http://www.zeltrax.com/classiccmp_forum/uts40/8406repair06.jpg

There you are !

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair07.jpg 



Since, I reassembled the beast, everything seems to be OK on the 
subsystem side, he's ready to follow the instructions of the UTS 40 !


http://www.zeltrax.com/classiccmp_forum/uts40/8406repair08.jpg

http://www.zeltrax.com/classiccmp_forum/uts40/8406repair09.jpg 



When I think that somewhere on the CPU board of the computer it is 
perhaps a simple component like that capacitor which is the cause of 
this breakdown since 17 long years ...

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Al Kossow via cctalk]

On 8/12/17 6:45 AM, Dominique Carlier via cctalk wrote:

> I did not know this type of case for capacitors
> 
> http://www.zeltrax.com/classiccmp_forum/uts40/M2894-63_board03.jpg


the black one in the lower left

it is a tantalum, which are known to short, try removing them

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

It seems that you are right, I start successfully the Memorex 102 with 
the PSU without flickering, this HDD uses the same voltages as the 
Mitsubishi M2894-63D 8" disk drive (24VDC - 5VDC).


It remains me to find out which components are faulty

http://www.zeltrax.com/classiccmp_forum/uts40/M2894-63_board01.jpg

http://www.zeltrax.com/classiccmp_forum/uts40/M2894-63_board02.jpg

I did not know this type of case for capacitors

http://www.zeltrax.com/classiccmp_forum/uts40/M2894-63_board03.jpg



On 12/08/2017 14:30, Dominique Carlier via cctalk wrote:
Note that this produces the same result with either of the two disk 
drive connected however your suggestion is interesting.
These two drives have the same age and some capacitors may become 
faulty simultaneously. I will analyze the hypothesis, thanks!
I have an old hard drive (MEMOREX 102) with the same connector for the 
supply, i'm going to test the PSU with this other device.


On 11/08/2017 16:13, Al Kossow via cctalk wrote:

On 8/11/17 5:40 AM, Dominique Carlier via cctalk wrote:

I continued my investigations about the power supply if the 8406 subsystem.
Here is what I observe:

With the PSU in charge, I mean with a disk drive connected but without the +24V 
:
+5V -5V +12V -12V: OK
If I connect the +24V to the drive this is at this moment than the power supply 
goes mad and sends erratic alternating
voltages to the + & - 5V and + & - 12v.

It sounds like something is wrong with the disk drives, probably a shorted 
capacitor on the drive.

If you apply a resistive load on the 24v does the supply still go crazy?

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Note that this produces the same result with either of the two disk 
drive connected however your suggestion is interesting.
These two drives have the same age and some capacitors may become faulty 
simultaneously. I will analyze the hypothesis, thanks!
I have an old hard drive (MEMOREX 102) with the same connector for the 
supply, i'm going to test the PSU with this other device.



On 11/08/2017 16:13, Al Kossow via cctalk wrote:


On 8/11/17 5:40 AM, Dominique Carlier via cctalk wrote:

Hi all,

I continued my investigations about the power supply if the 8406 subsystem.
Here is what I observe:

With the PSU in charge, I mean with a disk drive connected but without the +24V 
:
+5V -5V +12V -12V: OK
If I connect the +24V to the drive this is at this moment than the power supply 
goes mad and sends erratic alternating
voltages to the + & - 5V and + & - 12v.

It sounds like something is wrong with the disk drives, probably a shorted 
capacitor on the drive.

If you apply a resistive load on the 24v does the supply still go crazy?

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Al Kossow via cctalk]

On 8/11/17 5:40 AM, Dominique Carlier via cctalk wrote:
> Hi all,
> 
> I continued my investigations about the power supply if the 8406 subsystem.
> Here is what I observe:
> 
> With the PSU in charge, I mean with a disk drive connected but without the 
> +24V :
> +5V -5V +12V -12V: OK
> If I connect the +24V to the drive this is at this moment than the power 
> supply goes mad and sends erratic alternating
> voltages to the + & - 5V and + & - 12v.

It sounds like something is wrong with the disk drives, probably a shorted 
capacitor on the drive.

If you apply a resistive load on the 24v does the supply still go crazy?

Re: The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi all,

I continued my investigations about the power supply if the 8406 subsystem.
Here is what I observe:

With the PSU in charge, I mean with a disk drive connected but without 
the +24V :

+5V -5V +12V -12V: OK
If I connect the +24V to the drive this is at this moment than the power 
supply goes mad and sends erratic alternating voltages to the + & - 5V 
and + & - 12v.


The question remains complete, which is the guilty component? :

http://www.zeltrax.com/classiccmp_forum/uts40/subsystem05.jpg

Is it possible that rectifier diodes become mad only when they have a 
significant load? Or is it finally a faulty capacitor that a ESR-meter 
can not detect as broken when is it not in charge?


Dominique

On 10/08/2017 20:07, Dominique Carlier via cctalk wrote:

Hi everybody,

After talking about this subject but in a thread following a sale - 
http://www.classiccmp.org/pipermail/cctalk/2017-July/036578.html - I 
decided to start here a thread but this time fully dedicated to the 
restoration of this rare computer, I named the SPERRY UNIVAC UTS 40 
and its Subsystem 8406 (2 X 8 "DSDD).

http://www.zeltrax.com/classiccmp_forum/1.jpg
http://www.zeltrax.com/classiccmp_forum/2.jpg
http://www.zeltrax.com/classiccmp_forum/3.jpg

I take the opportunity to show you the 4 motherboards of that beast in 
details (hi-res)

http://www.zeltrax.com/classiccmp_forum/uts40/4boards_presentation.jpg

The CPU board
http://www.zeltrax.com/classiccmp_forum/uts40/cpu_board.jpg

The communication board
http://www.zeltrax.com/classiccmp_forum/uts40/communication_board.jpg

A memory extension board
http://www.zeltrax.com/classiccmp_forum/uts40/memory_extension_board.jpg

And a - I don't know exactly - board
http://www.zeltrax.com/classiccmp_forum/uts40/unknown_board.jpg
(Any information ?)

In the other thread some topics were discussed, I will copy some part 
here in the idea of grouping these information.


This machine has an historical importance for me. I have an 
unforgettable memory of the data center in which my father worked. He 
sometimes took me with him in the early evening to start some 
procedures to be done during the night (process, tape backup, 
printing), my father worked on a UNIVAC 9200 II and then on a SPERRY 
UNIVAC 90/30. I remember the look of this big room in the dark, it was 
beautiful like a Christmas tree ;-) (that's what I was saying when I 
was five).


Me in 1980 at 6 ... yes, I had hair like the kid in the movie 
"Shining" ;-)

http://www.zeltrax.com/classiccmp_forum/uts40/dce1980.jpg

I even remember the exact configuration of the 90/30 : 3 X disk pack 
drive of 30 MB each, 3 X nine track tape drive Uniservo 10/14, a punch 
card reader/writer, a frightening and noisy drum printer, an 
indefinable number of UTS20D terminals, a beautiful Uniscope 100 that 
was standing next to the control panel on the central console, and a 
little bit later (1983) ... a UTS 40 and its subsystem. All these 
beautiful machines shone in the darkness of this data center during 
the night, it was beautiful, there was also the characteristic smell 
of hot machines in these places, well ventilated but smokers allowed. 
It is indeed of this time that I come to me an attraction for the 
technology and mainly for computers, preferably big, imposing and 
spectacular.


In 1987, my father acquired a VAX 8350 (3 X CPU - 6 X RA82H - 2 X TU81 
plus - a lot of VT220s and one VT340), progressively they started the 
migration from the 90/30  to the 8350, some part of the 90/30 are been 
progressively decommissioned and I received sometimes some gears. 
Notes that at the age of 14 I had only an Amiga 500 and a Commodore 
64, when I received the UTS 40 and its subsystem 8406 (with a UTS 20D 
bonus) in 1987 from my father's hands, I considered this computer as 
my first "serious" machine. Besides the sentimental value through the 
paternal donation, this machine evoked me the loved mainframes and 
computer terminals from my childhood.


I used this machine from 1988 to 1999 (the date of the breakdown). I 
wanted to give it a major utility in my own "data center" and under 
CP/M I coded in BASIC a program to manage a database, a kind of big 
help-memory-reminder, in which I noted all that passed by my mind, a 
lot of funny stuff, dreams, projects, technical stuff, music 
annotations, poetry... ;-)


Here I was 16 and so happy to have fun with my dear UTS 40 ^_^
http://www.zeltrax.com/classiccmp_forum/1989.jpg

My UTS 40 was ON every day and I used it constantly, the machine 
seemed indestructible however after 20 years of good and loyal 
services it began to show signs of fatigue. I had sometimes an error 
message during the POC TEST at initialization (RAM or ROM error, I can 
not remember). At this time I incriminated my brave cat who was 
watching me tapping on the keyboard and sleeped regularly on the top 
of the screen that served her as heater, thus blocking the normal 
ventilation of the machine.
Important thing : after a POC test 

The SPERRY UNIVAC UTS 40 system + 8406 double-sided diskette subsystem : Restoration

[Dominique Carlier via cctalk]

Hi everybody,

After talking about this subject but in a thread following a sale - 
http://www.classiccmp.org/pipermail/cctalk/2017-July/036578.html - I 
decided to start here a thread but this time fully dedicated to the 
restoration of this rare computer, I named the SPERRY UNIVAC UTS 40 and 
its Subsystem 8406 (2 X 8 "DSDD).

http://www.zeltrax.com/classiccmp_forum/1.jpg
http://www.zeltrax.com/classiccmp_forum/2.jpg
http://www.zeltrax.com/classiccmp_forum/3.jpg

I take the opportunity to show you the 4 motherboards of that beast in 
details (hi-res)

http://www.zeltrax.com/classiccmp_forum/uts40/4boards_presentation.jpg

The CPU board
http://www.zeltrax.com/classiccmp_forum/uts40/cpu_board.jpg

The communication board
http://www.zeltrax.com/classiccmp_forum/uts40/communication_board.jpg

A memory extension board
http://www.zeltrax.com/classiccmp_forum/uts40/memory_extension_board.jpg

And a - I don't know exactly - board
http://www.zeltrax.com/classiccmp_forum/uts40/unknown_board.jpg
(Any information ?)

In the other thread some topics were discussed, I will copy some part 
here in the idea of grouping these information.


This machine has an historical importance for me. I have an 
unforgettable memory of the data center in which my father worked. He 
sometimes took me with him in the early evening to start some procedures 
to be done during the night (process, tape backup, printing), my father 
worked on a UNIVAC 9200 II and then on a SPERRY UNIVAC 90/30. I remember 
the look of this big room in the dark, it was beautiful like a Christmas 
tree ;-) (that's what I was saying when I was five).


Me in 1980 at 6 ... yes, I had hair like the kid in the movie "Shining" ;-)
http://www.zeltrax.com/classiccmp_forum/uts40/dce1980.jpg

I even remember the exact configuration of the 90/30 : 3 X disk pack 
drive of 30 MB each, 3 X nine track tape drive Uniservo 10/14, a punch 
card reader/writer, a frightening and noisy drum printer, an indefinable 
number of UTS20D terminals, a beautiful Uniscope 100 that was standing 
next to the control panel on the central console, and a little bit later 
(1983) ... a UTS 40 and its subsystem. All these beautiful machines 
shone in the darkness of this data center during the night, it was 
beautiful, there was also the characteristic smell of hot machines in 
these places, well ventilated but smokers allowed. It is indeed of this 
time that I come to me an attraction for the technology and mainly for 
computers, preferably big, imposing and spectacular.


In 1987, my father acquired a VAX 8350 (3 X CPU - 6 X RA82H - 2 X TU81 
plus - a lot of VT220s and one VT340), progressively they started the 
migration from the 90/30  to the 8350, some part of the 90/30 are been 
progressively decommissioned and I received sometimes some gears. Notes 
that at the age of 14 I had only an Amiga 500 and a Commodore 64, when I 
received the UTS 40 and its subsystem 8406 (with a UTS 20D bonus) in 
1987 from my father's hands, I considered this computer as my first 
"serious" machine. Besides the sentimental value through the paternal 
donation, this machine evoked me the loved mainframes and computer 
terminals from my childhood.


I used this machine from 1988 to 1999 (the date of the breakdown). I 
wanted to give it a major utility in my own "data center" and under CP/M 
I coded in BASIC a program to manage a database, a kind of big 
help-memory-reminder, in which I noted all that passed by my mind, a lot 
of funny stuff, dreams, projects, technical stuff, music annotations, 
poetry... ;-)


Here I was 16 and so happy to have fun with my dear UTS 40 ^_^
http://www.zeltrax.com/classiccmp_forum/1989.jpg

My UTS 40 was ON every day and I used it constantly, the machine seemed 
indestructible however after 20 years of good and loyal services it 
began to show signs of fatigue. I had sometimes an error message during 
the POC TEST at initialization (RAM or ROM error, I can not remember). 
At this time I incriminated my brave cat who was watching me tapping on 
the keyboard and sleeped regularly on the top of the screen that served 
her as heater, thus blocking the normal ventilation of the machine.
Important thing : after a POC test error asimple reset was enough to 
restart the machine.This can be a useful data regarding components that 
were tired at that time. In the case of an eprom that breaks down, the 
change from the operating state to the non-operating state is direct, is 
it not? (i mean, without return possible to an operating state).


Anyway, one day I turned my UTS ON and instead of the POC TEST and the 
short BIP, just a long "BEEEP" and nothing on the screen. As I 
remembered the RAM / ROM issues displayed via the POC TEST, I suspected 
the ROMs on the "program cartridge". In fact I knew a lot less 
electronics compared to today. Fortunately I had printed all of my 
writings with the Manesmann-Tally dot matrix printer that was connected 
to the machine shortly before the failure.