2 Faulty Beebs

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

2 Faulty Beebs

Postby JannievanZyl » Fri Jan 12, 2018 8:09 pm

So, for my sins to repair (with the group's help!) 2 Model B's recently, a friend dropped off his own two faulty Mobel B's! :D

The one is completely dead and the other does the constant beep thing.

Had a quick look at the dead one and the display chip was not generating any of the lower frequencies. Plugging the display chip from the constant beep one in, gives the same constant beep. So now I have two beebs with the same symptom.

Still reading all the threads before I'm going to dive in, but will soon be bugging everyone for more advice! :D

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sun Jan 21, 2018 7:26 pm

With this system I get the long, continuous beeeeep.

The first obvious issue with the 1st one is that their is no clock on pin-37 of the CPU.

What also worried me was that the voltage of the 8MHz signal from the video chip was much lower (~3.3V) than the 1, 2 and 4MHz signals which are ~4V. But it seems to drive the flip-flops attached to it quite fine. I replaced 2 of the 7474's (IC30 and 34) and this seemed to drive the level of the 8MHz a bit higher. It does look like the level is good enough.

But I still get no clock on the CPU. I'm trying to trace backwards but it get quite hairy quite quickly. :shock:

There is an 8-input NAND gate ((IC23) that seems to enable (or is involved with) the clock circuitry to the CPU but it's inputs are all static.

So it seems some other magic must happen before the CPU is supply with a clock?

Any hints / suggestions on what I can look at to get the clock to the CPU?

I burned a copy of tricky's ROM but it also gives the long beep. Not 100% sure the ROM works obviously.

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sun Jan 28, 2018 2:12 pm

tricky,

Some question on your diag ROM,

Burned a ROM and first tested in a working beeb.

In the working Beeb:

1. I get the motor control LED come on, the relay clicking and the various patterns on the VDU.

2. But the Caps-lock and Shif-locks don't come on.

3. I'm also not sure anymore if I heard the sounds.

As I said this was in a Model B that tests fine with the OS ROM plugged in.

Any idea?

User avatar
tricky
Posts: 2098
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: 2 Faulty Beebs

Postby tricky » Sun Jan 28, 2018 4:42 pm

That would suggest that the system via isn't working, was it ok with the OS ROM back in?

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sun Jan 28, 2018 8:11 pm

With the normal OS ROM, it works fine. System runs, can input a BASIC program, etc.

On the faulty system;

The faulty unit (with the continuous sound) is driving me nuts. :?

I tested its system VIA in another machine and it works fine. I took out IC 24 and it failed on the logic tester so put a new one in, same problem.

What I can see is that there is no interrupt on the 6502 and it also looks like some address bus lines are static, while others have activity on them.

Could IC 32 be a factor? I don’t have a replacement at this time.

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Mon Jan 29, 2018 5:19 am

Ok, an update.

Just to recap (if you'll pardon the pun), a friend brought me his two faulty Model B units.

- 1 PSU is dead and I had to re-do the recapping on the other (C9 was inserted wrong way around).

- 1 system completely dead
- 1 system gives the continuous beep

On both, I traced and measured, swapping chips 1 at a time between the two systems, figuring the chances are small that both the same chips from both systems are faulty. Used a working 6522 as the system VIA as all the symptoms point to it.

Eventually, I took all the major chips and plugged them into a working beeb. These all tested faulty: :shock:

- 1 x 6502 - completely dead
- 1 x 6502 - actually boots but then hangs with garbage on the screen
- 3 x 6522 VIAs dead (only 1 of the 4 is working)
- 1 x CRT Controller
- 1 x Video ULA
- 1 x System ROM

So, I suspect a huge power issue on the one system that blew the board and took most of the chips with it.

Is it common for so many chips to go? At one point, I was doubting myself but these definitely test faulty/

I'll need to dig around for a few chips to try and get at least 1 system working. At this point, I need at least a 6502 and a 6522. Not sure what else is faulty, I definitely found some faulty glue that I also replaced, so I suspect there might be another few.

But it pains me greatly not to fix both of them. Anyone on this group to who I can ship one or two boards for repair?

beeb-4.jpg

User avatar
tricky
Posts: 2098
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: 2 Faulty Beebs

Postby tricky » Mon Jan 29, 2018 8:17 am


duikkie
Posts: 2782
Joined: Fri Feb 07, 2014 3:28 pm
Contact:

Re: 2 Faulty Beebs

Postby duikkie » Mon Jan 29, 2018 8:30 am

You need only a 6502 for one to work the second 6522 can put on later as userport

i think is board where use to swap ics

JannievanZyl wrote:Ok, an update.

Just to recap (if you'll pardon the pun), a friend brought me his two faulty Model B units.

- 1 PSU is dead and I had to re-do the recapping on the other (C9 was inserted wrong way around).

- 1 system completely dead
- 1 system gives the continuous beep

On both, I traced and measured, swapping chips 1 at a time between the two systems, figuring the chances are small that both the same chips from both systems are faulty. Used a working 6522 as the system VIA as all the symptoms point to it.

Eventually, I took all the major chips and plugged them into a working beeb. These all tested faulty: :shock:

- 1 x 6502 - completely dead
- 1 x 6502 - actually boots but then hangs with garbage on the screen
- 3 x 6522 VIAs dead (only 1 of the 4 is working)
- 1 x CRT Controller
- 1 x Video ULA
- 1 x System ROM

So, I suspect a huge power issue on the one system that blew the board and took most of the chips with it.

Is it common for so many chips to go? At one point, I was doubting myself but these definitely test faulty/

I'll need to dig around for a few chips to try and get at least 1 system working. At this point, I need at least a 6502 and a 6522. Not sure what else is faulty, I definitely found some faulty glue that I also replaced, so I suspect there might be another few.

But it pains me greatly not to fix both of them. Anyone on this group to who I can ship one or two boards for repair?

beeb-4.jpg

User avatar
sbadger
Posts: 242
Joined: Mon Mar 25, 2013 1:12 pm
Location: Farnham, Surrey
Contact:

Re: 2 Faulty Beebs

Postby sbadger » Tue Jan 30, 2018 4:58 pm

JannievanZyl wrote:Eventually, I took all the major chips and plugged them into a working beeb. These all tested faulty: :shock:

- 1 x 6502 - completely dead
- 1 x 6502 - actually boots but then hangs with garbage on the screen
- 3 x 6522 VIAs dead (only 1 of the 4 is working)
- 1 x CRT Controller
- 1 x Video ULA
- 1 x System ROM


That's an awful lot to have died.... is there any chance the testing itself is at fault?
A3020| A3000x3| BBCBx3 | Electrn | Masterx3 |RiscPC| RPix3
A600 | C64 bbin x2|C64C | Toastrack |QL | XB360&1X |GB |GBC |GBA |GBASP | DS | 3DS XL x2| MD | MS
Atari 7600 | PS1-2-3-4| PSP |Vita |SNES |GC |N64 |Wii & U |Switch |JammaCab |Sony PVMx2

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Tue Jan 30, 2018 8:50 pm

sbadger wrote:
That's an awful lot to have died.... is there any chance the testing itself is at fault?

My first thoughts exactly.

If everything tests faulty, check the meter (or tester!) type of thing.

But I took the chips and plugged them int o working beeb, one at a time. If they failed, I would plug the working chip back in, test and then repeat the test with the suspect IC. So every chip was (at least) double checked and some - like the flaky CPU - I tested a few times as the symptoms were pretty weird.

Just for sanity, I'll use a second workig beeb to confirm.

It seems like 1 board blew most of it's chips and it is highly suspicious that the one PSU is also dead. Something likely happened to the PSU and gave a huge over-voltage or something like that.

User avatar
1024MAK
Posts: 7011
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: 2 Faulty Beebs

Postby 1024MAK » Tue Jan 30, 2018 11:06 pm

A severe overvoltage (greater than about 6.5V but various between different chips), or a reverse voltage (below -0.7V approx) is more than capable of killing multiple chips.

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

User avatar
DutchAcorn
Posts: 1671
Joined: Fri Mar 21, 2014 9:56 am
Location: Maarn, Netherlands
Contact:

Re: 2 Faulty Beebs

Postby DutchAcorn » Wed Jan 31, 2018 6:18 am

I fixed a TRS-80 that had suffered from a high voltage spike. It’s a lot of work, tracing the signals through the chips one by one using a logic analyser (did not need a scope).

You say you only checked the major chips. It’s fair to assume other chips have also died.

Most chips you can still get, some are harder...
Paul

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Wed Jan 31, 2018 12:33 pm

DutchAcorn wrote:I fixed a TRS-80 that had suffered from a high voltage spike. It’s a lot of work, tracing the signals through the chips one by one using a logic analyser (did not need a scope).

You say you only checked the major chips. It’s fair to assume other chips have also died.

Most chips you can still get, some are harder...


Exactly, I still expect to find a whole bunch more.

Wonder if the 6502 from an old Commodore floppy drive will work in here?

I have a huge moral problem to not fix these machines. But it's time consuming and I also want to get to my 3 A3020s I started with and have another 80-odd machines staring at me to be restored. :)

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Wed Jan 31, 2018 12:35 pm

1024MAK wrote:A severe overvoltage (greater than about 6.5V but various between different chips), or a reverse voltage (below -0.7V approx) is more than capable of killing multiple chips.

Mark


If C9 on the PSU was reversed during capping (I saw it was and fixed it before I turned it on), would it have had such an effect? Or would the PSU just not start up?

User avatar
1024MAK
Posts: 7011
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: 2 Faulty Beebs

Postby 1024MAK » Wed Jan 31, 2018 2:10 pm

JannievanZyl wrote:If C9 on the PSU was reversed during capping (I saw it was and fixed it before I turned it on), would it have had such an effect? Or would the PSU just not start up?

I think it unlikely that an incorrectly fitted C9 would have caused the damage. The PSU may have eventually started (as a reverse fitted electrolytic capacitor with a low voltage would have had a small amount of capacitance). But the PSU would not have been able to operate at the normal output current, so the output voltages would have been lower than they should have been, and may not have been stable.
C9 supplies the current that switches the main switching transistor (Q2) on and keeps it on until the end of the switching cycle. If C9 is too low in value, the transistor will either not switch on, or will not be able to stay on for long enough to transfer enough power to the transformer.

I have drawn a partial schematic, it's in this post.

Apart from C1, C2 and C9, does it look like any other parts in this PSU have been worked on or replaced?

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Wed Jan 31, 2018 6:20 pm

1024MAK wrote:Apart from C1, C2 and C9, does it look like any other parts in this PSU have been worked on or replaced?
Mark


I'll need to check. I did not really pay attention to the PSUs until now that I've realised how many chips are popped.

I'll go and look in more detail now.

User avatar
hoglet
Posts: 6844
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: 2 Faulty Beebs

Postby hoglet » Wed Jan 31, 2018 7:38 pm

Maybe it was a victim of a lightning strike?

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Thu Feb 01, 2018 6:47 am

hoglet wrote:Maybe it was a victim of a lightning strike?


Would for explain it, but these are Cape Town machines and if lightning strikes here the whole population is under the bed it's so rare! :lol:

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sat Feb 03, 2018 9:31 pm

Definitely need some help here :)

I've got a known good CPU, System VIA (IC 3) and keyboard in the system.

Still get the continuous tone. Even with tricky's DIAG ROM.

Also had no LEDs on the keyboard coming on, so replaced the '259 (IC32) and now the Caps Lock LED comes on. But still the same continuous tone.

Here's the bit that I think will make sense to someone who knows the circuit;

When I turn the system on, there is clock on the CPU for a brief period and then it stops. If I keep Break in, clock appears continuously on the CPU.

So while RST and not-RST is in their initial states, I get clock. But when I release Break it stops.

It looks like not-RST (what is the syntax to type a 'not' signal? :) ), forces the output of one of the D-type flipflops IC34-pin6) to stay low, thus allowing the clock to reach the CPU. But once it goes high the clock is blocked. The signal that controls this comes from the area around or before IC23 and IC24.

Any pointers that can cause this?

Is there any documentation that describe what must happen straight after reset?

User avatar
danielj
Posts: 5600
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: 2 Faulty Beebs

Postby danielj » Sat Feb 03, 2018 9:53 pm

Not an expert, but I'd follow the clock from the vidproc (IC6) and see where it stops..?

Usual notation for not is /, so not-RST is /RST :)

d.

User avatar
1024MAK
Posts: 7011
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: 2 Faulty Beebs

Postby 1024MAK » Sat Feb 03, 2018 10:10 pm

The CPU clock in the Beeb depends on the current address from the CPU. For selected address ranges (the 1MHz I/O device addresses) the clock "slows" to a 1MHz clock.

Are the various clock frequency outputs from the videoproc / video ULA okay (don't worry if the 8MHz looks like a sine or triangle wave).

The service manual does explain the system design, but not in any great detail. Still worthwhile reading through it.

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

crj
Posts: 522
Joined: Thu May 02, 2013 4:58 pm
Contact:

Re: 2 Faulty Beebs

Postby crj » Sat Feb 03, 2018 11:03 pm

danielj wrote:Usual notation for not is /, so not-RST is /RST :)

"Usual?"

I've never seen that one.

nRST, #RST, ~RST, ¬RST and RST with a line over the top, but not /RST. (-8

User avatar
hoglet
Posts: 6844
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: 2 Faulty Beebs

Postby hoglet » Sat Feb 03, 2018 11:08 pm

If you have ever used PALASM, you would have come across the use of / for negation:
https://en.wikipedia.org/wiki/PALASM

Dave

User avatar
1024MAK
Posts: 7011
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: 2 Faulty Beebs

Postby 1024MAK » Sat Feb 03, 2018 11:28 pm

Well, we prefer a bar over the function or signal name. But I keep forgetting to ask grovel to Daniel to get him to provide the forum with it...

Maybe [o]text overline[/o]? Them active low signals won’t enable themselves.

So, on systems with no bar, the next nearest ASCII symbol got used. Alas, everyone and their dog decided on their own standard....! :roll: #-o

I use constructs like /CS and /OE simply because when I was younger, that was what I found in the technical documentation that I was reading. The / symbol being far more common in what I was reading than any of the alternatives.

Mark
For a "Complete BBC Games Archive" visit www.bbcmicro.co.uk NOW!
BeebWiki‬ - for answers to many questions...

dp11
Posts: 725
Joined: Sun Aug 12, 2012 8:47 pm
Contact:

Re: 2 Faulty Beebs

Postby dp11 » Sun Feb 04, 2018 12:34 am

At least one cad system uses /R/S/T which renders as RST with a bar across the top.
I've never come across ¬RST but have used RSTB in the past.
Many early CAD systems didn't allow non text as net names. If they did when you came to export information and read it into another system it usually ended up in a mess. EDIF netlist tried to solve this by making the rules very strict, hardly anyone in the early days could generate valid EDIF.

User avatar
daveejhitchins
Posts: 3850
Joined: Wed Jun 13, 2012 5:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: 2 Faulty Beebs

Postby daveejhitchins » Sun Feb 04, 2018 7:49 am

dp11 wrote:At least one cad system uses /R/S/T which renders as RST with a bar across the top.
Yes, Altium uses that method, with an option to just use one instance of '/' at the beginning of the required overscored text.

CUPL uses '!' to indicate a negated label . . .

Dave H :D
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sun Feb 04, 2018 9:07 am

1024MAK wrote:The CPU clock in the Beeb depends on the current address from the CPU. For selected address ranges (the 1MHz I/O device addresses) the clock "slows" to a 1MHz clock.

Are the various clock frequency outputs from the videoproc / video ULA okay (don't worry if the 8MHz looks like a sine or triangle wave).

The service manual does explain the system design, but not in any great detail. Still worthwhile reading through it.

Mark

Yup, the ULA gives out all the correct frequencies and it does get through to the CPU but *only* while RST is active (i.e. you hold down the Break key, or at initial startup).

Read the service manual but where I'm stuck is what exactly happens during the 'reset cycle'. Clearly the CPU tries to do something (or the system does) which fails and it then all hangs with the clock to the CPU also stopping.

If you look at the attached portion of the circuit, it's effectively the output of the 8-input NAND gate (IC23) that stops the clock from getting to the CPU. Through the inverter (IC33) it keeps the one input of the OR gate (IC29) high, so effectively the clock stays high.

So something higher up must be happening to ensure this signal allows the clock to get through. Something is not completing during reset properly.

I checked the system ROM and it works in another machine. Does the same with tricky's diag ROM as well.

I know I'm frustratingly close (but not there) to fixing this one. :)

Capture.JPG

User avatar
vanekp
Posts: 465
Joined: Thu Nov 30, 2000 7:09 am
Location: The Netherlands
Contact:

Re: 2 Faulty Beebs

Postby vanekp » Sun Feb 04, 2018 10:10 am

JannievanZyl wrote:Ok, an update.


Eventually, I took all the major chips and plugged them into a working beeb. These all tested faulty: :shock:

- 1 x 6502 - completely dead
- 1 x 6502 - actually boots but then hangs with garbage on the screen
- 3 x 6522 VIAs dead (only 1 of the 4 is working)
- 1 x CRT Controller
- 1 x Video ULA
- 1 x System ROM

So, I suspect a huge power issue on the one system that blew the board and took most of the chips with it.

Sound a bit like someone reversed the polarity of the power on the mother board easy to do with the wires that come from the PSU.

User avatar
vanekp
Posts: 465
Joined: Thu Nov 30, 2000 7:09 am
Location: The Netherlands
Contact:

Re: 2 Faulty Beebs

Postby vanekp » Sun Feb 04, 2018 10:20 am

JannievanZyl wrote:So something higher up must be happening to ensure this signal allows the clock to get through. Something is not completing during reset properly.

I checked the system ROM and it works in another machine. Does the same with tricky's diag ROM as well.

I know I'm frustratingly close (but not there) to fixing this one. :)

Capture.JPG

Have you checked that the RSTA (not) is working (left middle on your diagram) comes off R20/C10 and also the two RST and RST(not) from IC33 pin 3 & 4 generated off the NE555.

JannievanZyl
Posts: 145
Joined: Sat Feb 11, 2017 8:56 pm
Location: Cape Town, South Africa
Contact:

Re: 2 Faulty Beebs

Postby JannievanZyl » Sun Feb 04, 2018 11:12 am

vanekp wrote:
JannievanZyl wrote:So something higher up must be happening to ensure this signal allows the clock to get through. Something is not completing during reset properly.

I checked the system ROM and it works in another machine. Does the same with tricky's diag ROM as well.

I know I'm frustratingly close (but not there) to fixing this one. :)

Capture.JPG

Have you checked that the RSTA (not) is working (left middle on your diagram) comes off R20/C10 and also the two RST and RST(not) from IC33 pin 3 & 4 generated off the NE555.

Yup, All working fine.


Who is online

Users browsing this forum: No registered users and 15 guests