BBC B+ Repair log

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

BBC B+ Repair log

Post by mw0uzo » Tue Jan 17, 2017 11:06 am

This is a repair log for a broken BBC B+ with the dreaded Beeeeeee of death.

The machine is in very poor condition outside:
SV204767.JPG
Inside not so bad, some corrosion throughout.
SV204765.JPG
It was stuffed with ROMs, I took them all out apart from the MOS/BASIC rom and cleaned contacts, but still no joy.

The first problem found was the RESET signal being held low. This was a problem with the 555 reset timer circuit. The 555 timer IC, two 1Mohm resistors and 3 100nF capacitors were replaced. This restored the reset signal, but still Beeeee....
SV204771.JPG
A capture of /RS, CLK, SYNC, RnW, D7-D4 showed that the CPU was attempting to run, but the instructions were not valid, after one or two instructions the CPU did the same thing repeatedly and then hung totally.

I further cleaned the MOS ROM and socket, still Beeeeeee

Taking the ROM out and reading it in the EPROM programmer, showed it was knackered, the data was gone! I found the DFS ROM that was in it, that was shot too, and I tested a few other ROMs that read fine.

So, I've got to program a new EPROM. I have a the 27C256 from the Masters fiasco and a 27128 old Pendown ROM for the DFS. I don't have a EPROM eraser, it's on order. So in the meantime, I've got flourescent lighting in the workshop so I thought I'd try placing them in the diffuser right next to the tube to see if they will erase. Will it work? Ha, we'll see - any UV getting through the phosphor might not be high energy enough to erase. Maybe a week in there will erase...
SV204775.JPG

User avatar
1024MAK
Posts: 7548
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: BBC B+ Repair log

Post by 1024MAK » Tue Jan 17, 2017 11:55 am

mw0uzo wrote: Taking the ROM out and reading it in the EPROM programmer, showed it was knackered, the data was gone! I found the DFS ROM that was in it, that was shot too, and I tested a few other ROMs that read fine.
So is the MOS/BASIC "ROM" actually a EPROM, or a non-windowed EPROM/PROM?

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

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Tue Jan 17, 2017 11:58 am

1024MAK wrote:
mw0uzo wrote: Taking the ROM out and reading it in the EPROM programmer, showed it was knackered, the data was gone! I found the DFS ROM that was in it, that was shot too, and I tested a few other ROMs that read fine.
So is the MOS/BASIC "ROM" actually a EPROM, or a non-windowed EPROM/PROM?

Mark
It was a non windowed PROM or ROM. 23256-1109 (C) ACORN
The data within was stripes and blocks of 04 06 14 and 16

User avatar
BigEd
Posts: 1890
Joined: Sun Jan 24, 2010 10:24 am
Location: West
Contact:

Re: BBC B+ Repair log

Post by BigEd » Tue Jan 17, 2017 3:34 pm

If you have a week, using daylight might do it.

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Tue Jan 17, 2017 3:52 pm

BigEd wrote:If you have a week, using daylight might do it.
Yeah I wondered this, what would be better. It's pretty grey and cold out there. It probably won't work under the fluorescent tube either :lol:

Just checked them, no obvious bit changes. I'll just have to wait for the eraser to arrive or get some new EPROMs.

I've got hold of the B+ MOS ROM and BASIC 2 from an emulator, so I assume if I join these two files together, MOS ROM first, that should be the right configuration for the B+ ROM...

User avatar
CMcDougall
Posts: 6088
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: BBC B+ Repair log

Post by CMcDougall » Tue Jan 17, 2017 4:50 pm

Daylight won't wipe them, may be in 10yr LOL

It's Basic2 first 8000-BFFF, then second MOS C000 to FFFF.

I have my worker imaged, & should be same as emulator files, will check later....
ImageImageImage

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Tue Jan 17, 2017 5:04 pm

CMcDougall wrote:Daylight won't wipe them, may be in 10yr LOL

It's Basic2 first 8000-BFFF, then second MOS C000 to FFFF.

I have my worker imaged, & should be same as emulator files, will check later....
Ah ok, great you've saved me some cursing!
So in the image it should be
BASIC 2.0 0000-3FFF
MOS 2.0 4000-7FFF
Looking at the Beeb memory map I can see exactly why that is. :)

User avatar
CMcDougall
Posts: 6088
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: BBC B+ Repair log

Post by CMcDougall » Tue Jan 17, 2017 5:10 pm

Yes!

also don't blow original DFS2.0j , use the upgrade with is DFS2.26, then all games work & won't get dumb disc errors.
ImageImageImage

User avatar
CMcDougall
Posts: 6088
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: BBC B+ Repair log

Post by CMcDougall » Tue Jan 17, 2017 8:18 pm

me wrote:real MOS should be same as emulator files, will check later....
os 64 comp.jpg
diffs
just checked, and are same, BUT :shock: with bEm22 or BeebEm414
this area is different, so like elkOS1 & BeebOS1.2, this may be Fred /Jim 'ram' area to show system state,
where the "(c) thank bob pete etc" gets wiped :o

here is mine a B+64k to compare to yours:
Attachments
bplus64.jpg
B+64k
ImageImageImage

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Wed Jan 18, 2017 12:16 am

Ok, I'll put in DFS 2.26. Thinking about what other ROMs that could go in. MMC DFS and the MMC COP114 tools rom. ADFS. That could go in with the DFS. I might as well 'compact' all the ROMs into 32K devices as they've all got to be recreated. Are there any other essential ROMs you use? I have yet to identify which are the most useful yet.

Thanks for checking the ROM you have. I have the 64b os rom from b-em with the message gone. I can recreate the message with the hex editor from OS 1.2

That board is idential to mine, interestingly, that trimpot is cranked too, I assumed mine had been tweaked and backed it off quite a bit. What is it for? Audio? ADC reference voltage? can't be that, its too far away

When I get this working, i'll need to somehow sort the case. Anyone got an OK condition top case half that they don't mind selling? Anything is better than that mess :lol:
Maybe even just the stick on plastic bit. Probably not available? They must have been available at some point...

What about case yellowing? I've seen the hair bleach creme/UV method. Is there something easy like dunking it in bleachy water or scrubbing it with thickened bleach? I've got some Oxy type stain remover powder. Scrub with that? Like an abrasive paste? The Masters are a bit yellowed too, I could follow up on them if sorting this one works. The paste idea sounds good to me (with eye and hand protection :lol: ) then I can be really careful not to get the paste on any of the black plastic parts.

User avatar
vanpeebles
Posts: 438
Joined: Wed Nov 28, 2012 10:01 am
Location: UK
Contact:

Re: BBC B+ Repair log

Post by vanpeebles » Wed Jan 18, 2017 10:36 am

Are there many B+ owners on the board? I don't see that much posted about them :)

User avatar
flaxcottage
Posts: 3085
Joined: Thu Dec 13, 2012 8:46 pm
Location: Derbyshire
Contact:

Re: BBC B+ Repair log

Post by flaxcottage » Wed Jan 18, 2017 11:14 am

mw0uzo wrote:Ok, I'll put in DFS 2.26. Thinking about what other ROMs that could go in. MMC DFS and the MMC COP114 tools rom. ADFS. That could go in with the DFS. I might as well 'compact' all the ROMs into 32K devices as they've all got to be recreated. Are there any other essential ROMs you use? I have yet to identify which are the most useful yet.
I obtained a B+ 64K a while back. The first upgrade I performed was to add an extra 96K using the information from this website - http://www.firths.org/bbc/my_b+128k.html This gave me 6 sideways RAM banks

I then added a TurboMMC card storage system.

My ROM complement is as follows;

1. OS and BASIC
2. MMC DFS and DFS 2.26 in a single 32K PROM
3. Basic Editor and Exmon 2 in a single 32K PROM
4. Computer Concepts Mega 3 ROM
5. Computer Concepts Interbase ROM
6. 96K sideways RAM upgrade

The machine boots to the TurboMMC FS. I have set up *DIN volume 0 to boot on Shift-Break to show a menu that allows access to what I need. One of the options is to load various ROMs into the sideways RAM. So I can load the MMC copy ROM (this MUST be in sideways RAM to work as it uses SWR as its workspace) when it is needed. Other ROMs I use are ADT2, ADI2, ADFS 1.33, various languages, eg. Logotron LOGO, Wordwise Plus, GXR B+ and the Datacentre RAMFS.

This system is a beast and is really by favourite BBC micro variant. The only thing missing is a network upgrade but the MMC system and Econet conflict over workspace.
- John

Why do I keep collecting Acorn gear? I'm going to need a considerably bigger man-cave. :?

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Wed Jan 18, 2017 1:47 pm

Thanks for the extra information flaxcottage! Shame that COP114 won't work in ROM, but that memory upgrade should fix that. That's definitely going in there once it's fixed.

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Wed Jan 18, 2017 6:42 pm

It's looking a bit better after a clean.
Mixed some oxy-action and toilet bleach together with some warm water and dabbed it on the case every now and then over the afternoon.
Cleaned they keyboard keys and removed the yukky broken plastic faceplate and cleaned off the glue.
Still quite yellow though, the difference between the hidden plastic and sunlight exposed plastic is remarkable.
Before
SV204766.JPG
SV204782.JPG
After .. ha it's just cleaner. Or is it slightly less yellow than the top picture ... taken under the same light conditions? Middle image was taken outside. Opinions? Maybe tomorrow if I can find a container big enough, I'll try strong, straight Oxy Action and leave it dunked in.
SV204784.JPG

User avatar
CMcDougall
Posts: 6088
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: BBC B+ Repair log

Post by CMcDougall » Wed Jan 18, 2017 7:46 pm

You also need an original Ruler /key strip with 64k on right, rare as hens teeth, I have one! 8)

Don't bother with stupid TurboLOL MMC copier rom, the 33p China MMC with MMFS1.21+ rom copies back & forth with a basic prog :lol:

also ADFS1.3 will just shove PAGE to high so nothing works :roll:
ImageImageImage

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Wed Jan 18, 2017 10:17 pm

CMcDougall wrote:You also need an original Ruler /key strip with 64k on right, rare as hens teeth, I have one! 8)

Don't bother with stupid TurboLOL MMC copier rom, the 33p China MMC with MMFS1.21+ rom copies back & forth with a basic prog :lol:

also ADFS1.3 will just shove PAGE to high so nothing works :roll:
Ha, yeah the one on this is a standard one, at least its not snapped.
TurboLOL :lol: I've got some of the cheap SD card reader boards, the plan is to make some up for these Beebs. I do want the DataCentre though for the Master development setup...

If ADFS is *UNPLUGged page goes back down?

User avatar
CMcDougall
Posts: 6088
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: BBC B+ Repair log

Post by CMcDougall » Wed Jan 18, 2017 10:23 pm

^it sure does, but still better in nearest skip... :lol:
If both DFS & ADFS, goes upto &1F00 :x

The MMC China card works perfect on my B+64 8)
ImageImageImage

User avatar
flaxcottage
Posts: 3085
Joined: Thu Dec 13, 2012 8:46 pm
Location: Derbyshire
Contact:

Re: BBC B+ Repair log

Post by flaxcottage » Wed Jan 18, 2017 10:38 pm

One has to realise that Col has a negative thing for ADFS. :lol: :lol: :lol:
- John

Why do I keep collecting Acorn gear? I'm going to need a considerably bigger man-cave. :?

BBCB+64K
Posts: 221
Joined: Mon Mar 22, 2010 9:51 pm
Location: North Norfolk
Contact:

Re: BBC B+ Repair log

Post by BBCB+64K » Thu Jan 19, 2017 11:35 am

CMcDougall wrote:^it sure does, but still better in nearest skip... :lol:
If both DFS & ADFS, goes upto &1F00 :x

The MMC China card works perfect on my B+64 8)
Ha! My B+128K has PAGE at &2200 :shock: Due to it having ADFS, DOSFS and an extra copy of the SRAM Utils.

Steve

duikkie
Posts: 2867
Joined: Fri Feb 07, 2014 3:28 pm
Contact:

Re: BBC B+ Repair log

Post by duikkie » Thu Jan 19, 2017 11:43 am

someone yells: SMART-SPI :shock: , it handles SWR well if you read how , not everything works only with pb0/pb1 :)

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Fri Jan 20, 2017 11:03 am

EPROM eraser arrived!

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Fri Jan 20, 2017 12:30 pm

One new system ROM
SV204798.JPG
BuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuBEEP! \:D/
SV204802.JPG
System ROM image for repairs
BpSYSTEMROM.zip
(24.41 KiB) Downloaded 14 times

User avatar
vanpeebles
Posts: 438
Joined: Wed Nov 28, 2012 10:01 am
Location: UK
Contact:

Re: BBC B+ Repair log

Post by vanpeebles » Fri Jan 20, 2017 12:42 pm

Top work! =D>

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Fri Jan 20, 2017 2:29 pm

The story isn't over yet. Plugged in the disk drives and the system VIA went down. Or something more annoying - no video. Dreaded Beeeeeeeeeeeeeee returns. Either these old VIAs are really flaky, or my disk drive unit/cable until is killing things.

User avatar
danielj
Posts: 6262
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: BBC B+ Repair log

Post by danielj » Fri Jan 20, 2017 3:15 pm

Those old 6522s are pretty flakey :D

d.

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Fri Jan 20, 2017 3:43 pm

danielj wrote:Those old 6522s are pretty flakey :D

d.
Ha, another 40 pin IC to desolder... yay :lol:

Had a quick poke around, sometimes init starts then hangs other times it just executes nothing then hangs.
Verified the system ROM - OK.
Perhaps a chip select logic support chip has died.
Doesn't get as far as R/W to the VIA, so could not be that.

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Fri Jan 20, 2017 5:59 pm

Initial suspect is the CPU.
- Hangs with no SYNC a lot when Phi1 Phi2 ok and /IRQ, /NMI and /RES high. When it runs for a longer time than usual, the 1mhz IO clock stretching happens.
- Executes different amount of ROM each time, sometimes none, sometimes more.
- OS ROM chip select not stuck.
- DBE, RDY tied to 5V so no CPU waits
- A3-A0 works initially, then when CPU hangs all address lines go high.

Thoughts? The CPU should not stall completely?

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Thu Feb 09, 2017 3:55 pm

It's groundhog day in mw0uzo's lab LOLolollolol
Replaced soldered in CPU, still broken.
Replaced 2 x soldered in VIAs, still broken :x
Looks like it's going to be a swine, out with the logic analyser etc again...

mw0uzo
Posts: 202
Joined: Wed Dec 21, 2016 9:00 pm
Contact:

Re: BBC B+ Repair log

Post by mw0uzo » Mon Feb 13, 2017 2:21 pm

- Main 16Mhz clock OK
- 8,4,2,1 Mhz outputs from VidProc OK

- /1E clock OK.

No display output at all, no cursor etc. So VidProc not being initialised.

- ROM ok in programmer, burned a new one just in case. No change.
- Changed CPU and both VIAs, no change.

Captured D7-D0
D2 seems to like being high, but not stuck. Hint of a problem? (Reconnecting probes and capturing again and the problem seems to have gone.)

Captured RD0-RD7
Seems ok, nothing stuck, a few extra short, high pulses on RD2. Ram probably ok.

So if RAM ok and ROM ok and no init, then perhaps its the buffer/latch IC49 74LS245N connecting RAM/VidProc/RD bus to main CPU bus.
Desoldered IC49 and IC12, socketed, swapped, problem remained.

Captured data bus, RnW, SYNC, RES, CS for ROM etc
Sometimes, it does not execute ROM! The pattern for SYNC is different.
Other times it does and manages to do two 1Mhz bus writes before it all goes wrong.

Burned and fitted tricky's test ROM. Copied all contents of 0000-3FFF to 4000-4FFF, as OS ROM resides in the upper half. Tricky - can you confirm what I have done here is OK? Your B+ seemed to work with the ROM from 0000-3FFF...
Tried it, no display, no sound, no flashing keyboard LEDs, nowt.

So something pretty bad must be wrong. Chip select decode logic? /CS for the ROM appears to work OK...

User avatar
1024MAK
Posts: 7548
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: BBC B+ Repair log

Post by 1024MAK » Mon Feb 13, 2017 2:57 pm

I forget the details about Tricky's test ROM. But keep in mind, the address decoding logic on the board decides where in the CPU memory map the ROM appears. So in the EPROM (assuming a 27128 / 27C128) chip is used, the code should start at ROM address 0000 and go all the way to the end (so that the reset vector etc is in the correct place).

If using a different EPROM, where the code is programmed depends on the logic levels on the higher EPROM address pin(s), so if a 27256 / 27C256 is used, A14). Ideally these should be low, then the code is burned from ROM address 0000 etc. But if the higher address pins are high, then the code should be burned in the corresponding EPROM address range. Or just burn the code at every possible memory range.

Also if you are using a larger EPROM, you do need to check that any higher EPROM address pin(s) are supplied with a fixed stable logic level.

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

Post Reply