BBC B+ fault

[Sniffer]

Maybe I should start a new thread on this, if so just say.

The reason I asked initially about the B+ memory mapping was that having just sorted my B, due the help received here, I powered up a B+ and got a long beeeeep. Another machine out of the loft.

Pressing CTRL BREAK gets rid of the long beeeeeep but I'm left with random exclamation characters over the display and can't change modes. Well, I can change mode but then there is no keyboard response.

I ran Trickys tester rom but can't make head nor tail of the output. The initial screen appears with lots of exclamation markks at the bottom and it continues on doing it's thing but the audience is non the wiser.

Apart from changing each ram chip, is there a better way? I've changed two already without success. The first I suspected due to voltages from IC49 and the second due to the wave form of pin 14 (of the ram) being vastly different from than all the others.

I'm starting to enjoy this tinkering again.

[Kazzie]

I'm sure we can get some sense out of it for you.

Firstly (with all the B+-related posts recently I've got them all mixed up in my head) are we talking about a 64k or a 128k model?

Random exclamation marks are a good indicator of where the fault may be. Working from a table of ASCII values (which is almost identical to MODE 7 teletext), a space is encoded as 0x20, and an exclamation mark is 0x21: it seems bit D0 of some RAM addresses has been switched on by mistake.

So if it's a RAM fault, and a 64k machine, I'd suspect IC55, which provides RD0 for all 64k addresses. However, other chips that connect directly to the RD0 line include the data bus transceiver, IC49 (74ls245), the latch for MODE 7 teletext, IC54 (74ls273), a chip on the expansion card (for a 128k model), and possibly even the video processor, IC53.

If you see screen corruption (possibly of a different nature) in other MODEs, then that would suggest that it isn't IC54. Narrowing it down among the other possibilities might not be as straight forward, though.

Edit: A video of the Test ROM footage could let us help interpret what it's trying to say.

[Sniffer]

Thank you for the comments Kazzie.

Firstly (with all the B+-related posts recently I've got them all mixed up in my head) are we talking about a 64k or a 128k model?

64K

If you see screen corruption (possibly of a different nature) in other MODEs, then that would suggest that it isn't IC54. Narrowing it down among the other possibilities might not be as straight forward, though.

Much screen corruption and long beeeeep on initial startup, CTRL BREAK gets rid of long beeeeb and all corruption other than random exclamation marks. Any attempt to change mode, whilst it seems to work, results in an unresponsive keyboard and complete screen corruption.

Edit: A video of the Test ROM footage could let us help interpret what it's trying to say.

I did do a video but it is 250MB in size. Also, the variant flickering of the monitor versus the fone camera is not good. I could upload to a website if you think it will help.

[DutchAcorn]

Actually, just a photo of the first (mode 7) screen may be sufficient. Does the sound and the caps lock/shift lock toggling work with Tricky’s rom installed?

[Kazzie]

Edit: A video of the Test ROM footage could let us help interpret what it's trying to say.

I did do a video but it is 250MB in size. Also, the variant flickering of the monitor versus the fone camera is not good. I could upload to a website if you think it will help.

Ah the curse of the rolling shutter...
If you don't think the video's legible enough to be of use, don't worry about it. A sequence of still photos (in order) might be more practical.

A quick photo of the corrupted screen when you change mode might also be useful.

[Sniffer]

Ok.
On power on with long beeeeeeeep;

After CTRL-BREAK without long beeeeep;

Typing MODE 6;

In MODE6;

and MODE0;

After a memory check, which I assume from previous info means nothing;

Test ROM pic to follow. Hope this helps.

[stephen_usher]

It does look like bit 1 is being stuck high in some locations.

They're not random, some of those exclamation marks stay in the same place all the time, which suggest that those memory locations are hard stuck whilst others are marginal.

It definitely looks like it's the IC holding bit 0 rather than anything else.

[Sniffer]

Here's a gif of the start of the test rom but I doubt it's of any help. It is too big to upload here so I've made it available at;

https://www.ictprovision.com:80/pics/TEST2.gif

[Sniffer]

Actually, just a photo of the first (mode 7) screen may be sufficient. Does the sound and the caps lock/shift lock toggling work with Tricky’s rom installed?

Sorry, missed this post. Yes, the sound works at the start and the caps lock/shift lock toggle.

[DutchAcorn]

I agree with Kazzie that this is most probably caused by a faulty IC55. Sometimes the first bit (bit 0) is high when it should be low and vice versa.

The way it behaves it could well work if you piggyback a working memory chip on it (but you can never really tell).

[Sniffer]

I agree with Kazzie that this is most probably caused by a faulty IC55. Sometimes the first bit (bit 0) is high when it should be low and vice versa.

The way it behaves it could well work if you piggyback a working memory chip on it (but you can never really tell).

Piggy backing didn't have any effect but obviously that's not conclusive. This afternoon I'll set about removing the ram, starting with IC55. Hopefully I'll get a result. I've already done IC65, as checking voltages on IC49 gave me a high reading on the pin leading there and I've done IC67 as it's pin 14 gave a wildly different waveform than all the other ram chips.

[Sniffer]

Success.

I socketed IC55 and replacing the 4564 did the trick. Cheers guys.

[tricky]

I really need to do some more work on that ROM.
The best way to use it at the moment is to load it into an emulator and run it side by side and compare the differences.
I wrote it with the model B in mind, so it doesn't set up ACCRON.