Date   : Mon, 24 Mar 2008 12:18:09 -0000
From   : ken.lowe@... (Ken Lowe)
Subject: Reviving a dead BBC+

Hi all,

Following up on an email to the board early last year, I'm still trying to
revive an old BBC B+ 64K. I get a continuous beep with nothing shown on the
screen - not even the cursor. As part of the investigation, I managed to get
hold of a copy of the B+ Service Manual, and went through the recommended
checks (copied below), with the following results:

Step 1: Power Supply is OK
Step 2: Can't test - don't have a PET tester!
Step 3: IC78 Pin 39 (HS) held low, Pin 40 (VS) held high
Step 4: IC42 Pin 40 (not RS) pulses low when break key pressed
Step 5: Activity detected on IC42 Pin 7 (SYNC) and Pin 34 (R/notW)
Step 6: Phi1 and Phi2 are both cycling. Not able to check phase shift - only
got 1 scope probe just now
Step 6b: IC33 Pin 3 is low; IC41 Pin 8 is low
Step 7: Address lines are showing signs of activity
Step 8: Clock divider is working fine
Step 9: CRTC notCS (IC78 pin 25) does NOT pulse low after BREAK. This signal
stays high.

Based on the above, my first thought was the 6845 CRTC, which I swapped out
with a known working one from a BBC B 32K. Unfortunately this made no
difference. I then tried reseating the OS / Basic ROM as I was told that
this could also be the cause of the problem. Again, this made no difference.
I'm confident that I have the correct OS / Basic ROM in the machine (socket
71) because I've been able to read the contents of the ROM on another system
and it looks correct.

I don't really want to just keep swapping chips out without any
justification, because invariably I have to destroy the chip to remove it
from the board (they're soldered directly onto the board and I struggle to
remove the larger ones without damaging them). Any recommendations on what
to go after next? I thought about the system VIA, but I'm not sure if that
would prevent the cursor appearing on screen? What about the Ferranti
Vidproc ULA? I assume this wouldn't prevent the Beeb from booting, and would
only affect the display?

Thanks
Ken.


Extract from B+ Service Manual:

Follow the sequence of checks shown below.

1 If there is no noise on power-up, and no LEDs light up then check the
power supply (see 9.2).

2 If a PET tester is available then use it. The PET tester will work on the
B+ providing the CPU is running and it can access the ROM, although it may
give strange screen output, and some of the tests will fail. Please refer to
the information manual supplied to dealers for details of the operation of
the PET tester when used with the B+. If PET will not work at all then
either the CPU isn't running or it cannot access the ROM.

3 Check HS and VS signals (pin 39 and pin 40 IC78) using an oscilloscope. HS
and VS should be clean TTL voltage levels, HS pulsing every 64?s and VS
pulsing every 20ms.

Results: if they are stuck or floating then carry on with the checks be1ow.
If they are working then the CPU must have programmed the 6845 and so must
have gained access to the OS ROM. Check using PET (see information manual
supplied to dealers). If the signals are there but are not pulsing at the
correct intervals then look for a data line fault to CRTC.

4 Check that the notRS pin of the CPU (pin 40 IC42) is high when the
computer is switched on. It should pulse low on power-up and when the BREAK
key is pressed. If it is stuck low then look for shorts or damaged
components around the 555.

5 Check that there is activity on the SYNC (pin 7 IC42) and the R/notW (pin
34 IC42) lines of the CPU. If SYNC is stuck then the CPU has stalled, and
R/notW won't be working anyway. Check for address and data bus short or open
circuit, or a complete failure to select the OS ROM (see 9.5).

6 Check the CPU clocks. Phi1 and phi2 should be as shown in figure 1 (see
section 5.2). If not then check the 2M circuitry from the video processor
IC53. IC33 pin 3 should be low. If not then check SYNC 1M at pin 8 IC41
which should also be low. If SYNC 1M is high then check IC25. If SYNC 1M is
stuck high then find which one of the inputs to IC41 is stuck low. The 1MHz
device attached to this input must be checked.

7 Check activity on the CPU address lines. If after a BREAK the activity
starts and then stops, this suggests that the CPU cannot read the OS ROM.
Check the OS ROM by replacing it with a known good one. Check that it is
enabled and that all address lines are present. Check following BREAK that
notOE pulses low at 2MHz (inverse of phi2), and notCS goes low and stays low
for a time.

8 Check all clocks from video processor 8M 4M 2M 1M, see 9.3.

9 After BREAK check CRTC notCS pin pulses low (pin 25 IC78).