Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

discuss both original and modern hardware for the bbc micro/electron
Post Reply
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

Hello all - I've just discovered this after managing to pick up a BBC Model B Issue 7 which I am trying to revive.

So far, I have replaced the dodgy X caps in the power supply and the power supply is now measuring a healthy 5.13v on the 5v rail and -4.89 on the -5 rail.

When I turn the BBC on, I get the long continuous beeeeeeeeeep and the screen just flickers.

I have checked the clocks being output by the IC6 and they all seem fine.

On the 6502, Reset is high and brought low when I press the break key on they keyboard. Clock in on pin 37 is at 2MHz and nice and square, but clock out on pin 3 has a significant slope on the rising edge, but does measure at 2MHZ.

Pin 37
pin37.png
Pin 3
pin3.png
IRQ and NMI are both high. No activity on SYNC, R/W is high.

All of the address pins seem to be high, although a little bit strange is that they are at 3.8v. vcc is at 5.1v.

I have pulled out both the ROMs and been able to read them in my minipro. They seem to be fine.

A long shot, as it seems to not be able to read the ROM, but I have tried moving S25 North and removing S25 completely.

I have also tried pulling as many ICs as I can that are socketed. IC 73, IC 7, IC 69 (mine doesn't have Econet/Disk Drive/TTS). IC 4 is not socketed so I haven't tried pulling that yet. None of the remaining ICs are running noticeably hot.

Does anyone have suggestions for what I could try next?
wiggy
Posts: 97
Joined: Fri Feb 12, 2021 12:19 pm
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by wiggy »

No activity on SYNC is bad. That means the 6502 isn't even trying to execute instructions.

You probably need to get !Reset on the second channel of your 'scope and trigger on that so that you can see what's happening just after !reset has been released: hopefully you'll be able to see it trying to access the ROM for a few cycles and be able to work out what's going on from there.

The address lines may not be being driven (hence just floating), depending on where you are looking at them. In any case, the high level from TTL is rather a long way off the Vcc rail, so this might not be an issue. (Yes the CPU is NMOS, but NMOS is TTL-implemented-using-FETs, and doesn't have the same high-side drive strength as CMOS). However, the trouble pulling up Phi2 out does suggest a high-resistance path in the Vcc supply, so it'd be worth removing the CPU and checking its pins and those in the socket are clean and making good contact.
Ronin47
Posts: 133
Joined: Tue Apr 20, 2021 5:32 pm
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by Ronin47 »

The Beeb I fixed recently was giving a permanent beep, and although it had a bad CPU, it also had a bad keyboard that was causing it to do the same.

If you have a spare known good keyboard and CPU it would be worth swapping over and seeing if it makes a difference.
User avatar
1024MAK
Posts: 10927
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by 1024MAK »

Check both the +5V/VCC and the 0V/GND/VSS pins on the 6502 are actually electrically connected. So for example, measure the voltage across the top of pins 8 and 1, and between 8 and 21 with the power on. AND with the power OFF, measure using the 200Ω range on your multimeter from these pins to the power lead connectors on the board. Poor power or ground pin connections can cause all kinds of strange symptoms.

If you are trying to go for a minimal configuration, the keyboard does not have to be connected. The BASIC ROM is not needed. But you won’t get into BASIC and the sound system will give a continuous tone like you have now. But you should get a message on the screen.

If you have a breadboard, and you suspect the 6502 but don’t have a spare, you could build a NOP generator on the breadboard.

Dave H. on here can sell you a working NMOS 6502 at a very good price :D.

Mark
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

Thanks for the replies!

Resistance measured between +5v lead and pin 8 measures around 0.4ohms
Resistance measured between ground and pin 1 and 21 is also around 0.4ohms.

When powered off

Voltage between 8 and 1/21 both at 4.98

I set the scope to trigger on Reset (CH1) and then measured Sync on ch
20210621_132357.png
So there is some sync activity, but it stops very quickly (35us)
20210621_132756.png
I thought I might see if I could see A0 twinkling away, it should at least be reading the NMI vector from the ROM and hey presto it is doing something:
20210621_134222.png
and likewise - A1:
20210621_134820.png
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

Bonus! I hooked the keyboard back up and watched the interaction between RSET going high again after I press break and the length of the SYNC pulses. It varies significantly. I suspect the suggestion to pull out the 6502 and build a NOP tester is a very good idea.

Can anyone recommend a 1MHz oscillator I can use for the job? I'm a bit of a newbie at this!
wiggy
Posts: 97
Joined: Fri Feb 12, 2021 12:19 pm
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by wiggy »

Hmmm... your pattern on SYNC doesn't match mine....

(Ch1 = Sync, Ch2 (bottom) = !Reset, Ch3 (pink) = OS ROM !CS)

... admittedly, this screenshot is from the German version of OS1.20, but so far as I can tell from the disassembly on MDFS.net the instructions after reset are the same for both. Of course, if you don't have OS 1.2 then some more checking might be in order. The very first SYNC pulse appears to be part of the 6502 resetting - the second pulse looks right for the first actual instruction (at the address fetched from the reset vector).

So that suggests the 6502 isn't actually executing from ROM, but is executing garbage until it locks up.

Have a look at your OS ROM's !CS pin (again, just after reset) and see if it's getting selected: if not, dig into the address decoding logic; if so, check to see if the data pattern(s) coming out of the ROM match those arriving at the 6502. If not suspect the data buffers/bus muxes...
Attachments
SYNC after reset
SYNC after reset
User avatar
1024MAK
Posts: 10927
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by 1024MAK »

It’s always worth checking the ROM socket contacts. Both visually (under a good light and maybe with a magnifying glass) to see if any socket contacts are bent out of shape. And also test with a multimeter on resistance.

IC51 - (OS ROM socket):
Pin 1 - A15 from 6502 unless the board has been modified, in which case it will be connected to +5V/VCC.
Pin 20 - /CS from link S21 (check this link).
Pin 27 - A14 from 6502 unless the board has been modified, in which case it will be connected to +5V/VCC.

IC52, IC88, IC100 and IC101 (sideways ROM sockets):
Pin 1 - open circuit/not connected on board.
Pin 20 - /CS from IC20 (74LS139).
Pin 26 - depends on link S32 or link S33 but normally A13.
Pin 27 - connected to pin 28 (+5V/VCC).

All other remaining pins are all wired in parallel between all five sockets.

Mark
User avatar
1024MAK
Posts: 10927
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by 1024MAK »

beezly wrote:
Mon Jun 21, 2021 2:02 pm
Can anyone recommend a 1MHz oscillator I can use for the job? I'm a bit of a newbie at this!
You don’t have to run a NMOS 6502 at 1MHz. The three most important things are:
  • The clock must be run at between 100kHz and 2MHz.
  • The higher period and the low period don’t have to be exactly the same, but they should be similar or equal.
  • The clock must have fast rise and fall times.
A NE555 can’t do the fast rise/fall times, but 74 series logic chips (LS, ALS, HC, HCT) can.
A microcontroller can also be used to clock a 6502.
Or you can use a higher frequency crystal and then use a 74 series counter to divide it down to a lower frequency.

Mark
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

I managed to put together a NOP generator - I ended up using the 2MHz clock off the BBC as it seems absolutely fine.

This managed to get me into a more consistent state, but alas this CPU looks poorly.

Sync was bouncing up and down in a regular way
sync.png
but when I checked the Address lines, they do not form regular patterns as would be expected.
a0-a1.png
marcelaj1
Posts: 311
Joined: Wed Apr 29, 2020 5:07 pm
Location: Surrey
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by marcelaj1 »

I had not heard of a NOP generator so did a search, there appear to be a number of ways to do it. The one I thought may be helpful in this circumstance would be this one that replaces the roms.....
http://www.6502.org/mini-projects/nop-gen/nop-gen.htm
This forces NOP, no need for a clock or anything, it won't work if there is a fault on the board but if the 6502 "runs" then you have checked the clock and data buses. It looked cheap to make, an added bonus :D

Ashley.
wiggy
Posts: 97
Joined: Fri Feb 12, 2021 12:19 pm
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by wiggy »

That doesn't look like it's executing NOPs: NOP (0xEA) should be a 2-cycle instruction, so SYNC should be equal lengths high (1 cycle) and low (also 1 cycle).

Double check (on the CPU's pins) that the data bus is right and there are no bad contacts - if it looks like the right answer, then yes that looks like the CPU is a dud.
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

Many thanks for your post wiggy - you made me go back and check my homework and sure enough I had not built the NOP generator correctly.

I have now fixed it and I see evenly spaced sync pattern and a nice cascade of 1s and 0s through the address lines.
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

I have been experimenting a bit more with an Arduino Mega 2560 and this 6502.

I wrote some code which would get the Arduino to output a nice clock signal, configurable between 2MHz and 125khz-ish and I thought as I have so many spare pins on the mega, I would try and connect everything up and see if I could monitor the address lines and feed different opcodes to the CPU on the data lines.

With all of the (relevant) pins connected up to the Arduino, the CPU would exhibit similar behaviour to how it is in the Beeb... SYNC for a few cycles (generally more than in the bbc) and then it would lock up.

Interestingly, if I removed all the data/address connections from the arduino and hardcode the data pins to 0xEA - (so effectively the arduino is just providing clock, NMI, RSET, IRQ, RDY and monitoring SYNC) the 6502 starts behaving itself again. Is it worth me acquiring another 6502, and if so, how do I summons Dave H (I don't appear to be able to send him a private message as I'm a newbie).
User avatar
1024MAK
Posts: 10927
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by 1024MAK »

The ability to reply to, or to send Private Messages (PMs) is disabled for new members at first. However you can ask for full PM functionality to be enabled by using the ‘Contact us’ link (envelope icon) at the bottom of every page. This sends a message to the admin’s who will consider your request. They are more likely to enable full PM functionality if you demonstrate that you are a friendly member. So please do consider telling us about yourself and your hobbies in the introduce yourself section.

You may get Dave’s attention by asking in the wanted section. Or I could point him in your direction.

You can currently receive PMs, so it’s possible for members to send their email address to you by PM.

Mark
User avatar
daveejhitchins
Posts: 6597
Joined: Wed Jun 13, 2012 6:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by daveejhitchins »

PM sent . . . Dave H.
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

With thanks to Dave H and all those who helped advise - I now have a working BFG Compqpar!
PXL_20210626_153045309.jpg
This looks like suspect RAM to me. If I switch the S25 jumper south, the machine won't boot. If I remove the S25 jumper - it boots OK (but with 16K of RAM obviously).

By comparing the ASCII between "BFG Compqpar" and "BBC Computer" it looks like D2 is the problem, (every byte that is wrong is 0x04 out). Given that, and the fact the problem seems to be in CAS1 (because it fails to boot with S25 south), does it make sense to replace IC55?
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

I swapped out IC55 with a new one and hey presto... it's alive and no longer a BFG Compqpar!

Many thanks to those who helped revive this Beeb. I'm very happy!

Obligatory Elite photo for proof :D
PXL_20210630_151615303.jpg
User avatar
IanS
Posts: 1829
Joined: Mon Aug 31, 2009 7:02 pm
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by IanS »

beezly wrote:
Wed Jun 30, 2021 4:21 pm
Obligatory Elite photo for proof :D
The colours look wrong, looks like you have Green & Blue swapped.
beezly
Posts: 10
Joined: Wed Jun 16, 2021 7:12 pm
Location: Huddersfield, UK
Contact:

Re: Issue 7 BBC Micro won't boot - beeeeeeeeeeeeeep

Post by beezly »

Interesting! I have a SCART cable hooked up to an HDMI box - and sure enough - the blue and green pins were hooked up backwards. Thanks for pointing it out... I was wondering why some of the colours schemes were a bit garish!
Post Reply

Return to “8-bit acorn hardware”