exxos's Atari, Amiga & retro forum

So, despite a lack of tools (I do now have a logic probe) I've been trying to figure out what is wrong with the BBC B I have with just that, a multimeter and my brain. Interestingly, the logic probe can detect frequency, though the limit is about 1Mhz.

Previously I had removed all the chips from their sockets and cleaned the pins, followed by a squirt of contact cleaner, after which I started getting some proper symptoms: Caps lock on, blank screen and continuous tone from the speaker.

Those particular symptoms appeared after cleaning IC6, the video ULA.

I decided to have a closer look at the mobo and cleaned off a lot of old flux, same with the KBD, and then finally got round to soldering in the replacement LEDs for caps lock etc. I found that the mobo had what looked like a short on one of the front ports, so that was fixed.

Next on the agenda was going through with the LP to see whats what. I verified all the LS chips have 5v and GND, though without my TL866 I can't test any. I also tried to test for clock signals, and did manage to confirm some with the LP, but the others were a bit of a bust.

I followed a suggestion on startdot.org.uk to check the voltage with a multimeter, and that for pins 3, 37 & 39 on the CPU read about 2v-2.5v under normal operations, but didn't get much of a result there. Well, certainly not what was suggested.

I also found a minimum configuration on stardot.org.uk and did try that, however it made no difference. Also removing chips one by one as suggested in a troubleshooting guide made no difference other than everything fell silent without IC6, the video processor, but then it does provide a lot of clock signals for the board.

Also, the CPU is supposed to go high on pin 7 when the CPU fetches and decodes an instruction, and right now that pin shows nothing at all, bar a really brief flicker on power on indicating low, which could simply be the LP powering on. Double checking with the multimeter shows a reading of 0v, so I don't think the CPU is doing anything,

The other strange thing is the KBD. The schematics show a 5v feed to the KBD, but measuring the voltage at the 74LS chips on it gives me just over 2V and the LP says the GND pin is also high. It's really weird, though disconnecting the KBD makes no difference to the machines behaviour other than the caps lock light can't come on.

Though I've been told the voltages I'm getting from the PSU (4.8v at the connector, 4.5v-4.6v at the ICs), that seems kind of low for a machine with the IC count it has, so I've got a full PSU recap kit on the way. Don't know if it'll help, but worth a try at about £12. I've also got a new 6502 on the way because that has badly oxidised legs as well and I can't help but think a lot of this is connection problems for a lot of the chips/

My main concern is the old sockets though. The wipes are badly oxidised, and the OS ROM is looser than I'd like in a socket. I'd remove and solder in directly if I knew the ROM was good, but I don't.

I'll keep plugging away.

Ok, so a bit more Beebing today on account of it being a bit grey and miserable, plus I thought it would take my mind off BTs humongous screw up on Friday!

I figured after all the flitting about looking at clock signals as best I could, plus after getting some advice over at stardot.org, that I should do some back to basics investigation and go back to make sure all the voltages are good.

Looking generally at TTL ICs, VCC can be as low as 4.75v for normal operations. The BBC Model B Issue 4 I have has 4 connections for power, VCC 1-3 and a -5v connection which I believe is required for the sound chip amongst other things.

Checking these I got the following readings:

VCC 1 - off-load 4.94v on-load 4.83v
VCC 2 - off-load 4.93v on-load 4.76v
VCC 3 - off-load 4.93v on-load 4.76v
-5v - off-load 4.21v on-load -5v

From this we can see that VCC2 & VCC3 are sailing close to the wind with 4.76V at the board.

Further checking some of the main ICs, I got these readings:

6502 - VCC=4.72v
OS ROM - VCC=4.74V
Basic ROM - VCC=4.75V
IC3 6522 - VCC=4.68V
IC69 6522 - VCC=4.67V
All RAM ICs - VCC was between 4.55V and 4.68V
IC2 MC6845 - VCC=4.66v
IC6 - VCC=4.7v

I also did a random check of the 74LS chips on board and got readings from all of below 4.75v.

So power is looking like a strong candidate for the problems I'm seeing. I've ordered a recap kit for the whole PSU, not just the 2 RIFA smoke generators and starter cap, which I'll hopefully have sometime this week.

Fingers crossed they'll perk up the PSU, next on the agenda is a spot of looking at a HiRes photo of the underside of the mainboard to look for any obvious problems, though my initial look shows a real clean board free of rework.

PSU re-cap done and voltages are slightly better across the board, with a minimum of 4.77v across the board and a maximum of 4.8v. So over the 4.75v minimum, though still a little low.

The other thing I noticed thanks to @cmorley was that of the 3 voltage tuning resistors, R25, R26 & R30, R26 was missing leaving behind only it's snipped legs!!

So I've reached out to Mark at Retroclinc, from where the cap kit came, to ask a bit more about these resistors and the values they usually have by default to see if I can tune the PSU to deliver something a bit closer to 5V, just in case the voltages are still a bit low.

I think in the meantime I'll also have a scout for some images to see if I can find values from there and compare to my own PSU.

Mark is super helpful. He deserves his reputation for being the #1 8-bit Acorn guru.

Interestingly I've found quite a few PSU pictures and there's quite a few missing R26.

Yes they were clipped at the factory. I have some PSUs clipped and others not.

cmorley wrote: 05 May 2021 19:56 Yes they were clipped at the factory. I have some PSUs clipped and others not.

Fair enough.

I cheated with my Master and put a MeanWell psu in instead but obviously that's not the cheapest option.

Does work well though.

PhilC wrote: 05 May 2021 22:04 I cheated with my Master and put a MeanWell psu in instead but obviously that's not the cheapest option.

Does work well though.

The thought had crossed my mind, believe me! a PD-2505 looks like it would fit the bill size and voltage wise.

My PSU is now fully recapped and seems to be in spec now, well just, but good enough.

I've also replaced IC14 as suggested by a friend, but still no dice. Constant tone, caps lock light and no video out.

I now have a super basic scope, and whilst it can't read the frequencies of the clock, I can pick up wave forms and see that when checking pins 4-7 of IC6 the values double with each step, so 5 is double that of 4, 6 is double that of 4 etc, which I'm taking as a good sign that IC6 is doing something right.

I'm less certain about the clock, as trying to get a handle on IC43, I'm not 100% sure that the outputs there and on IC40 are correct. Damn my proper scope being in storage!!!

Also, when trying to read pins 1-6 on IC43, touching them with a scope probe and logic probe clearly makes something happen. The tone from the speaker starts to warble, which is new, and has me wondering if there is a clock problem, so I have replacements for IC40 & IC43 on the way.

One final question I have is around RAM. There is a mix of different chips in the machine. You have a mix of the standard Hitachi 4816 chips, and Mostek 4516 chips.

I have the datasheet for the Mostek chips and it seems like a reasonable replacement, but I cannot find the matching sheet for the 4816 chips, so was hoping somebody might be able to point me in the right direction as I'd like to compare the chips to see if they are a good replacement or if I need to get some more original 4816's.

Thanks in advance.