Silent, nothing visible

discuss both original and modern hardware for the bbc micro/electron
Coeus
Posts: 1764
Joined: Mon Jul 25, 2016 12:05 pm
Contact:

Re: Silent, nothing visible

Post by Coeus » Tue Aug 04, 2020 10:13 pm

kfro wrote:
Tue Aug 04, 2020 9:10 pm
See the resistor R102 and the capacitor C42. When the clock goes high, the cap will be charged up via the resistor. The time it takes the cap to charge is related to the capacitance times the resistance. The rule of thumb for the time it takes for the cap to charge to the supply is about 3RC. But the gate will switch long before then (at 2 volts apparently). For short we just use the term RC. I do not think this is particularly critical, I think it's just there to remove noise (signal that we do not want) from the clock. The cap charge time will have the effect of delaying the signal on pin 13 and the NAND gate will not switch low until both pin 12 and 13 are high. When the clock goes low pin 11 will switch straight away, So I guess if you compare pin 12 to pin 11 you should be able to see this difference. If you need any more excuse to play with the new scope that is :-)
So thinking about a clean input signal, I was thinking that this arrangement would shorten the duty signal, though I could not think why that would be required. For noise, I wonder if there is any reason why +ve noise would be more likely than -ve

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Tue Aug 04, 2020 10:22 pm

oh hang on the video ULA is IC6 not IC7

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Tue Aug 04, 2020 10:32 pm

So thinking about a clean input signal, I was thinking that this arrangement would shorten the duty signal, though I could not think why that would be required. For noise, I wonder if there is any reason why +ve noise would be more likely than -ve
clocks like this suffer from problems like ringing where the signal overshoots and then you get a damped oscillation until it settles down. This is probably what they were trying to remove. I suspect they were only concerned with the leading edge of the clock so didn't bother to do anything with the pos to neg transition. That would have needed another gate, which would have added to the cost and BCP real estate. "If it an't broke, don't fix it"!

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Silent, nothing visible

Post by Kazzie » Wed Aug 05, 2020 3:02 pm

Also, flip-flops - as used the Video ULA used for dividing the 16MHz down to 8, 4, 2, 1MHz - only pay attention to one of the clock edges: rising or falling, but not both. (That's the whole point of it: by ignoring half the transistions, the clock rate is halved!) On that basis, why bother cleaning up the unused* edge? :wink:

*The 16MHz clock is also used around the FDC, if I recall, but I'm not sure how exactly it gets used.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Wed Aug 05, 2020 5:13 pm

Yep, just found it, it's fed into the clock of IC86 pin 1 if S27 is in the north position. Also I suspect that we need to recheck pins 4, 5, 6 and 7 on IC6 for the divided clocks. We should see 1, 2, 4 and 8 meg clocks respectively.

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Wed Aug 05, 2020 5:46 pm

Hang on, there may be some user error here...

I have IC7 with the Ferranti ULA in.

But now you are talking about IC6 which has an Acorn chip in.

Which is it before I short out the wrong pins here?

Asking for a friend.
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Wed Aug 05, 2020 5:54 pm

Ok...

I checked the pins on IC6 and initially had no sound and no signal.

Then my probe slipped and shorted between Maybe pin 3 and 4 (I can't be 100% sure)

Then the boooooop sound started and I had correct signal on those pins on IC6
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Wed Aug 05, 2020 6:27 pm

Ok, IC7 is another ULA for the serial / Cassette interface. Make sure that the divided clocks on IC6 are there consistently. You could try flexing the board slightly and power cycling a few time. Once you are happy that they are good we should move on to seeing if the CPU (IC1) is running.

Pin 8 - VCC, +5V
Pins 1 and 21 GND,
Pin 37 - 2MHz clock in, this can drop to 1MHz by the way.
Pin 40 - not reset, should be high (+5V)
Pin 5 and 4 not NMI and not IRQ will mostly be high
Pin 34 R/not W, you should see activity
Pins 9 to 20 and 22 to 25 - the address bus, you should see activity
Pins 26 to 33 - the data bus, you should see activity

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Wed Aug 05, 2020 7:15 pm

Ok. Measurements from IC1

Pin8 - +5v
Pins1 and 21 - GND
Pin37 - 2mhz
Pin40 - +4.4v (which seems a bit odd, I again confirmed all the vcc were +5v and they are)
Pin5 and 4 - +2v and +5v respectively. (It seemed pin5 was a bit wavery in the 100ths of a volt range)
Pin34 - There was very slight activity
Pins9-20 and 22-25 - about 1.3mv @ 16mhz
Pins26-33 - Good activity - decent waveforms

So it looks like that is doing good.

I wonder if the 6502 is getting out of reset (having read other threads)?
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Wed Aug 05, 2020 8:49 pm

sorry pin 5 should have been pin 6, pin 5 is not connected :shock:

The reset seems a little bit odd, it's not out of spec as such, but I'd really expect to see +5V. Can you check that it goes low when you press the BREAK key on the keyboard.

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Wed Aug 05, 2020 9:14 pm

It doesn't.

No change on break or when I short across the reset switch pads.
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Wed Aug 05, 2020 10:58 pm

ok, I'm fairly sure that pressing break etc should cause not reset on the CPU to go low momentarily. So what do you get on IC33 pin 3 if you press break?

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Silent, nothing visible

Post by Kazzie » Thu Aug 06, 2020 6:14 am

kfro wrote:
Wed Aug 05, 2020 10:58 pm
ok, I'm fairly sure that pressing break etc should cause not reset on the CPU to go low momentarily.
You're right. In fact, it should remain low while the break/reset button is being held.

IC16 (a 555 timer) is used to generate a reset pulse on power on or press of the appropriate button. Is there any response from Break on pin 3?

If so, check pins 3 and 4 of IC33 (74ls04), which inverts the signal ready for the CPU.
Colday wrote:
Wed Aug 05, 2020 5:54 pm
Ok...

I checked the pins on IC6 and initially had no sound and no signal.

Then my probe slipped and shorted between Maybe pin 3 and 4 (I can't be 100% sure)

Then the boooooop sound started and I had correct signal on those pins on IC6
This is interesting, as it indicates that the clock circuitry had stalled, but a brief accidental short got things running again. I wonder if the 16MHz clock has been running fine all this time, but the Video ULA in IC6 has been failing to divide down? If it fails to create the 8MHz clock, it won't be able to use that to generate the 4MHz clock, and so on.

This isn't a type of failure I've come across before, but the intermittent boop sound indicates that the 4MHz clock wasn't present, and after your hand slipped it was.

If you happen to have the lack-of-boop again, it may be interesting to probe the clock signals starting from 1MHz (pin 4) and working your way up to 16MHz (pin 8 ), to see at which point you find a clock signal.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 10:37 am

Agreed!

I'm thinking that the clock issue maybe due to poor connections, either between scope probe and crusty pin or the normal dry joint, broken trace, IC socket contact, etc.. At some point all the socketed ICs need to be pulled and pins and sockets cleaned and some suitable contact clearer / protector applied.

I suggest the we park the clock issue for now, but keep it in mind and concentrate on the reset issue, since this seems to be a hard fault and I do not think the two are interrelated. Yes, just refreshed my memory on how the 555 works, as long as trigger is low the output will be high. So, once the reset goes through the inverter IC33 pins 3 and 4 not-reset will be held low.

If IC33 pin 4 is correct, i.e. high while the break key is pressed then I think this points most strongly to IC33, but it could be something else in the circuit pulling it down but that probably wouldn't prevent not-reset from going low. (BTW do we have a convention for saying "not reset" or "bar reset"?)

I did a bit of research and the other things that are connected to not-reset are:
  • External VIA IC59
  • IC34 pin 4
  • 1MHZ Bus connector
  • TUBE connector
  • IC89 ECONET advance serial controller
I might have missed something of cause :|
Last edited by kfro on Thu Aug 06, 2020 11:04 am, edited 1 time in total.

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Silent, nothing visible

Post by Kazzie » Thu Aug 06, 2020 10:51 am

From previous experience, and given that most of the chips in the Beeb are TTL or NMOS (which can't pull pins up to 5V that well) I'd expect the fault to be with IC16 and it's passive components, IC33, or a broken track.

ICs can fail in all sorts of weird and wonderful ways, but I'd probably investigate those possibilities first.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 12:02 pm

Thanks for your help you two!

I shall have a look round those mentioned.

This board does have Econet, and all IC's were pulled (and cleaned) prior to the board going in the dishwasher.

Don't forget that, in everything you suggest (well, the fault still has to be somewhere but the cause may be unusual).

I am happy that we are making progress though, I suspect the reset is the issue as there just doesn't seem to be any reaction anywhere. Whether that being held low and with no output, that would create the squiggly lines on the screen is anyone's guess.

Also, don't forget this was put into the dishwasher as it had been sitting in a puddle of rusty water and the back of the board is still contaminated with rust, although it has been cleaned considerably. The board and traces look remarkably good.
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 5:01 pm

LOL, this machine is obviously a survivor if it's been through all that. We should do all we can you get it running again :D

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 5:27 pm

IC33 - pin3 - nominal (mV)
pin4 - +4v

No difference if |I press *Break*
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 7:34 pm

So it looks like IC16 the 555 is the problem or one of it's passives, back in a bit need to have dinner

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 7:38 pm

I'm sure I checked IC16 previously.

I will go and get readings from all the pins.

What should I expect?
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 8:08 pm

Ok, so IC16 has nothing but noise.

It is a tiny voltage (mV), it appears to be the same 'waveform' but it is literally a mess.

The scope doesn't even pick up a frequency. From any pin.
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 9:11 pm

You should have +5V on IC16 pins 4 and 8
pin 2 should go low when you press BREAK
pin 7 and 6 should go low when you press BREAK
pin 1 should be ground
pin 3 should go high when you press BREAK

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 9:16 pm

If you do not have +5V on pins 4 and 8 you've lost the supply voltage.

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 9:18 pm

Right, I'm back up in the workshop.

I do actually have +5v on pin4

but only +3.3v on pin8

Also, how hoot should the VIA at IC6 be getting?

I know they get toasty but this one is currently at 43 degC.

Actually thought it was hotter than that... Lol.
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Silent, nothing visible

Post by Kazzie » Thu Aug 06, 2020 9:37 pm

here's a set of refernce readings from my Beeb, before I turn in for the night:
1 0.03V
2 4.50V
3 0.04V
4 4.96V
5 3.32V
6 0.04V
7 0.04V
8 4.96V

When break is pressed, pin 2 reads 0V, pin 3, 6 and 7 are 4.40V.

If you want to read up on what exactly IC16 is doing, the 555 timer is configured as a monostable circuit. A web search will turn up loads of tutorials explaining how it operates, but feel free to throw any specific questions our way. :)
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 9:44 pm

Thanks Kazzie.

I've now gone back and checked and my IC 16 is within tolerances of your readings.

Just doesn't change when I press break

But if I short across the reset switch pads, it drops to zero, and the long boooop changes ever so slightly

And pins 2, 4 and 7 match your readings also.

So, the timer seems fine...
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

kfro
Posts: 33
Joined: Mon Nov 25, 2019 10:26 pm
Contact:

Re: Silent, nothing visible

Post by kfro » Thu Aug 06, 2020 10:08 pm

Thank Kazzie that will be useful

IC6 is the video processor and it does get, hot, my machine has had a heat sink fitted for example.

When you short the reset switch pad does IC16 pin 3 go high.

I still think you have intermittent IC socket connections. My IC16 is in a socket and assuming yours is too. Do you have any contact spray? WD40 would work at a pinch but use very little for obvious reasons.

I think the sound you are hearing is because the sound chip is not getting initialised because the processor is not running.
Last edited by kfro on Thu Aug 06, 2020 10:10 pm, edited 1 time in total.

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 10:10 pm

I will freely admit that I miscounted the pins on IC16 earlier, for some reason counting them top to bottom both sides rather than anti-clockwise.

Yes, I'm tired and inexperienced but learning so much.

I now have a check sheet that I'm putting together of all these tests and checks that once this machine is running should be a comprehensive set of checks for any failure. Lol.

I do think we are getting closer.

It's somewhat exciting when you test something and find it is either working...or not!
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

Colday
Posts: 341
Joined: Thu Jul 25, 2019 10:18 pm
Location: North Essex, UK
Contact:

Re: Silent, nothing visible

Post by Colday » Thu Aug 06, 2020 10:12 pm

kfro wrote:
Thu Aug 06, 2020 10:08 pm
Thank Kazzie that will be useful

IC6 is the video processor and it does get, hot, my machine has had a heat sink fitted for example. - I have some to fit to all my machines

When you short the reset switch pad does IC16 pin 3 go high. - Yes

I still think you have intermittent IC socket connections. My IC16 is in a socket and assuming yours is too. Do you have any contact spray? WD40 would work at a pinch but use very little for obvious reasons. - My IC16 is soldered directly to the board.

I think the sound you are hearing is because the sound chip is not getting initialised because the processor is not running. - But I thought we proved the processor was good. We had all the right freq's and activity on the right bus' etc?
BBC B's... I now have 6!
I also have 1 boxed with manuals, unmolested model A.
And also an unmolested model B. (but not boxed sadly)
12x floppy drives (only 1x currently works I think)...
Learning to repair and refurb keyboards next! No more sticky keys!

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Silent, nothing visible

Post by Kazzie » Fri Aug 07, 2020 7:12 am

We're not that confident the CPU is getting everything it needs. In order to run, it needs:
  • A clock signal going in on pin 37 (and seen coming out again on pin 39)
  • A reset signal on pin 40, going low for a few clock cycles, to kick-start the CPU
  • Access to instructions from ROM using the data and address bus, to keep it doing stuff and not crashing on an illegal opcode
  • For the Beeb's MOS, working RAM in order to boot correctly
We've determined that you've got various clock signals being generated some of the time, but I don't recall if you've checked that they're coming to the CPU on pin 37. Do you see a clock signal here?

IC16 is responding to shorting the reset switch. Does pin 40 of the CPU go low when you short the reset switch? Without this, the CPU will never start executing instructions.

Check for any activity on pin 7. This shows when the CPU is fetching an instruction. If there's no activity, then the CPU isn't doing anything. It may not have been kickstarted correctly by a reset signal, or it may have encountered an illegal opcode (instruction) and crashed.

Two external signals can grab the CPU's attention, the IRQ and NMI pins (pins 4 and 6) allow other chips to send an interrupt request, or an urgent non-maskable interrupt to the CPU. They are active low, like the reset line, and if they're stuck low then the CPU will never get the opportunity to run its own code. You've already probed the IRQ pin earlier, and found that it was high, though it may be as well to double-check. Check if pin 6 is stuck low.

(If you've removed FDC or Econet chips from the board, you may need to refit a jumper link to stop NMI from dropping low in its absence.)

If you're happy that there is a clock and reset signal coming to the CPU, no problems on IRQ or NMI, but there's no activity on the sync pin, then the next step will be to inspect the address and data buses, and their connection to ROM or RAM.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

Post Reply

Return to “8-bit acorn hardware”