Yet another RPC600 fault

Arc/RPCs, peripherals, RISCOS operating system & ARM kit eg GP2x, BeagleBoard
Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Yet another RPC600 fault

Postby Boydie » Wed Oct 18, 2017 5:42 pm

My (currently) spare RiscPC motherboard has stopped working. There was quite a lot of battery damage a couple of years ago, but I cleaned it up with vinegar and water and it was still working a few months ago.

I've just tried it again, and the significant floppy codes were for vidc virq and vidc sirq. So, another clean-up with soft toothbrush/vinegar followed by spin in the dishwasher, thorough dry in low fan-oven, and final brush-down with IPA was in order. After this, continuity between IC33 and RP16 was confirmed, as was between RP16 and VIDC.
Things have now improved somewhat. The code for virq has vanished, but I'm stuck with sirq error. After POST, the floppy makes its usual noise and the keyboard lights respond. No monitor hooked up yet, so can't see what's on screen.

I've trawled previous posts, but they all refer more to virq errors. Whilst some mention sirq, none seems to have solved it.

It's an issue 2 board (0197,100) so slightly different to the layout in the circuit diagrams.

I'm admittedly lousy at reading the circuit diagrams. I've learnt enough to be able to trace the video circuit, but the sound one has me stumped for where to look next. I'm guessing it's a bad track somewhere, but which components should I be looking between? Is there an sirq equivalent for RP16/vidc?

edit: Just hooked up a monitor. Following POST, it boots into supervisor and lookis okay. However, in the desktop the light grey (as in the RISC OS banner) has white stripes down it, as does the light grey taskbar. Dark grey (desktop background) doesn't.

dp11
Posts: 681
Joined: Sun Aug 12, 2012 8:47 pm

Re: Yet another RPC600 fault

Postby dp11 » Wed Oct 18, 2017 7:26 pm

Take out the Vram module and see if the desktop returns to normal. If so cleaning of the Vram and socket is needed. If that doesn't fix it you need to start to look for broken tracks

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Thu Oct 19, 2017 6:39 am

Vram's already been removed, as has everything else. All that's left is an ARM610 board, and a single stick of FPM RAM (several sticks tried, in both slots).

I'm resigned to tracing out tracks. Am I right in assuming the white stripes on the display will be down to a bad track either between one of the other RPs and VIDC, or the RP and its supplying IC?

But which ICs should I be looking between for Sound IRQ? My knowledge of the sound system is minimal, and whilst I've been sble to work out from the schematics what happens with sound once it's left VIDC, I can't work out how it gets there in the first place.

Or do both sound and video take the same/similar ICxx->RPxx->VIDC route?

User avatar
danielj
Posts: 5227
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester

Re: Yet another RPC600 fault

Postby danielj » Thu Oct 19, 2017 6:52 am

The vidc is responsible for producing sound, so that's where it starts. Go to town checking the continuity around those resistor packs, there are loads of fine links there that easily get knackered.. :?

sirbod
Posts: 663
Joined: Mon Apr 09, 2012 8:44 am
Location: Essex
Contact:

Re: Yet another RPC600 fault

Postby sirbod » Thu Oct 19, 2017 7:20 am

Pin 135 on VIDC20 is the sound IRQ, you also want to check pin 138.

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Thu Oct 19, 2017 7:58 am

sirbod wrote:Pin 135 on VIDC20 is the sound IRQ, you also want to check pin 138.

Pin 135 has continuity to pin 112 on IOMD, pin 138 has continuity to pin 113.

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Sat Oct 21, 2017 4:00 pm

Not had chance yet to check out continuity. However, I've just tried the machine with vram inserted - now post shows errors for both virq and sirq, but the stripes on the desktop have disappeared.

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Sun Oct 22, 2017 9:48 pm

The story so far...

Continuity checks out for IC33->RP16, IC30->RP13, IC26->RP11, and IC22->RP6. Also checks out for all those RPs to VIDC, and all those ICs back to IOMD. Resistance for RP16 to VIDC is slightly higher (nearer 2 ohms as opposed to the 1-1.5 ohms for the others) but that's all.
IOMD pins 112 and 113 have continuity to pins 135 and 138 on VIDC.

There are strange vertical stripes on certain colours if no vram is installed. If vram is installed, these go away.

POST shows sirq fault whether vram is installed or not.

Changing ram stick and/or ram socket makes no difference.

Presumably this suggests a bad connection in an area which is served by main memory (as I assume sound always is, and video if no vram present) and bypassed by the presence of vram. However, looking at the circuit diagrams, I'm struggling for where to look next.

Any suggetsions?

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Tue Oct 24, 2017 11:04 pm

After another vinegar/water/ipa wash, virq stays off no matter what now. The major niggle now is sirq timing error (which may or may not be linked to the banding on screen if main ram is used for screen memory).

I've also tested all the RPs for internal continuity, and they all have the correct internal resistance.

Another thread commented on the absence of continuity between Pin 20 of IC5 and pin 124 of vidc on a Mk 3 board, and wondered if this was normal. Mine doesn't have continuity between these either, but making the link doesn't solve the sirq error. Should pin 124 of vidc be connected in a Mk2/3 board. If so, to what?

IC36 (the sound "chip") looks as though it may have taken a bit of a battering by acid. However, this is downstream of vidc, correct? Whilst it will clearly impact on the machine's ability to reproduce vidc's sound output, is there any way this can be contributing the the sirq issue?

Pin 123 of vidc (pclk) is connected to pin 107 of iomd.

edit: Pin 124 of vidc is connected to pin 6 of ic4 on the Mk2/3 boards. This is as opposed to pin 4 of ic4 (and thence to IC5 and vidc) on the Mk1 board. Apparently IC4 is a hex inverter (AC04 series). What pin 6 gets its input from (well presumably pin 5) is anyone's guess...

Boydie
Posts: 136
Joined: Sat Oct 24, 2015 8:25 am
Location: Sunny Wigan

Re: Yet another RPC600 fault

Postby Boydie » Wed Nov 01, 2017 4:42 pm

RPs 14 & 15 check out okay to +5v one way, and appropriate pins of VIDC the other, so that's the upper 32 data bits to VIDC checked out, right?

Interestingly, changing the OS from RO3.5 to 3.6 introduces a new wrinkle. Under 3.6. the power-up screen is black text on red background, rather than the more usual white-on-black. Desktop colour palette is somewhat screwy too; under 3.5 it's as you'd expect.

Any suggestions where else to look?


Return to “hardware”

Who is online

Users browsing this forum: No registered users and 3 guests