well done ,these PCs are not only historic they are still capable machines believe it or not and a growing community of modders

Good to see some Risc PC content! I had very similar issues to you I replaced a SIMM socket, with regards to the VRAM I ended up replacing the chips which resolved it, you may need to run some tests and look for compatible replacements.

RP14 is a common failure point following a Varta leak, along with the other buffers and resistor packs nearby. Look there for the VRAM issue. When it's detected you'll see 1mb more reported at boot. So 32mb+1mb vram would say "RISC OS 33MB". Generally it's a good idea to hold DEL until you hear the boop noise, and get yourself a battery on the CMOS otherwise you'll be needing to power on with DEL held every time you power on after more than a few minutes without power. The PCF8583F will fill with garbage that can cause a machine to behave very unpredictably, and differently across powerups. Battery and confirmation it's holding the settings is a very good idea before moving onto the RAM issue. Fit a single 1.2v AA NiMH low self discharge cell in a box, leads long enough to reach the AA sized cavity at front of the machine (bottom slice). Use header pins and female dupont connectors so it can be removed easily from the PCB.

I enjoyed your video but I unfortunately know very little to not about Acorn RISC PCs. In any case, good luck with the further repair. Cheers

When the VRAM is working you’ll see it in the RAM count at POST - so it will say 33MB or 65 MB if you have 1MB VRAM.

With that riser, looks like this would have had some extra slices added to the case. Would have been nice to see. These were the fancy computers in my school, but they never had the extra case slices.

Hey Mark!  Enjoy watching your content.  Did you check to see if the individual ram ics had 5v across vcc and gnd?  If I was you I would look at the data sheet of those chips and test that to see if they are getting power, through that particular chip or none of them, to see if you need to keep looking to where the power feeds into the socket.  After that, I would check all of the address lines to see if they are getting where they should.  Good luck!

Also, the diag sequence indicated by the floppy drive LED during boot is fantastic as a debugging aid. It may show problems that aren't obvious, like VCO timing issues (which on one of my boards was due to the 2 small electrolytics to the right of the VIDC chip). The lighter green main board (the one you have) seems to have been bitten by the "capacitor plague". I have two of them and both had many caps with extremely high ESR, but also have a dark green main board with all caps in spec.

I had two of the 1MB vRAM modules. One worked fine, the other simply wasn't detected. I wanted to upgrade the non working one to 2MB, so basically rebuilt it with new HM538253BTT-7 modules (fitting the additional on the other side as well). Not sure of it was one of the ram modules that was faulty, but i did find that several of the electrolytics were basically resistors. I replaced those also and it works fine (2MB detected).

I had a 2 slice RiscPC with a StrongARM 233MHz and a 5x86 CPU About 15 years ago. Then the battery leaked and utterly killed it, I ended up selling off the working parts and throwing the motherboard away. I probably could have fixed it now with what i've learnt since.

are you lookin Adrian's digital basement video channel on youtube about his acorn archimede repair and restore ? :)

You mentioned you checked around there already, but I'd check very carefully every connection in and out of IC33 and RP16, they are notorious for rotten vias which are terrible to repair (Hence I'm working on a PCB to replace those 2 packages on the back side), the 74ACT244's also have a habit of failing if they've had corrosion on the legs, as does the CMOS chip which is why I made a podule version for easy testing, although your not getting a red screen so we'd pick that's ok now (that 47ohm resistor nearly always falls off 😂). The VRAM it's also not unusual for the traces to have hairline corrosion where the trace meets the contacts right at the edge of the solder mask, but at least on first glance it doesn't look like the blue crystals of death were growing on that socket like they were on the RAM. But the purple and blue screen shouldn't usually be visible at power on suggesting a fault somewhere still

Try swapping the two drives around to boot from the other one.

18:41 It's so satisfying to see someone else running into those "let's clean it until it breaks" issues 🫠🤘But I think in your case it was a good thing to finally have it break and be able to search for the real failures.

The missing resistor at 16:33 was there earlier in the video (12:32) its 470 by the looks of it.

wonder if the vram caps need to be changed. that's smd caps so... suspect?

🍪

I made a promise to myself, that I won't get more into British micros (beyond the few I already own) because it's such a deep rabbit hole. But every time there's a Risc PC video, I want one. Even despite the horrific battery leakages. Send help. (Or send a Risc PC!)

I dunno, I can't say I know SWI &4000F either, so I reckon it's fine. Actually considering it says "Application may have gone wrong", it seems like something running on startup is crashing which could be due to hardware but I'm pretty sure the safe bet is to blame the software (developer), even in 1994.

Oh shit, don't look in the 'Convict' folder. Not sure if verb or noun but either way...