Hope for dead Vic 20?

[AnacreonZA]

A friend of mine has loaned me his Vic 20 to see if I can revive it for him. He said that is stopped working some time in the 80s, but since he was so attached to it he has kept it in very good condition ever since. I have some limited experience with older video game hardware so I said I would have a look at it but I made no promises.

At first I had hoped that it may just be the RF modulator that had failed, but after finding a DIN plug and soldering a new video cable up it still does not generate readable text. I have attached a screenshot of what it does when booted up. Basically vertical lines bending over to the right at the top of the frame.

The machine (a later model Vic with the C64 style PSU) seems to be booting up properly. If I type LOAD, and then press play on tape and then RUN/STOP the cassette stops turning. If I then type LOAD again with the play button already down the tape starts up again. I'd assume this means the CPU and BASIC is at least working? I tested the PSU with my meter and it also seems to be producing the right voltage (about 5.2v DC and 9.6v AC)

I also own a PET/CBM 4032 so I removed the 6502 from the Vic and booted the PET with the Vic's CPU and it worked perfectly. I took the risk of booting the Vic with the PET's 6502 but it made no difference.

I also have 2 game carts for the Vic (Avengers and Cosmic Jailbreak) and if I boot the machine with those inserted and then press F1 (which I figured out is necessary to start the game from VICE) I hear sound effects and I can hear that the game is running. That gives me at least some hope that the VIC chip itself is working. I just get no proper video. The screen is either black or constantly trying to sync. I do get a flashing screen when the player dies however - just no actual pixels I can see. I know my video cable is working because I bypassed it by soldering wires directly to the pins of the DIN socket going into the motherboard and I get the same result.

Sadly I have no working CRTs left available to me so I use a SCART to HDMI converter for display - but it has always displayed everything I've thrown at it before - even if it does take a few seconds to sync up sometimes. For example my (modded) PAL Master System, NTSC NES and NTSC Super NES all work perfectly through composite video to the device so I doubt there is something special about the Vic's composite video that is confusing it. In the screenshot it says SECAM, but it skips between PAL and SECAM as it tries to detect the signal.

I have also replaced three of the capacitors (electrolytic ones I had spares of) but that made no difference. I have not yet started replacing all the ceramic ones. Would this be the next logical place to look or would there be some other place to look before I undertake that difficult task (at least for someone with limited soldering skills?)

[Mike]

I tested the PSU with my meter and it also seems to be producing the right voltage (about 5.2v DC and 9.6v AC)

Check the ripple on the DC rail while the VIC-20 is on.

The VIC chip is quite sensitive to AC components on the 5V and will refuse to display a picture if the ripple exceeds somewhat over 0.1 Vrms.

Main cause for this is a PSU where the stabilizing caps have degraded too much.

[AnacreonZA]

Ah - an opportunity to see if my dad's old scope still works. I've never used one before but something new to learn.

Otherwise I have lots of fairly decent 5v DC power supplies from other projects - up to 3 Amp. Does the 9VAC only run the datasette? Will the machine display anything with only 5v DC? Is it safe to boot it that way?

Thanks for the suggestion.

[Mike]

You wouldn't actually need a scope to measure the ripple on the 5 V DC rail, simply take your multimeter and set it to VAC. The 5 V DC component then is 'ignored'.

However, a scope with(!) probe is quite handy to check the various oscillators, and even the non-periodical signals on the busses can give good hints which parts are still alive ... if you can match the circuit diagram with the board before you.

[AnacreonZA]

I see.

I recently acquired a C64c and after trying the PSU from that on the Vic I get the exact same result. The Vic's PSU powers up the C64c as well. Seems like its not as simple as I'd hoped.

[Spirantho]

Check pins 2 and 3 of the VIC chip (is yours PAL or NTSC?) - you should see the sync pulse on pin 3, and should get other data on pin 2.

There's a schematic here:
http://zimmers.net/anonftp/pub/cbm/sche ... 1027r1.gif

Check the output of Q2 with the scope and see if the sync pulse is there too - I suspect it isn't, as that's connected to the video output.
Change C17 if you've not done so already (it's electrolytic).

I would also strongly recommend trying your VIC chip in another VIC-20 before you do too much.

[Clockmeister]

Looks like VIC is dead. Swap with a known good one to confirm.

[Spirantho]

Make sure you put the suspect VIC in a known good machine not the working VIC in the suspect machine, as it may have more than one fault.

Speaking of which, look what arrived in the post this morning....

[AnacreonZA]

It's a PAL unit. I would love to be able to swop the VIC out but I have no source of such chips - although I've been watching ebay.co.uk for a while to see if any pop up. When disassembling the machine I did see a brown mark where the VIC RF shielding box touched the paper RF shield - so things seem to have gotten fairly hot in that area at some point.

[Spirantho]

Those chips in my picture are all PAL. I'm going to put a few up on eBay as I've not seen any on there, and a few on my shop, and keep others for my repairs. They're all 1984 so they're really late in the production run... I reckon I can get plenty more.

You're in the UK, yes? Do you want me to send you one to test in your VIC? Or you could send me your VIC to test in one of my VIC-20s.... that'd be a better way but I can understand you may not want to. Then if it doesn't help you can just send it back. Be aware, though, that if there is more than one problem, you may be sending back part of the solution, if you see what I mean.

If you find the new 6561-101 is what you want I would need payment of course (they cost me to get them ).... I'm going to put them on eBay for £29.99 but you can have it for £19.99 if it fixes your Vic.

[Clockmeister]

Those chips in my picture are all PAL. I'm going to put a few up on eBay as I've not seen any on there, and a few on my shop, and keep others for my repairs. They're all 1984 so they're really late in the production run... I reckon I can get plenty more.

You're in the UK, yes? Do you want me to send you one to test in your VIC? Or you could send me your VIC to test in one of my VIC-20s.... that'd be a better way but I can understand you may not want to. Then if it doesn't help you can just send it back. Be aware, though, that if there is more than one problem, you may be sending back part of the solution, if you see what I mean.

If you find the new 6561-101 is what you want I would need payment of course (they cost me to get them ).... I'm going to put them on eBay for £29.99 but you can have it for £19.99 if it fixes your Vic.

Have you tested those VIC's? People have been stung with bulk fake SID's, so nothing would surprise me.

[Spirantho]

I've spent a lot of time and money finding reliable Chinese suppliers, to try and find genuine SIDs that fully work with good filters. I've been getting CIAs, CPUs, VIC (i and ii) etc. at the same time as ordering SIDs, so that even if the SIDs I buy are faulty, I'm still getting my money's worth on the postage cost.

These all check out fine, and are genuinely brand new from a 1984 batch!

(and my current working SID count is about 70.... )

China has a lot of dodgy companies, but just like in the West, there are also very good companies. It's just a matter of finding which is which.

These 6561-101 chips came with 4xMOS8501, 8xMOS8360, 6x6569R3 and the 2 samples of MOS6581R4AR. Of these, one 8501 and two 8360 chips were DOA, the rest all work perfectly as far as I've tested so far, but generally if they boot up they're likely to be ok. The two SID chips checked out perfectly, by the way. Filters seem perfect.

[Clockmeister]

Fair enough, I'm glad you found reliable suppliers as there are a few horror stories out there with either completely bogus remarked chips or factory reject chips that have been remarked and have faults out of asia.

I have a spare 6561 from the last few I won on Ebay from a gent in the UK. Two were dead (well, not completely dead, as sound worked fine but there was no display), one was used but OK and a couple of NOS . It didn't bother me because the seller was honest about the condition and could only guarantee that the two NOS ones should work, the rest was as-is, and the few quid postage (to Australia) was more then I paid for the lot anyway. One NOS one lasted about 12 months before it gave out though, and it even had a heatsink and fan to keep it cool

I've noticed a lot of mask ROM's and DRAM are failing in 80's retro stuff in general, perhaps those IC's have reached their end of life.
What is your experience in that regard, are you seeing increasing DRAM/ROM failures?

[Spirantho]

No more than any other chip.
It's the complex chips that are the trouble. I'm quite used to bad RAM now, so I have lots of various types of RAM chips (enough to keep VIC-20s and their RAM expansions going for quite a few years ).
I do find that 41464 chips are very prone to death, though (and taking out their brethren on the bus in the process). Pretty much every Spectrum +2A/+3 that needs repairing needs new DRAM, same with Amstrad Plus/GX4000 machines. I find 4116 chips get a bad reputation but they're no worse than any other for failing in a working circuit.

Apart from the chips with design flaws (I'm looking at you, Mr. MOS8501) there's not much difference in mortality rates between chips, in my experience.

[Clockmeister]

Well that is interesting. My experience is that about 70% of failures in retro equipment is due to RAM/ROM failure.
Dead glue logic is pretty rare (though my last repair was a failed 74LS14 on an MC-10). Custom chips and dead CPU's/VDG's would make up the other 30% of chip failures.

I suppose one explanation might be that there are environmental differences affecting some components more severely.