Ram Expansion Issue, No Block 5 RAM

[PTherapist]

I recently bought my first VIC-20 from eBay and everything seemed to be working fine.

So I bought a 35K RAM expansion for it and ran into an issue. Basically Block 5 doesn't seem to be working. With the full 35K enabled via dip switches, the RAM test only shows RAM present in RAM 1,2,3 & Blocks 1,2,3. But the RAM tests otherwise pass ok. Of course the VIC-20 will also not run any cartridge dumps which need the RAM in block 5, it just resets to BASIC.

Thinking the expansion may have been faulty, I requested a replacement. The replacement arrived, also exhibiting the same issue so I sent the replacement back. The expansion itself looks fine and I removed the 2 RAM chips and tested them with my MiniPro on my PC and they checked out ok also.

The only other cartridge I have for the VIC-20 is an 8K game - Jupiter Lander. This works absolutely fine, so it appears as if the cartridge port itself is ok.

For reference my VIC-20 is the later rev. d model with the rainbow logo and DIN power connector. I also have tested loading stuff from an SD2IEC, as well as a Pi1541, just to rule that out. I've also cleaned the cartridge port and the expansion's edge connectors, to rule out any poor contact issues.

Anybody have any ideas what may be up with my VIC-20?

[Vic20-Ian]

if the problem is visibility of block 5

1. Have a look at this, block A000 is not visible to Basic

https://en.wikipedia.org/wiki/Commodore ... _expansion

if the problem is real and a block 5 program will not run

2. Why assume it is the Vic20?

It could be the ram pack, if manually assembled I would look at soldering.

Look over the ram pack pcb, maybe there is a dry joint near the dip switch.

[Mike]

With the full 35K enabled via dip switches, the RAM test only shows RAM present in RAM 1,2,3 & Blocks 1,2,3. [...]

May I assume you are using this RAM check?

[T]he VIC-20 will [...] not run any cartridge dumps which need the RAM in block 5, it just resets to BASIC.

[...]

The only other cartridge I have for the VIC-20 is an 8K game - Jupiter Lander. This works absolutely fine, so it appears as if the cartridge port itself is ok.

This also means the signalling of BLK5 from your VIC-20 works. And this more or less rules out a fault in your VIC-20.

Thinking the expansion may have been faulty, I requested a replacement. The replacement arrived, also exhibiting the same issue so I sent the replacement back. The expansion itself looks fine and I removed the 2 RAM chips and tested them with my MiniPro on my PC and they checked out ok also.

As this already happened with two cartridges of the same kind, this might be due to a design fault in that RAM cartridge.

A photo of the cartridge PCB (component and soldering side) could help narrowing in the possible culprit.

[PTherapist]

May I assume you are using this RAM check?

Yep, using that test.

This also means the signalling of BLK5 from your VIC-20 works. And this more or less rules out a fault in your VIC-20.

That was what I initially thought too, which is why I contacted the seller and asked for a replacement. The seller says he tested them before shipping using that RAM Check program, which he linked to, which is how I initially discovered it.

Could it be a problem with 1 of the logic chips in the VIC-20 that would allow cartridge games to work ok in block 5, but not RAM? Or does that seem completely unlikely?

As this already happened with two cartridges of the same kind, this might be due to a design fault in that RAM cartridge.

A photo of the cartridge PCB (component and soldering side) could help narrowing in the possible culprit.

Here are pics of the cartridge in question -

[Mike]

That was what I initially thought too, which is why I contacted the seller and asked for a replacement. The seller says he tested them before shipping using that RAM Check program, which he linked to, which is how I initially discovered it.

That's also good.

Could it be a problem with 1 of the logic chips in the VIC-20 that would allow cartridge games to work ok in block 5, but not RAM? Or does that seem completely unlikely?

The logic in the VIC-20 doesn't differentiate between RAM and ROM in the select logic. When the chip select works for ROM in BLK5 (as with your game cartridge), it also works for RAM, but ...

... one of the DIP switches is supposed to set BLK5 on the cartridge between read/write and read-only. For the RAM check to work, this must be set to read/write. When you load cartridge data into BLK5, the switch also must be in the position read/write. Only in a few cases, where cartridges perform write operations to BLK5 (mainly as copy protection measure) you'll need to flip the switch to read-only *after* load and *before* reset.

Here are pics of the cartridge in question

The R/W pins of the RAM chips are correctly connected to pin 17 (VR/W) of the cartridge connector, so we can rule out another problem, where CR/W on pin 18 was used by mistake (CR/W on its own has the wrong timing for RAM chips).

Please check against the instruction manual of the cartridge for the correct setting of the BLK5 write inhibit DIP switch.

[PTherapist]

That was what I initially thought too, which is why I contacted the seller and asked for a replacement. The seller says he tested them before shipping using that RAM Check program, which he linked to, which is how I initially discovered it.

That's also good.

Could it be a problem with 1 of the logic chips in the VIC-20 that would allow cartridge games to work ok in block 5, but not RAM? Or does that seem completely unlikely?

The logic in the VIC-20 doesn't differentiate between RAM and ROM in the select logic. When the chip select works for ROM in BLK5 (as with your game cartridge), it also works for RAM, but ...

... one of the DIP switches is supposed to set BLK5 on the cartridge between read/write and read-only. For the RAM check to work, this must be set to read/write. When you load cartridge data into BLK5, the switch also must be in the position read/write. Only in a few cases, where cartridges perform write operations to BLK5 (mainly as copy protection measure) you'll need to flip the switch to read-only *after* load and *before* reset.

Here are pics of the cartridge in question

The R/W pins of the RAM chips are correctly connected to pin 17 (VR/W) of the cartridge connector, so we can rule out another problem, where CR/W on pin 18 was used by mistake (CR/W on its own has the wrong timing for RAM chips).

Please check against the instruction manual of the cartridge for the correct setting of the BLK5 write inhibit DIP switch.

Switch 1 in this case switches between R/W or read only. Doesn't make a difference what I set that to, still nothing showing up for BLK 5 and I can't load any BLK 5 programs. Switch 2 is BLK 5 and on or off this also makes no difference. All the rest of the switches work exactly as they should.

[Mike]

Switch 1 in this case switches between R/W or read only. Doesn't make a difference what I set that to, still nothing showing up for BLK 5 and I can't load any BLK 5 programs. Switch 2 is BLK 5 and on or off this also makes no difference. All the rest of the switches work exactly as they should.

This seems to be Ruud's design.

What looks a bit suspicious is how the right RAM chip, the 6264, sits in the socket. The 6264 is responsible for BLK5 here. Barring any possible perspective distortion, the lower pins of the chip don't seem to make any contact with the socket, as if the pin pitch of the chip is a little bit wider than the socket. This is not as apparent with the other RAM chip.

Do you have an Ohm-Meter to measure the resistance between chip pin 14 (from above, the lowest pin on the left side) against the corresponding solder point on the bottom of the PCB? If possible, extract the chip once again from the socket, and re-seat it to nudge it a little bit more away from the edge connector side.

Besides, the PCB design misses out on stabilizing capacitors for the three ICs, and the routing of Vcc could also be improved.

[PTherapist]

I think it may be just the angle of the photo making it look not in place there.

But I just tried reseating the 6264 anyway and also checked the resistance with my multimeter (digital multimeter) on the said pin, before and after reseating - got a reading of 0. Also a 0 reading from the socket itself with the chip removed.

Still no blk 5 of course, nothing has changed there.

[Mike]

Okay.

We're on the limit though what can be sorted out by remote diagnosis. Further, 'non-destructive' tests would require another VIC-20 to check the RAM expansion with, and another RAM expansion to check with your current VIC-20. If you add your home town to your profile, maybe another Denial member lives near you and you two could then arrange a meet-up.

As more invasive action, you could extract the two RAM chips and desolder the TTL logic chip, and then compare the PCB with Ruud's schematic. I don't expect any difference though.

However, as I already wrote, this RAM cartridge suffers from those design flaws I already mentioned: lack of stabilizing capacitors and problematic routing of the Vcc rail. Your VIC-20 is just sensitive enough to expose those flaws. I am pretty confident - given that your VIC-20 works with a ROM cartridge that uses BLK5 - your computer will also work with a better designed RAM cartridge. Even if that is more expensive.

Sorry I can't be of more help this time.

[PTherapist]

Thanks for the help anyway.

I'll probably have to live with it for now and look into either replacing the expansion or the VIC-20 with a different revision later in the year.

[beamrider]

Can you get a refund on the RAM expansion explaining it doesn't work and perhaps try a Penultimate+ instead?

[eslapion]

Here are pics of the cartridge in question -

I notice there is no gold plating on the edge connector. Another fine piece of junk made by an amateur who doesn't know any better...

[PTherapist]

Well despite my last post stating I'd replace it "later in the year", I happened to come across one of these on eBay for a fairly affordable price -

(image grabbed off google images)

Once it arrives, hopefully next week at some point, it'll be interesting to see if this works ok on this VIC-20. If it does work, it would at least confirm the VIC-20 itself is ok as everybody is suggesting. Hopefully so and I could probably live without the extra 3K.

[PTherapist]

Just an update to say the 32K expansion I pictured above turned up today and is indeed fully working, Block 5 RAM detected & all RAM testing perfectly. So yep I can confirm the VIC-20 itself is absolutely fine and the original fault must indeed be down to the poorly designed 35K RAM expansion as suggested.

Thanks all for the help.

Can you get a refund on the RAM expansion explaining it doesn't work and perhaps try a Penultimate+ instead?

Now that I know for definite my VIC-20 is ok, getting the Penultimate+ is something I'll probably consider in the future.