Freeze after warmup

[Linzino]

Hi

One of my Vic 20s (C64 power plug) has the following problem:
it freezes if IN BASIC mode after some time but it works rock solid if I use a cartridge.
Once it freezes I sometimes get a white screen with visible borders and a restart is only
possible after some time. Something is heating up too much and causing a problem, ONLY
visible in BASIC.

Is the BASIC rom the most likely culprit?
Should I test the RAM chips?

Linzino

[mrr19121970]

Which cartridge are you using?

You should try text (Scott Adams adventures), graphics (avengers) and multicolour (gorf) too to check the symptons.

1st guess is a damaged VIC-I chip. Try Ray Carlsons heat sink recommendations with epoxy and copper making contact with the RF cage.

http://personalpages.tds.net/~rcarlsen/ ... 20sink.jpg
http://personalpages.tds.net/~rcarlsen/ ... nk%202.jpg

[Linzino]

Hi

All cartridge games seem to work rock solid except the ones that require BASIC to start.

The Vic-I seems OK to me.

[Kakemoms]

Well, some of the Vic-20's had poor solders and as things heat up over time it could be a solder that has a crack in it.

If so, I would suspect something connected to the Basic ROM or related memory (<$0400).

[srowe]

My VIC suffered this same issue, swapping the BASIC ROM with another system and the problem followed the faulty ROM.

[Linzino]

Hi

My description of the symptoms was not accurate.

Vic-I is and CPU are OK because swapped with working and tested ones.

Things are worse: some game cards do not work:
- Dig Dug is stuck/frozen at start-up with no music
- Deamon Attack seems to work but the main ship is not displayed.

BUT most other single color (Avenger) and multi-color games (Gorf) seem to work with no issue.

Which chips are the most likely culprits?

I now think the BASIC rom cannot be the (only) problem.

[mrr19121970]

As I suggested over PM.

Ram or char rom maybe. Psu could be a cause too?

[Linzino]

I fear it is something else because:
- Dig Dug is not using any built-in characters and a failing character rom should not freeze such a game
- RAM may be the issue but I cannot detect any corruption on the screen even if I re-map video memory to different
locations (poke 648,x: sys 58648)
- PSU is OK because used with lots of other Commodore computers

[srowe]

Things are worse: some game cards do not work:
- Dig Dug is stuck/frozen at start-up with no music
- Deamon Attack seems to work but the main ship is not displayed.

If these games use timers then perhaps one of the VIAs is faulty?

[Linzino]

I have also swapped both VIAs with working ones and nothing changed.

I confirm:
- CPU OK (swapped)
- both VIAs OK (swapped)
- VIC-I OK (swapped)
- PSU OK (works fine with all other VIC20s and C64s)

Symptoms are:
- Dig Dug frozen at startup with no sound and
- Daemon Attack (main ship not shown)
- Basic freezes after a few minutes

The other cartridge games work solidly with no freeze.

No corruption is seen on screen at start up and no corruption could be
detected if I remap the video memory to different locations through poke 648,x:sys58648.

[srowe]

OK, so we're starting to run out of candidates.

My top suspect would be RAM, if page 0 or 1 were faulty then all sorts of odd behaviour could occur.

This is a CR model right? So you have a single 2K x 8 RAM chip?

[Linzino]

This is a CR PAL model with
2 6116 chips (2kbyte each)
and
3 2114 chips (0.5kbyte each) including color ram

Previous problems fixed
This model had a loading issue which I fixed by swapping a faulty 7406 (UB4).
By doing this I inadvertently cut one line which I repaired with a bridge.
One more line was slighly scratched but continuity seemed OK to me
on all lines.

[Linzino]

The specific graphical problem seems to occur in those games that seem to do some tricks
with the raster and probably with the timings to get more colors such as:
- Atlantis (ships have just one color instead of the many colors expected)
- Daemon Attack (ships are not shown but the bottom is shown)
- Dig Dug (freeze at startup but I do not know if it is using any tricks with scanlines)

BOTH plain high resolution and multicolor modes are OK in most games.

Should I also suspect the 7402 (MASTER CLOCK OSC)?

[Mike]

That one beforehand:

Should I also suspect the 7402 (MASTER CLOCK OSC)?

If that was faultly, the VIC-I wouldn't display anymore, so we can exclude that for the moment. I would suspect, that the circuit trace of the IRQ line is the culprit. Can you check for continuity between Pin 21 of VIA #2 (UAB1) and Pin 4 of the 6502?

Missing interrupts will make the cursor stop blinking and the keyboard wouldn't work anymore. To differentiate this from a real CPU crash, try the following: as long you can type in something in BASIC (before BASIC "freezes"), type in this small program and run it:

Code: Select all

1 A=A+1:PRINT A, TI$:GOTO1

When the interrupt stops working, TI$ freezes. However when the CPU still operates, A keeps ticking, and that will show.

[Linzino]

The CPU sort of froze but the CPU is OK because I have swapped it with a good one. I can recheck.

I have attached:
- a photo of a bad fix for a 7406 with a slighly scratched line and a repaired one
- Atlantis on a good Vic 20 with colored ships
- Atlantis on the bad Vic 20 with missing and wrong colors
- a screenshot of the counter program frozen (I used TI instead of TI$ to see smaller intervals):
the time and the interger counter stopped at the very same time)

At the moment I don't have a multimeter at hand. Once I get one I will check the lines you suggest.