C128D displaying random characters on startup screen (plastic case model)

Got an issue?
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

Something interesting happens when I plug in a game cartridge like Avenger in.

The games appears to behave completely normal. Joystick, sound and graphics look fine. Note that there are no vertical lines on the screen.

IMG_20230114_120930.jpg
IMG_20230114_120808.jpg


User avatar
Gyro Gearloose
Member
Member
Posts: 471
Joined: Sun Nov 15, 2015 5:20 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by Gyro Gearloose »

That's great, although I can't explain it. I don't know the 128 nearly as well as the 64, but clearly we can see all the big chips must be OK.

I think it's a jellybean logic chip, or maybe much worse, an intermittent trace/via somewhere.

What else can you test? Can you check for MOS branded logic chips? Those are never good news.
The price one pays for pursuing any profession, or calling, is an intimate knowledge of its ugly side.
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

Hi Gyro Gearloose,


I really appreciate all your help! Thanks.

Yes, it looks encouraging. There are some issues. I think you are right about the jelly bean logic chips.

It does seem like the main chips are working thankfully.


I'm only mediocre hobbyist at best.

I have been studying hard about what is classed as ground referenced in regards to safe hook up procedures with oscilloscopes.
I could be wrong here. Any advice would be great.
I have been using a multimeter and carefully measuring continuity of the mains power earth pin of the C128's power supply as well as the mains earth of the oscilloscope.making sure that the active line and the neutral line is wired the same was around on both devices.

Here's my thoughts.
I was thinking if I plugged the oscilloscope and the C128 power supply into a power board. Making double sure that only the ground of the BNC connector for oscilloscope and the ground of the PCB board of the C128 have continuity.
Checking as always to see if there is no voltage/ current differential between these two grounds.
I could then possibly use the oscilloscope probe at 10x without the little black ground clip that attaches to the probe to take some measurements.
Of course not having the gound clip as close to the probe tip as possible will no doubt lead to issues with accurate readings.

I would like to just be able to see some wave forms from some of the support logic chips.

I suspect that a jelly bean logic chip may be outputting a marginal signal every now and then.


Of course I checked the house wiring is up to code with RCD breakers.

I will go through the PCB board and check for MOS chips and report back.


I will post some more detailed observations. Maybe you may have some suggestions.
rmzalbar
Member
Member
Posts: 257
Joined: Wed May 08, 2019 9:06 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by rmzalbar »

It looks like a bad or flaky address since the stripes are showing on the same 8-pixel increments on a character mode screen. Maybe a bad char rom pulling something down when enabled, or a bad address decoder 74LS leaking across on certain combinations?
Smooth operator
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

Hi rmzalbar,


Both very probable causes, thanks for your thoughts. I will investigate and report back. :D
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

Hi rmzalbar,



Following on from you hypothesis of a corrupted CharROM chip, I made some observations of the lines generated by the C128 computer on the the screen.

Here is the problem. Lines in the character display area of the C128 & in C64 modes.
Resizer_16738515366551.jpg
Resizer_16738515366550.jpg
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

Now if I run a diagnostic cartridge program such as this...
IMG_20230118_150037.jpg

Or this......


IMG_20230118_153227.jpg
I think there is an TOD error on one of the CIA chips in this image. I think it's because the MOS 8520-A CIA is meant for an Amiga computer. I understand that the TOD circuitry is not active in these Amiga CIA chips.

One can see the expected problematic lines.


When running a SID cartridge something unique occurs. The lines only seem to be visible where there is text.

IMG_20230118_160141.jpg
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

However if I run a C64 deadtest cartridge, such as this.....

IMG_20230118_151616.jpg
Something totally different occurs.

This deadtest cartridge operates in Ultimax mode. I do believe that this particular version bypasses the CHAR ROM chip all together and generates it's own custom character set.

Notice that the vertical lines are no longer visible (just to reiterate that these bars are not to be confused with RF interference or the commonly named Jailbars).

I will have to do some further testing and post further observations.
banman
Member
Member
Posts: 419
Joined: Sat Jun 15, 2019 12:21 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by banman »

rcade wrote: Mon Jan 23, 2023 1:36 am Mine did that also- it easily breaks free during shipping. It does fit in the holes, or at least mine did and I resoldered it.
Hi rcade,


I stand corrected unreservedly. Upon removal of the back PCB board RF shield I noticed this.
IMG_20230404_141705.jpg
IMG_20230404_141716.jpg
IMG_20230404_141700.jpg

Looking at the back PCB board one can see signs of flux residue on areas where larger thru hole components were added at the assembly line after the wave soldering bath. I am very confident that this board hasn't been opened or modified by anybody but me since leaving the Commodore factory.
Looking closely at the area that contains the RF shield can for the 2 video chips and associated components one will notice that there is evidence that the RF shield can pins may have been all the way thru to the back side of the PCB board.
It appears that the RF shield can pins have been pulled back out while the solder was still hot....

I did do a closer examination of the RF shield can pins is very bright light and I was able to scrape off solder residue with a very sharp knife. Admittedly on initial inspection the RF shield can pins look as if there is no solder at all.
I cleared the PCB board holes out a little and friction fitted the RF shield can back to the PCB board. When I clipped the top RF shield can lid on the spring tabs were strong enough to pull the RF shield can back out of the PCB board.

I replaced the RF shield can back to the PCB board. It took a great deal of time and heat to accomplish this by way of the large mass of metal to heat up. One thing I did notice. The addition of the RF shield can to the PCB board stiffened the PCB board up considerably. Without this this very large PCB board is very flexible and bendy, which is probably not a good thing

A question....
Could there have been a possibility that at the factory assembly line the RF shield can plus lid was correctly installed, then before the solder had a chance to transition to a solid the spring tabs on the RF shield lid had enough force to pull the RF shield can back out of the PCB?
rmzalbar
Member
Member
Posts: 257
Joined: Wed May 08, 2019 9:06 am
Contact:

Re: C128D displaying random characters on startup screen (plastic case model)

Post by rmzalbar »

Have you tried swapping the char rom yet or scoping the data lines on the char rom while displaying a corrupted text screen? All 8 of them should be showing a changing square wave pattern. You'll need to trigger on chip select..
Smooth operator
Post Reply Previous topicNext topic

Who is online

Users browsing this forum: No registered users and 5 guests