A3000 strange interlaced resolutions display

[schlika]

Hi All !

My Amiga 3000 has a strange fault with interlaced resolutions only, through vga output and internal video bus. I tried everything I could but nothing has any effect. It is a "Revision 6.1 modified" mainboard.

I did a few updates to it :
- replaced battery (no leakage)
- burnt Kickstart 3.1.4 on period correct eproms
- updated scsi chip to a WD33C93A-PL
- updated buster to rev-11
- replaced all the GAL's by PAL's
- full 16Mb fast FPM ZIP + 2Mb chip

A friend of mine lent me his "vanilla" perfectly working A3000, and was kind enough to let me try and replace every chip.

I did, starting with Amber, and moving on to every socketed chip, and even desoldered/resoldered HY480. No difference.

Please have a look at the screenshots. Non interlaced resolutions look OK, but interlaced ones are awful, with or without the deinterlacer (tested on my Dell 2410). Video output through the db23 connector is perfectly fine, even for interlaced.

I also have a ZZ9000 card, and internal video connector also suffers from the same trouble as the vga output.

As you see, I tried both AmigaOS 3.1.4.1 & Amix, both have the same issues.

So it seems to be a non socketed component active on the internal video bus, the vga output, but not on the db23 output, and only for interlaced resolutions.

Any clue ?

Thanks !

[DarrenHD]

Have you had any battery damage on your A3000? I have an A3000 rev 6.1 also and I have a similar, but worse, issue. I did have battery leakage though. My interlaced resolutions are totally corrupt, but only thru VGA/Amber. If you look at this video a guy had my exact problem and ended up replaced the chip marked U480, DM74ALS74AN. It resolved the issue.

[ Show youtube player ]

[schlika]

Cool thanks a lot I'll definitely try that !

[pipper]

I also have a ZZ9000 card, and internal video connector also suffers from the same trouble as the vga output.

Thats a pretty strong indicator that the A3000 flickerfixer/scandoubler circuit is not to blame, but what feeds both circuits: the digitial R/G/B lines coming from Denise. But since the regular RGB-Out is fine, it can't be Denise to blame.

If so, the next chips in line to test are U451 and U452. Try checking their output level and continuity with Amber, U470, U471 and U472, U475 and U479.
I would assume the fault to live in one of the common lines that feeds into one of these chips and the internal video connector at the same time.
Its also possible that the issue lies with the daughter board and that the digitial RGB signals get somehow corrupted on there - maybe try swapping the daughterboard with your friend's?
If you test with fully white, black, red, blue, green screens, you should be able to see if there are any stuck color bits.

[schlika]

Allright, I'll try that too ! Thanks

[schlika]

Thank you so much guys !!

I replaced U480, way more work than anticipated due to some light battery leakage that rendered the desoldering process a nightmare, I finally managed to somehow solder a socket and plug in a new DM74ALS74AN ... Boom, it works !! I am super happy !!

[trixster]

Great result!

[DarrenHD]

Originally Posted by schlika

Thank you so much guys !!

I replaced U480, way more work than anticipated due to some light battery leakage that rendered the desoldering process a nightmare, I finally managed to somehow solder a socket and plug in a new DM74ALS74AN ... Boom, it works !! I am super happy !!

Glad it works. I still have to replace my U480 chip...