A600 faulty video

[jimbob]

Hello, I have here a rev2D A600 with broken video output. Otherwise it seems to be working, boots ok from floppy or IDE, plays sound etc, just faulty video output. Attached is a photo of what is supposed to be the purple kickstart floppy insert animation. You can sort of make out the tick and part of the floppy animation.

I have just changed all the caps on this board and it made no difference to this fault.

The denise chip gets very hot to the touch. Denise gets hot in other A600s I have but this one is painful to the touch after a second or two. This makes me suspect denise itself or the DACs hanging off its outputs.

Otherwise I would suspect U12 video encoder and and/or associated transistor circuits. Also blowing on this chip affects the colours in the outputs. (though maybe this is also cooling denise/DACS).

Composite looks the same as RGB. RF suffers lots of interference, sometimes looks the same.

Would appreciate any troubleshooting tips anyone has. Also where to get replacement DACS, CXA1145 or PLCC Denise?

Is there such a thing as a video testing program, maybe which cycles colours to produce predictable wave forms at points in the circuit which can be tested with a scope, or boots to some sort of testcard screen which helps identify various types of faults.

[Khyron]

Start with the basics..
measure the supply voltages at each part for:
U31/U32 (Pin20), U12 (Pin 19&12), +VID for the Restor DACs & Buffers (Q211-Q213), and +VOUT (Q231-Q233) for the output buffers..

If you have a scope, you can do a quick test on U31 & U32 and just compare the input vs the output of each color bit on each buffer to see that they're the same while the amiga is sitting at the insert disk screen, there should be enough bit movement to see what is going on.. and do the same type of check for the RGB input on U12 vs the RGB out of U12, and the RBG out of U12 vs the RGB at the DB23. These test should pretty much localize the problem to a specific part.

The dacs on the A600 are just a simple restor ladder highly doubtful they're damaged...
The CXA1145 is a little hard to find, but you can usually salvage them out of game consoles.. or ebay.. the Erso ES71145 and the fujitsu MB3514 are also compatible parts that should be easier to find..

and good luck!!

[altcomputing]

Originally Posted by jimbob

Also where to get replacement DACS, CXA1145 or PLCC Denise?

If you know how to solder them back on board, I could offer you a replacement Denise. My wife's sister is flying back to Scotland this Monday, so no international postage would be required.

EDIT: I suppose the video encoder chip has nothing to do with RGB output... So most likely Denise is broken.

[jimbob]

Thanks for the tips Khyron. After a closer look at the schematic I see what you mean about the DACS. U31 and U32 just buffer the outputs of Denise into binary weighted resistors then a summing amplifier. I had thought the chips were the DACs after reading in another thread about troublesome DACs.

This prompted me to suspect U31 or U32 even more strongly since if the buffer had failed somehow, this could cause excessive current draw from Denise explaining the heat problem. Because of the heat I was reluctant to power it up long enough to probe about so I just used my trusty multimeters continuity/diode tester and found what I'm pretty sure is a failed input on pin 2 of U32. Straight away, this will mess up the MSB of the red signal and the excessive current draw and heat probably messes everything else up too.

These buffers are 74HCT244. I think I can just replace them with any standard CMOS equivalent. I'm not quite sure if they are inverting or non-inverting buffers though, Anyone?

@altcomputing, thanks for the offer but I'm pretty hopeful denise itself is OK now, see above. I also thought the video encoder has nothing to do with RGB output but it's not true. U12 does provide analogue RGB to the 23pin port. The digital RGB signals taken from the buffer outputs are not used by standard monitors, I think they're compatible only with some old style 8 colour type monitors since only the MSB from each colour is used.

[Khyron]

Originally Posted by jimbob

These buffers are 74HCT244. I think I can just replace them with any standard CMOS equivalent. I'm not quite sure if they are inverting or non-inverting buffers though, Anyone?

Non-inverting, the same HCT type/SMD package should be easy to find at pretty much any local electronics shop.. they should only cost about $0.30 to $0.50 each..

[jimbob]

Thanks Khyron, I've ordered some of these -

http://uk.farnell.com/jsp/search/pro...sp?SKU=1103140

I didn't check too carefully they will fit the pads but I have A600s with different sized chip packages in this position and the same pads so I'm not too worried. Should be standard stuff.

Meanwhile I'm gonna lift pin two on that chip and see what happens. Hopefully I will get a proper display with just wonky red colours and a cooler denise.

EDIT: hmm lifting pin 2 of U32 doesn't seem to have made any difference. More is wrong here than just one dodgy buffer input.

[jimbob]

@alt computing, I sent you a PM about your spare denise.

Damn, I reckon Denise is dead after all. I have replaced the buffers U31 and U32, no improvement and it still gets severely hot.

I found a nice testcard program on aminet-

http://aminet.net/package/docs/hard/TestCard

It has a bunch of test screens some of which give very predictable RGB signals at denise outputs. These show up nicely on the scope from working amigas but I just get various weirdness on this broken one. Course since I can't exactly see whats going on I can't be 100% certain I pressed the right keys to get the pattern I wanted. . .

Since the rest of the amiga seems to be fine I'm kinda tempted to plug in my indiECS just to check that all else is well with the input side to denise. But in case something else is wrong I don't feel too great about plugging it in to a potentially denise killing motherboard.

Really the heat should be a dead giveaway that denise is faulty but since it seems to nearly work, I can't help thinking something else is wrong.

[jimbob]

Still looking for a replacement PLCC denise for this A600. If anyone has one, let me know what they want for it.

@Altcomputing, dunno if you didn't get my replies to your PM or I didn't get your response but please get back back in touch if you can still have one to sell.

[cpiac64]

i have a 600 dead, i can remove denise and send you with a agnus defect for test

[jimbob]

Hi cpiac64, I'm not sure what you mean about an agnus chip, but I'll PM you about getting that denise.

[cpiac64]

denise sent 2 days ago

[jimbob]

Received the magic squares today from cpiac64, thankyou very much. Hopefully they will do the trick.

I got an A604 the other day and while swapping it into my main A600, I thought I'd risk trying the indivision in this one, results were mixed.

Attached a photo of the kickstart screen on indivision, looks much better but maybe indicates worse trouble. Any ideas what might cause those vertical stripes?

I loaded workbench from IDE and the workbench screen looks totally normal, only the mouse pointer kinda shudders vertically and won't respond to vertical mouse movement, horizontal is normal. Forgot to check if I could move the pointer with the keyboard before putting the indi back, damn.

[cpiac64]

in the a500 when i plug the indivision, need insert denise on top for use the mouse, in the a600 i dont know

[jimbob]

Originally Posted by cpiac64

in the a500 when i plug the indivision, need insert denise on top for use the mouse, in the a600 i dont know

Hmm, well you don't insert a denise into the Indi for an A600. But I did wonder whether a dead real denise on the motherboard could cause trouble with the mouse. I'm just hoping that the vertical bars don't mean a broken agnus and the dodgy mouse a broken whatever else.

Anyway I'll try the chip you sent me, hopefully tomorrow. Plan is to remove the overheating one, check all it's lines connect where they are supposed to , then solder on a socket and try the replacement.

[jimbob]

Socket installed, I tried the old denise in it to make sure it still behaved the same, it does. New denise is a big improvement but still not all good. The picture now appears kinda properly and the mouse works again so the overheating denise must be dead or severely wounded.

RGB and composite have some vertical bars like I showed above but not stable, move the mouse and the bars change. also some shuddering/flickering of the whole picture. I'll try to post a video later. No picture on RF?

@cpiac64, I think you might be right about agnus. I'll try the one you sent and see if the fault changes to what you had. Need to get a socket for it first.

PS. Soldering experts, I'm pretty new to hot air desoldering, How long at at what temp would you expect to remove a 52 pin plcc. I used the correct nozzle but it seemed to take a very long time to melt the solder. The time it was taking made me nervous and I managed to damage one pad when it finally let go.

[cpiac64]

one pad under denise is (nc not used) and is simply to damage

for remove i used a hot air gun

[jimbob]

Well I got one of those EZcap video capture things today and it works quite nicely. I was gonna use it to show the fault on this A600 but it took so long to arrive that I got impatient and tackled replacing agnus yesterday.

I got the chip off nicely but made a bit off a botch of putting a socket on. After some swearing and a couple of hours I had it sorted and had learned a few things like it is a billion times easier to solder and clean up a socket if you revove the plastic in the center which is for robot placement

So new agnus with known fault supplied by cpiac64 showed its known fault and the previous fault of vertical striping and other picture disturbance caused by mouse movement. So this problem was not caused by agnus.

Knowing that mouse signals go direct to denise, I though maybe the socket I put on wasn't so great. There were no shorts between pins but maybe some conduction when the mouse signals changed was disturbing other signals. So I reworked the denise socket and behold it is fixed. Board appears to be functioning perfectly now

Originally Posted by cpiac64

one pad under denise is (nc not used) and is simply to damage

for remove i used a hot air gun

Yes I think pins 7, 20, 34, and 47 are unused. Pin 19 which is marked as burst on the schematic also appears to be unconnected. Unfortunately it was pin 1 I damaged, managed to make a pretty neat repair though.
BTW you have PM.