Amiga 1200 crashing and freezing

[orange]

Please help me. For years I had this problem with A1200:
When I turn it on, if its not used for 15 min, everything works fine (after). But, if I try to start games, DOpus, anything else within 15 min of power up, Amiga freezes and crashes. It seems that its more serious when 4Mb RAM card is attached (and I really need it). When I disconnect 2,5"HDD and 4Mb RAM, it almost works perfectly (I can't remember if it crashes at all then)


There is someone that had similar problems with expanded Amiga 1200:

http://groups.google.com/groups?selm=CC9q3z.2qC%40dcs.ed.ac.uk&output=gplain

The only problem is that I have 2,5" HDD and connecting to PC power supply (200W) didn't solve anything. Still, I get the same crashes in first 15 min of use. After that "warm up" time, everything works fine. I could leave A1200 on 24/7 but..
Someone told me it could be the bad capacitor.
What is wrong with my A1200 in your opinion?

[blobrana]

Hum,
The simplest solution is that it's a power issue...
(try a 500 power supply or convert a pc one)
Though it could be a software conflict...a program on the HD...Try running a minimum system...

And a worst case scenario, it's a virus or a dodgy memory module on the accelerator. (Virus scan, mem tester)

good luck.


------------------------
Half-dead cat found in box--RSCPA seeks Austrian scientist

[Framiga]

Hi

IMO it is a problem of lacks of contact somewhere.

When the system is warm, contacts usually are dilated and the contact become OK . . . now the problem is, where?

Is it a plain A1200 without an accelerator?

Have you checked:

socketed Kickstart ROMs

system oscillators (don't know if A1200 has socketed or not)

PSU plugs and so.

The worst thing, should be a cold/broken soldering that become conductive, when the mobo is warm.

Bye

[blobrana]

@Framiga
Your powers are weak, my friend.

you have much to learn.





(er, though i could be wrong) :-)

[Hyperspeed]

Sometimes I get the feeling that something needs warming up on my
Amiga. It usually takes 5 minutes before my display stops flickering
and juddering - I was wondering if a capacitor in my Scandoubler needs
replacing.

If I leave it for a whole day the juddering will take that much longer
to `warm up' and sort itself out.

Can someone give instructions on how to test capacitors (aside from
the usual check for leakage and bulging) and maybe a test we can
carry out on all the components of the motherboard.

Another thing that worries me greatly is the amount of oxidisation on
my Phase5 Blizzard 1260. The Gold contacts are literally matt-yellow!
I scraped the SCSI expansion connector to make it gold but this
surely wouldn't give as clean a contact as something to clean off the
oxidised material.

Would a cotton bud (Q-Tip?) dipped in brass polish do the trick? I've
heard of using isopropyl alcohol but is this any use as a contact
cleaner, does it actually de-oxidise in a way metal polish does?

I think it'd be good for an expert to help start a testing and
reliability thread!

:-D :-D :-D

[weirdami]

@orange

Mine would crash if it got too hot. Do you have something in your trapdoor? Maybe try propping it up to get air flow underneath.
 

[blobrana]

Ahh,
The power is a pc replacement....


Hum,
 perhaps Framiga is right then...
Yea it could be a contact issue, with the heat improving a dodgy contact somewhere...
And if it's a dodgy capacitor then replacement is a real pain...
Not a task to be taken on lightly
Heres a link to test them...

[orange]

Its actually connected to A500 PSU now, but it makes no difference. For some strange reason it worked fine yesterday, no crashes. Though it was a HOT day. There are so many solder points at motherboard, I can't check them all.
Is there any "usual" reason for freezing?
Rarely it crashes with blue/white stripes across the screen.

I had done a very stupid thing with A1200 once: connected 2,5" HDD to wrong ROW of pins. Could that have damaged 68020 (or something else) somehow ?

[orange]

I did some additional tests, here are the results:

When startup-sequence contains only memtest program that tests RAM "forever", and I don't use mouse or keyboard, everything is OK.
Then I put CLI command "list ALL" into startup-sequence, so that HDD works "forever". It worked fine, too, no freezing.

The obvious conclusion would be that something is wrong with chip controlling mouse and keyboard (was it CIA?). It somehow freezes A1200 if not heated.
What is the number label of CIA in A1200, so I could try to resolder it?

[blobrana]

Hum,
Well there are two CIA's in the A1200...and yeah, they`re soldered into the mobo so a real pain to replace...

Check these sites or this one for detailed pictures

I would just check the `legs` to make sure that they arn`t broken (you can use electronic paint to bridge any crack)

[orange]

I have found this post (from 1996)and it seems just like mine problem :

i have a problem with my a1200: when run in dblpal alice heats up and causes screen
corruption which makes the machine unuseable. If alice gets hot enuff the problem goes away -
or if she stays cool, but unfortunately at my normal room temp the damn chip hits just the
right temp to be annoying. other than cooling or insulating (dodgy) the chip, is there
anything i can do to solve this (some sort of h/w hack or summin)  - this is not just a one
off either - all new a1200 motherboards (from amitech) have this problem (i used to work for a
company who sold 1200's and i tested about 14-15 over different periods of time inc different
batches so i beleive it's common anyway)

:-o

[blobrana]

Hum,


Wow, never heard of that before.

So a small heat sink

And thermal-paste may cure it...


huh?

[Merc]

That sounds like the problem where if there were any colours onscreen that were full-white, ie 255,255,255, the gfx would get corrupted after a few minutes -- my brother's 1200 had this problem, but mine didn't.

We "fixed" it by writing a little program that watched the current screen and changed any colours that were full white to 254,254,254 instead, and that, believe it or not, fixed the problem!  :-D

[Ilwrath]

I have found this post (from 1996)and it seems just like mine problem :

[snipped Alice heat/flash description]

Only the Alice corrupted video output bug doesn't crash the computer.  In fact, it rarely even corrupts the viewable chip ram!  It just causes all sorts of visual effects to be displayed.  If you try to snapshot the screen while it's showing this effect, the screenshot looks normal.

It starts with vertical bars of white flashing lines to appear starting from the bottom of the screen, and working their way up.  Eventually it looks like you're trying to play your Amiga from inside a flashing jail cell!  

Also flickering "strings" start forming to the right of certain objects in the display.  Those look like HAM mode bleeding, except they flicker, as well.  (These are more likely to cause corruption in screen memory [in other words sometimes they show up in screenshots, and sometimes they'll make an actual change to a picture you're working on in DPaint, etc])

There are certain A1200 revisions that are more prone to this.  Every 1D3 A1200 I've used exhibits this problem to a certain extent.  It does seem to be heat related, though the Alice doesn't get extraordinarily hot.  

Anyhow, while this is a widespread problem, I don't think it's YOUR problem, orange.

That sounds like the problem where if there were any colours onscreen that were full-white, ie 255,255,255, the gfx would get corrupted after a few minutes -- my brother's 1200 had this problem, but mine didn't.

Now THAT'S one I've never heard of!  Quite an odd solution, too!  What little I know about the real electronics of how computers work would lead me to believe that is an artifact of one of the 'timing errors' that are discussed for the 1D3,1D4,2B motherboards.

[Merc]

Now THAT'S one I've never heard of!  Quite an odd solution, too!  What little I know about the real electronics of how computers work would lead me to believe that is an artifact of one of the 'timing errors' that are discussed for the 1D3,1D4,2B motherboards.

Well your description of the problem (lines working their way up the screen) seems similar to what would happen to his.. and it only happened on DBLNTSC/DBLPAL/other 31khz modes.  Why dropping the intensity down a bit helped, I don't know, but I think one of us read somewhere (Aminet probably) that this would fix it.