Have I bought a bad A1200

scoca

Member
Joined
Apr 12, 2013
Posts
68
Country
England
Region
Lincolnshire
My A1200 keeps re-booting itself.
It could be in the middle of a game, in workbench, any time really.

Sometimes the screen goes blue, sometimes green.

Is it a re-cap? How can I tell if it's already re-capped?
 
First step is to check the PSU and see if the voltages are correct.

Are you using a lightweight PSU with any extra peripherals plugged in such as a CD drive?
 
Does it have a memory expansion or an accelerator card? The connector they use are notorious for poor connections causing reboots and guru's...
 
I've just looked and it says it's an A500 power supply, I've no way to check the voltages, I'll have to borrow a multimeter.

There's an 8MB expansion card from AmiKit and a PCMCIA CF card.
 
On a cold boot, do you see any other colours on the screen apart from shades of grey and finally white?

When an unexpected reset occurs, is it always the same 2 colours (blue or green) that appear?
 
No shades of grey at all. Black screen then workbench.

- - - Updated - - -

Using it now and it locked up and screen went blue.
 
Take out the 8mb card and boot the Amiga again.

A blue screen usually denotes a custom chip problem with either Denise, Paula or Agnus

A green screen could indicate a RAM problem or a dodgy CIA chip / Agnus
 
Last edited:
The other thing you could try is to re-seat the ROM chips maybe a crap connection there. The ROMS are the 2 socketed chips on the board.

A good picture or two of a cluster of capacitors on the motherboard, and someone maybe able to tell if it has had a re-cap.
 
I'll try booting without the ram card later today.

I've photographed the motherboard so maybe someone could tell if it's re-capped.

I appreciate all this help IMG_20200723_074107__01__01_compress9.jpgIMG_20200723_074031_compress64.jpgIMG_20200723_074006_compress49.jpgIMG_20200723_074017_compress17.jpg
 
Last edited:
It doesn't look like your board has been recapped ... here's a picture of a recapped board done by retropassion.co.uk using hybrid polymer capacitors (their premium service), with timing fixes and the modulator also removed.



I am not affiliated with retropassion, just a happy customer of theirs :)
 
I'll try booting without the ram card later today.

I've photographed the motherboard so maybe someone could tell if it's re-capped.

I appreciate all this help View attachment 159143View attachment 159145View attachment 159146View attachment 159144

These are still original capacitors, and the good news is there is no leak yet or it's minimal as there is no visible (what I can see on your images) corrosion. Defunct capacitors won't cause reboot problems usually. However many other faults do, could be cracked solder joints on the PLCC chips. Green screen for example is chip RAM fault usually so could be cracked joints on those or bad RAM chips. But could be even a software issue.

Before you start shopping for recap try to diagnose the problem first as contrary to popular belief relatively few problems are caused directly by none functioning capacitors.

I recommend stripping the machine down (no expansions or peripherals) boot Amiga Advanced Analyzer and run a continuous RAM test for a few hours and see if there is a freeze or reboot during. You can get the ADF here:
https://archive.org/details/Advanced_Amiga_Analyzer_v1.11_1992-12-30_Wilcom_Australia

If you still experience reboots with this minimum setup you can also get hold of a DIAG ROM by Chucky (for about 10€ or so) which could also give you some more info about possible hw. faults. Then you can take it from there depending on what you find out. (Sending it for more diagnosis and repair, recap etc..)
 
Last edited:
I've let the analyser run for a while and no errors, lock-ups or re-boots.
IMG_20200724_113012__01__01_compress83.jpg

- - - Updated - - -

Running it again now with the fast ram installed.

- - - Updated - - -

Fast RAM ok.
IMG_20200724_121931__01_compress27.jpg
 
Of all the advice so far, tbtorro has given you the most accurate and technically sensible.

Strip it down, run it bare bones, test it for a loooooooooong time. Then add one component, and repeat test.

The testing you have done so far is OK, but test it for 300 passes, not 14. Give it a 24 hour test with nothing plugged in, and if it passes, then add the RAM card. Test again for another 24 hours.

If that works. Then boot workbench from a floppy disk and leave the hard drive disconnected. Again, let it sit idle for 24 hours.

You can start to see the logic behind this method, I hope?

Long tests. Slowly add components. Assume everything is the cause :)
 
I didn't realise it needed such a long test, but no problem, I'll run it again with no ram card, hard disk or compact flash for at least 24 hours.

Thanks again.
 
Last edited:
I didn't realise it needed such a long test, but no problem, I'll run it again with no ram card, hard disk or compact flash for at least 24 hours.

Thanks again.

No probs.

The long tests are to test components once theyre up to operating temp. Why? As things warm up their properties change. Solder joints move/flex. Chips may not be able to operate internally as they should. And so on and so on.

Assume everything is the cause, and slowly eliminate it. Assume the motherboard has a problem, eliminate it by running it alone for 12+ hours. If it works, assume the memory expansion is the cause, run it for 12 hours. Assume heat is the problem, by running a test for 12 hours your allowing heat to build up.

Here's a bonus: if it crashes during a 12 hour test, have a pedistal fan blowing air onto the motherboard (or ram expansion) to cool it. You are assuming heat is the causing and eliminating it as a possible cause.

Thats how we diagnose. Bit by bit. Step by step. Assume everything is the problem and prove to yourself its not (don't guess, prove it)
 
The testing is ongoing.
I tested with chip ram only - PASSED.
Tested with chip and fast - PASSED.
I plugged in the compact flash hd, loaded the test program and it crashed.

So, using the bit by bit method mentioned above I changed 1 thing, the ribbon cable from the motherboard to the CF adapter.

So far so good.

IMG_20200725_101347__01_compress92.jpg
IMG_20200726_100131__01_compress9.jpg
IMG_20200726_100509__01_compress24.jpg
 
So the memory expansion card is giving some grief. But it might or might not be the chips on the card.

The card looks new, so it might be the card, or it might be the connector. My next suggestion would be to try cleaning the edge connector on the motherboard, and give it a good clean!! The other thing to try is pulling the card back ever so slightly off the edge connector.

So you could try cleaning the edge connector, or mounting it a little out from its "perfect seating position", or you could try keeping the card disconnected/unplugged and now start testing the motherboard bare bones with some software.... try loading Workbench and see if it will sit idle overnight. Load a game and see if it will sit at the title screen overnight, etc.

- - - Updated - - -

Just to clarify for me, if you remove the hard drive and run the ram test again, will it give errors? or is it only when the hard drive is plugged in??
 
Maybe ask somebody to check your card in another A1200 and see if the results you are getting can be reproduced.

I know for fact that Amigastore.eu had received a batch bad memory chips earlier this year, I wonder if the manufacturer of your card got their memory chips from the same supplier.
 
Back
Top Bottom