A600/M-Tec A630 instability

[Damion]

I finally got a chance to play around with an A600/M-Tec 630 combo I picked up last year, and I'm finding it has some issues. :/

Basically, the thing suffers from random lock-ups and reboots. All I have to do is fire up the 3.1 "Blanker", and it's only a question of time (a few hours at best, usually a few minutes) before it crashes.

I've tried both an A500 brick, and a 360W ATX PSU.

It's not IDE related, as it'll do it straight from 3.1 floppies.

Although it doesn't get hot anyway, I pointed a huge fan right at the '030, no difference.

I cleaned all the contacts where the card interfaces with the A600 68K, have re-seated it several times, and even gave each pin on the M-Tec's PLCC socket a very gentle nudge outward.  

Thinking the '030 might be stressed (seems most of these are overclocked 33 MHz parts, no 40 MHz QFP '030 exists), I de-soldered the oscillator, installed a nice machine-pin socket and clocked her down to 30 MHz... no change.

I haven't tried an extensive test with the accelerator disconnected... but, it doesn't yet crash without it.

Turning off all cache, no fastrom, etc, "seems" to reduce the frequency of crashes.

I'm thinking my next step should be to replace the caps on the A600 mobo. Any other suggestions from those who have experience with similar hardware?

[giturox]

If the instability occurs after a few hours in electronic devices it usually is somehow related to temperature.

Electrolytic capacitors are prone to aging so they might need replacing. Especially if they have been used heavily.
If the electrolytic caps have lost much of their electrolyte the charge that they can carry is diminished. As they heat up this would become worse.

Read more here...
And even more here...

A more remote possibility is a failure of the voltage regulators for the supply for the 68030. If voltage regulators experiences in its lifetime over voltage (more than +35V) it could fail. I don't think this is the case here, as I believe the caps are problem here.

Best of luck and do tell what you find out...

[mfilos]

I never had any issue with my M-Tec 030@42Mhz on my A600.
It has the necessary fans to be sure that heat doesn't affect the CPU and everything works as intended.
Mobo revision it's not an issue (at least known) as it is with the Apollo 630. My mobo though it's an rev1.5
Try and clean the metal surface of the 68k along with the M-Tec's female socket metal surfaces (in case they don't make a nice and firm connection).
PSU is not an issue since I'm using an A500 brick also and I never had a problem while having HD, CF Adapter and 2xFans internal.
My version is the 4MB which is PCMCIA friendly.

It's weird though... these exact symptoms I'm having in my A1200T with my Apollo 060 - although these must me timing issues with the mobo... Anyway...

I hope you'll manage to fix it cause the M-Tec is a very stable and reliable piece of hardware for the A600.

[a1200]

Being as A600 motherboards are a lot cheaper to replace than the A630, and indeed some A600 accelerators, like the A1200 ones, do have issues with certain motherboards, I would be inclined to replace that first.

In the UK Amiga Retro Experience always sells boards, see:
eBay Link
but I understand NTSC A600s are a little rarer. Even if you have to buy the whole A600 to try it, the results of having a cute 030 powered machine are great see my old A600 here. If you have to get a grubby A600 cheap, then at least you can put the motherboard inside your existing case and you have a spare internal FDD and keyboard. I will put 10 dollars on the motherboard being the problem.

[Damion]

Thanks for the input gents, much appreciated. :pint: I'm glad to hear that the M-Tec is generally a stable unit.

I'm also leaning towards bad caps on the motherboard. At this point I think I'll buy a PAL board (would rather have one for demos anyway), replace the caps on both and keep the NTSC board for a spare. I'll update here when I finally get it sorted.

[marcfrick2112]

-D-: Just my 2 cents, I wouldn't insert/remove the A600's accel. any more than is absolutely necessary, I had heard that the prongs that lock in place over the 68000 have only so much 'life' in them.  ( my dad had a 600 with '030 accel. for some years, can't remember the specifics tho...)

[Damion]

marcfrick2112 wrote:
-D-: Just my 2 cents, I wouldn't insert/remove the A600's accel. any more than is absolutely necessary, I had heard that the prongs that lock in place over the 68000 have only so much 'life' in them.  ( my dad had a 600 with '030 accel. for some years, can't remember the specifics tho...)

Thanks, I'll keep that in mind. Worse case scenario I'll replace the socket, but hopefully it won't come to that. (I get real nervous putting an iron near old, almost irreplaceable Amiga HW.) Anyway, I'm determined to fix it, one way or another. ;-)

[a1200]

When I had my A600 accelerator I got a blunt slotted screwdriver and gently hit the bottom of the 68000's pins to make the pins form a lip for the accellerator to sit on instead of a smooth drop which the socket cannot "clip" onto.

[a1200]

If you do tweak the 68000 and break it by accident, I do have a shed load of new 68010's which are pin compatible with the 68000. As the 68030 would be doing all the work anyway, the 68010 would be as compatible as the 000.

[Damion]

Finally got back to my A600/'030, and am happy to say it's now running 100% perfect. :-)

The caps under the floppy had leaked so badly, I decided not to bother fixing the motherboard. Same with the second board I bought on ebay... some of the pads were simply destroyed.

Third time was the charm. This time a Rev 1 board (woot!) from a UK seller. The caps under the floppy had also leaked, but after removing them and the nearby IC, and cleaning with some baking soda/distilled water, the board was thankfully 100% untouched. Replaced all caps (PITA, as the Rev 1 boards have much smaller smt land sizes than the others), the corroded L347M IC, and the POS Jamicon caps on the M-Tec with some Nichicons. Initially the hard drive light didn't work, which turned out to be a little transistor near the connector for the activity LEDs.

Anyway, I'm so pleased it's working now. :pint:

[fiat1100d]

-D- wrote:

The caps under the floppy had leaked so badly, I decided not to bother fixing the motherboard. Same with the second board I bought on ebay... some of the pads were simply destroyed.

If you are going to trash some A600 board, let me know :-)