Why are there so many problems with Commodore's SMT motherboards?

[CCCP alert]

I find it odd that ALL my Amigas with SMT (surface mount technology) motherboards have problems and yet none of my other SMT based from early/mid 90s consoles have problems and none of my old skool big fat component Amigas like CDTV/A500/A1000 etc have problems despite being significantly older with much higher 'mileage'.

Are these component quality issues or motherboard design issues from C= engineers?

[Daedalus]

Primarily capacitor quality - SMD capacitors of the time were poor quality compared to the equivalent THT parts. Commodore used the cheapest parts that would do the job for the expected lifetime of the machine, whether they were SMD or THT. You occasionally get other failures like cracked ceramics on SMD motherboards, but in my experience they're about as common as cracked traces on an A500 motherboard. Also, there are plenty of other SMD systems that have similar issues - the Gamegear was notorious for capacitor failure even before the Amiga for example, and various PC-Engine consoles are well known to suffer similar failures due to capacitor leakage.

But I've also repaired a number of A500s that had faults due to capacitor failure - most capacitors in the A500 don't give obvious symptoms when they fail, but capacitors in the reset circuit, in the floppy drive and on the keyboard controller board can cause major issues. And I've repaired A500s with failed custom chip sockets - a fault that can't happen on SMD motherboards.

Bottom line: you're just lucky / unlucky - your experience with your collection doesn't really reflect the bigger picture.

[giantclam]

Originally Posted by Daedalus

Primarily capacitor quality - SMD capacitors of the time were poor quality compared to the equivalent THT parts. Commodore used the cheapest parts that would do the job for the expected lifetime of the machine, whether they were SMD or THT. You occasionally get other failures like cracked ceramics on SMD motherboards, but in my experience they're about as common as cracked traces on an A500 motherboard. Also, there are plenty of other SMD systems that have similar issues - the Gamegear was notorious for capacitor failure even before the Amiga for example, and various PC-Engine consoles are well known to suffer similar failures due to capacitor leakage.

But I've also repaired a number of A500s that had faults due to capacitor failure - most capacitors in the A500 don't give obvious symptoms when they fail, but capacitors in the reset circuit, in the floppy drive and on the keyboard controller board can cause major issues. And I've repaired A500s with failed custom chip sockets - a fault that can't happen on SMD motherboards.

Bottom line: you're just lucky / unlucky - your experience with your collection doesn't really reflect the bigger picture.

Yep, that pretty well sums it up ~ to remain competitive in the market, I always figured they were aiming at the 5~7year unit expected lifetime. There would've been some sort of impetus to use SMD caps, to save a bit of space on the component footprint and keep the overall height to a minimum...but end of the day, they did chose parts that lasted for the life of the design ~ it's only us avid humans that keep them alive and notice some things don't last forever =)

[CCCP alert]

I have 40+ Amiga machines of various models so it isn't just luck. I also have 100+ SMT other vintage equipment too that fairs better on percentages. That is enough of a sample of machines not to be just 'bad luck'.

I think then it's probably a combination of unintentional built in obsolescence due to lower resilience of these tiny parts and Commodore going all Tesco Value when choosing components to use seeing as they had to make a profit, unlike console manufacturers who can afford loss making hardware sales.

I would imagine the A1200 motherboard, hell A600 even, is more complex than a SNES motherboard etc so more to go wrong as they say.

Luckily my friend runs an electronics shop so I can get free advice and components ordered, even get work done at mate's rates.

[alexh]

There are other boards from the same era with the same failure rate or higher. Take the Acorn A4 laptop.

[Jope]

Any early 90s SMT Macintosh will also be super leaky.

[Megalomaniac]

My own dislike of the A600 notwithstanding, I know their initial return rate was a lot lower than for the A500, but it does seem like lifespan beyond 5-7 years was something compromised for this - but Commodore knew full well that it'd be obsolete hardware by then, and if your A600 died you'd treat yourself to an A1800 or A2400 or whatever (obviously not factoring in that they'd be history by then). The A600 actually cost more to manufacture than the A500+, despite these corners being cut.

[giantclam]

...and if we moved away from just computers and into the field of other consumer/domestic electronics, the same deal... not to mention the great tantalum capacitor debacle of the mid/late 90's that hit everyone that used them at the time .... now, it's MLC capacitors that transmogrify into fusible resistors for no apparent reason -- it's so easy to hide 'junk parts' when they're so small there's little room for unique identifiers... it was never just a C= thang ; production cost cutting relative to intended product lifetime is as much a thing today, as it was way back when Amigas were being manufactured... ie; if the car has a 5year warranty, you build control modules that last at least that long ...

[daxb]

That is the reason why everything should have lifetime warranty instead of a fixed (fantasy) number. And I don't mean by random. :-)

[Snoopy1234]

The capacitor plague occurred a decade later but maybe something similar happened with the Commodore supply chain...

https://en.wikipedia.org/wiki/Capacitor_plague

[Daedalus]

Originally Posted by CCCP alert

I have 40+ Amiga machines of various models so it isn't just luck. I also have 100+ SMT other vintage equipment too that fairs better on percentages. That is enough of a sample of machines not to be just 'bad luck'.

And I've repaired and serviced more machines than that, so by the very same token I can easily say it is.

I think then it's probably a combination of unintentional built in obsolescence due to lower resilience of these tiny parts and Commodore going all Tesco Value when choosing components to use seeing as they had to make a profit, unlike console manufacturers who can afford loss making hardware sales.

That's what I was saying about the capacitors. Other than the A1000, Commodore used the cheapest parts they could that should last the expected working life of the machine, or about 10 years. The difference is that cheap THT capacitors of the time were better than cheap SMT capacitors of the time.

I would imagine the A1200 motherboard, hell A600 even, is more complex than a SNES motherboard etc so more to go wrong as they say.

Yes, but the A500 motherboard is also more complex than a SNES motherboard so that doesn't really follow. FWIW, SNES boards also suffer from SMT capacitor leakage and the resulting fallout. Complexity doesn't necessarily translate into unreliability.

Out of interest, what issues have you had with all your many A600s and A1200s that wasn't capacitor-related? In particular, what common issues have you seen across multiple machines?

[r.cade]

It was all 90's electronics... and all for the same reason. Surface mount caps and lithium batteries.

[redblade]

New technology takes a while to work out the bugs, i.e different Amiga motherboard revisions (Although cheaper RAM chips would have been the main reason for Amiga mobo revision)