To all people recapping Amigas

[8bitbubsy]

There's no need to use non-polarised capacitors for the audio output part, because the audio is never swinging between negative/positive before it enters the caps. In fact, it's always positive. The caps themselves do the 'DC centering' so that the output swings between negative and positive. Remove the caps and hook up an oscilloscope and you will be amazed. I've ran normal 22µF 25v tantalums in the audio part of my Amigas for a long time.

Another thing:
Do not use cheramic caps in the audio circuitry. Cheramic caps can pick up vibrations, kind of like a mic, thus they are unwanted for audio applications. You'd not want to couple unwanted things into your audio, do you? Electrolytics or tantalums are perfect.

So unless someone proves me elsewise, I can prove them this: The "Commodore did a capacitor mistake" talk is wrong.

[ShK]

Old thread: http://eab.abime.net/showthread.php?t=67240

[8bitbubsy]

Originally Posted by ShK

Old thread: http://eab.abime.net/showthread.php?t=67240

I didn't know of this when I made that thread. And if you measure from the capacitors, you already get the negative/positive swinging, so it's easy to think that it did swing before the caps too (which it doesn't).

[kipper2k]

what is the minimum voltage rated tantalum that needs to be used, 10v, 16v ?. would it ever peak at 25v?

Also, is +ve towards the Amp ?

[Michael]

Just wondering, when should someone recap the board ?
Is there any signs that show that time has come?
Should a perfectly working board be recapped just in case?
Could a failed cap break something more important and difficult to replace?

[Jope]

If the caps have leaked, it's time to recap the board.

The biggest reason to do it is that the electrolyte in the caps corrodes the traces and vias, which will lead to the eventual destruction of the motherboard. Corroded traces and vias are consuming and tedious to repair.

Many but not all surface mount Amigas will see the caps leak. It's everyone's personal choice whether to do it pre-emptively or not.

Some meaningless statistics: I have four A1200 boards here with original caps, none have leaked. Two A4000s with caps that have leaked. Three A600s with leaked caps.

[Michael]

I guess I better keep my main A1200 board as is for now.
Only when things really fail (visually leaking caps) or something else starts to behave abnormally.

One thing bothers me however. I wonder if it is related to the two caps near the keyboard connector. I have a piggy back board on top of the chip that drives a PS2 keyboard and recently I have got a problem with the keyboard swallowing first 5 keys pressed after soft reboot. It works fine otherwise, and on power on it works immediately too.

Any familiar syndromes ?

[supaduper]

In all honesty anyone with a Amiga needs to recap them regardless of caps leaking or not and its better to replace them when they have not leaked because the electrolyte will not have done any damage to the pads on the board, like I keep saying please replace your caps and use Tantalum capacitors and the worry of leaking will be a thing of the past

The A600s and CD32 seem to be the worst for this problem so don't hang around just get these boards recapped !

And regarding the non polar caps I now replace those with a couple of non polar ceramic caps and you can definitely hear a difference using these as opposed to polarized ones

[Stedy]

@kipper2k

A 10V or 16V part will be ample, the audio output is only 2V peak to peak. +Ve to the amplifier output.

@thread.

Scope trace of Amiga audio output attached.

Take a look at this:

http://conradhoffman.com/capchecktut.htm

The pro-audio guys have more experience than us over many decades.

I've replaced 2 audio caps in my A600, The two incorrectly fitted ones in a CD32 and had one A500 fail for TBC reasons, the reset circuit among them. The main reason for the A600 was lack of use. Electrolytics need reforming charges if not used.

I'm an electronic engineer, one of my guiding rules is "If it ain't broke, don't fix it", the other is that Murphy's law prevails above Ohm's and Kirchoff's laws

Originally Posted by 8bitbubsy

There's no need to use non-polarised capacitors for the audio output part, because the audio is never swinging between negative/positive before it enters the caps. In fact, it's always positive. The caps themselves do the 'DC centering' so that the output swings between negative and positive. Remove the caps and hook up an oscilloscope and you will be amazed. I've ran normal 22µF 25v tantalums in the audio part of my Amigas for a long time.

Another thing:
Do not use cheramic caps in the audio circuitry. Cheramic caps can pick up vibrations, kind of like a mic, thus they are unwanted for audio applications. You'd not want to couple unwanted things into your audio, do you? Electrolytics or tantalums are perfect.

So unless someone proves me elsewise, I can prove them this: The "Commodore did a capacitor mistake" talk is wrong.

Good to know, I've just finished De-capping my A4000 and was considering non-pol or ceramics for the replacements. I'll go ahead and use the ones supplied by AmigaKit.

The caps in the audio section had leaked, so it was high time that they got replaced. The others around the board looked good, but it's hard to stop once I've got my hot air gun warmed up.

[Supamax]

Originally Posted by rootboy

Good to know, I've just finished De-capping my A4000

Sorry for the noob question: why the yellow tape?
Is it some sort of protecting mask from the air gun heat? If so, how?

[supaduper]

Originally Posted by Supamax

Sorry for the noob question: why the yellow tape?
Is it some sort of protecting mask from the air gun heat? If so, how?

Yes its Kapton tape, very heat resistant, never leave home without it !

[xArtx]

It appears none ever bothers with the electrolytic in the Amiga’s linear power supply
which are the ones that live the hardest life of the lot of them.
If they go near short it will heat up/fry the transformer, if they deform,
you’ll get hum through the rest of the computer and wonder why it has issues when you thought you installed new capacitors.

[Thorham]

Originally Posted by xArtx

It appears none ever bothers with the electrolytic in the Amiga’s linear power supply
which are the ones that live the hardest life of the lot of them.
If they go near short it will heat up/fry the transformer, if they deform,
you’ll get hum through the rest of the computer and wonder why it has issues when you thought you installed new capacitors.

Don't bother with old PSUs and buy a new ATX PSU (a GOOD one). Our Amigas deserve it

[FOL]

Originally Posted by Stedy

@kipper2k
I'm an electronic engineer, one of my guiding rules is "If it ain't broke, don't fix it", the other is that Murphy's law prevails above Ohm's and Kirchoff's laws

Very true, very true.

[Thorham]

Originally Posted by FOL

Very true, very true.

Yes, but what if you know that it's going to break?

[Jope]

You can only guess, not know. :-)

[Stedy]

@Thorham

Given enough time and the correct environment, any single component in a system will fail. The probability of it happening depends on numerous factors.

With any complex PCB, you need to weigh up the risks of damage during re-work, especially if the person doing it is not experienced enough. Hence, the term, if it ain't broke, don't fix it.

If you use your Amiga at least every 2 years for a few hours, and keep the operating environment below 50C, the capacitors and the rest of the Amiga should easily last 40 years.

Ian

[ShK]

I know that ceramic capacitors are not recommended, especially in the audio use.

How bad is it? Is there any difference on ceramic capacitors with the audio?

[demolition]

Originally Posted by ShK

I know that ceramic capacitors are not recommended, especially in the audio use.

How bad is it? Is there any difference on ceramic capacitors with the audio?

The problem with high capacity ceramics is that the capacity changes significantly with the charge, which makes it nonlinear. That being said, the audio from the Amiga is not exactly high quality and already quite distorted from Paula, so I'm not convinced the difference is audible. Also, the charge in the AC coupling caps in the output almost always has a constant charge, so any nonlinearities should not transfer significantly to the audio.

So go ahead and use ceramics or electrolytic caps, whatever you have available.