Audio output caps (again)

[jimbob]

Hello, I know this has been beaten to death a bit but I'm shortly to be recapping 3 or 4 amigas. Last time I did a couple I just used direct replacement SMT electrolytics not realising there was a a problem with the bipolar output.

Now I want to use more suitable parts but I want to avoid bulky film caps or leaded components. Anyone think of a reason why these new fangled MLCCs would not be a good choice -

http://uk.farnell.com/tdk/c4532x7r1e...f/dp/1843167RL

Apart from the fact they are out of stock, (and the stocked equivalants are ~5 times the price for some reason), these look perfect to me. Should last as good as forever and be easier to solder in behind the ribbon on A600s. Am I right to think that an 1812 SMD package should be a good fit for the pads that are there?

[jimbob]

Ok, no objections. I'll give it a try then. Seemingly these large value ceramics are designed to replace electrolytics and tantalums so I think they will be good for this. I've ordered some of these -

http://uk.rs-online.com/web/p/products/741-6865/

They are a bit pricey but I'm too impatient to wait for the cheaper ones to comes back in stock and they will probably outlast me. They are a bit bigger too, a 2220 package, kinda wide for the pads, maybe they will fit nicely on edge.

[FOL]

Originally Posted by jimbob

Ok, no objections. I'll give it a try then. Seemingly these large value ceramics are designed to replace electrolytics and tantalums so I think they will be good for this. I've ordered some of these -

http://uk.rs-online.com/web/p/products/741-6865/

They are a bit pricey but I'm too impatient to wait for the cheaper ones to comes back in stock and they will probably outlast me. They are a bit bigger too, a 2220 package, kinda wide for the pads, maybe they will fit nicely on edge.

I would stick with SMD electrolytics.
I would'nt say these are ment to replace electrolytics.

Which problem are you referring to on the audio, cause Im yet to see it / hear it.
Audio sounds perfect on every amiga with smd electrolytics.

[jimbob]

I've read in a bunch of threads here about cap replacement where the choice of polarised output coupling caps is thought to have been penny pinching by commodore and decreases the lifespan of those electrolytics due to bipolar audio output of the audio circuits opamp. I don't think it's a problem with sound quality but reliability.

I confess I haven't measured the signal at the relevant places or attempted to analyse the output circuit to see where it idles and if the signal is actually bipolar, just hearsay, but it seems reasonable since the opamp has a bipolar supply.

The data sheets for these MLCCs, (multi layer ceramic capacitors), does say they are intended as alternatives to aluminium and tantalum electrolytics so I don't think it can hurt.

[FOL]

Originally Posted by jimbob

I've read in a bunch of threads here about cap replacement where the choice of polarised output coupling caps is thought to have been penny pinching by commodore and decreases the lifespan of those electrolytics due to bipolar audio output of the audio circuits opamp. I don't think it's a problem with sound quality but reliability.

I confess I haven't measured the signal at the relevant places or attempted to analyse the output circuit to see where it idles and if the signal is actually bipolar, just hearsay, but it seems reasonable since the opamp has a bipolar supply.

The data sheets for these MLCCs, (multi layer ceramic capacitors), does say they are intended as alternatives to aluminium and tantalum electrolytics so I don't think it can hurt.

Fair enough,

But I would watch how you word it as it sounds your saying ceramics can replace normal polarised electrolytics.

[jimbob]

Originally Posted by FOL

Fair enough,

But I would watch how you word it as it sounds your saying ceramics can replace normal polarised electrolytics.

Yes I wouldn't do it without having learnt that this circuit was supposedly not suitable for polarised parts. But I think replacing polarised is exactly the intention of these new devices, the point being that they are able to cram increasing amounts of farads into small ceramics. I always thought that polarisation was an undesired side effect of electrolytic construction. Circuits generally don't require a polarised capacitor but using them = cheaper and smaller with the downside that you must insert them the correct way round and only in appropriate circuits, (ie. unipolar). I think ceramics are typically superior in every way other than size/farad to electros. I could be totally wrong about this and I expect there are at least some exceptions where polarisation is required.

Anyway, I won't be testing my theories in other positions in the amigas, on price if nothing else. Just the audio outs seemed like a reasonable plan. I'll let you know how it turns out.

[FOL]

Originally Posted by jimbob

Just the audio outs seemed like a reasonable plan. I'll let you know how it turns out.

Sounds like a plan, was just worried people that dont know much, miss understand and try and wedge ceramics every where, .

[Stedy]

Hi,

I would not replace all the electrolytics with ceramics, you could make the system unstable. Ceramic capacitors have very low Equivalent Series Resistance (ESR), typically between 0.01 and 0.2 ohms, electrolytics and tantalums have ESRs ranging from 0.05 ohms to 10-15. Too low an ESR can cause higher inrush currents and I have seen it cause issues with power converters. Unless you understand the operation of the circuit, always replace like for like technologies.

The bipolar audio circuit output stage is something I need to look at. With electrolytic capacitors, you can drive them 0.6 to 0.7V (a diode drop) negative, safely, indefinitely. I have seen this on DC clamped video circuits that have been in service for a number of years.

The parts you selected will work perfectly.

Ian

[jimbob]

Well, they fit and they work and I now know they are not necessary.

As you can see from the pic the ones I used are pretty chunky but still easier to solder in that position than the electrolytics. The 1812 package would be a nicer fit and easier still to solder plus a bit cheaper. Like these -

http://uk.rs-online.com/web/p/cerami...layer/7416815/

I will use these again just cause they are easy to work with and probably will never need replaced, but I no longer think that electros are a problem at the audio outputs.

I stuck a multimeter and scope on the caps and there is in fact a DC bias on the opamp outputs. I measured about 2.2V. This place mentions the A4000 output should have DC bias of 2.5V at the opamp-

http://amiga.serveftp.net/audio_repair.html

Using the audiotest utility from that same link, I measured max audio signal at about 3V peak to peak. So there is no problem using a polarised capacitor in this position. Except the fact they are a bit of bitch to desolder and resolder jammed in behind the keyboard connector.

EDIT: looking again at that A4000 audio page, it's audio output is quite different. It has an additional opamp stage. Apparently this has a +1V DC bias and a signal upto 8V p-p. I would say this is a problem for its polarised output caps. Not so on the A600 and I think A1200 also.

[Stedy]

Hi,

Nice soldering job on the Ceramic capacitors, very neat.

A while ago I measured the audio output of the A600, there is no DC bias, that is why there is an AC coupling capacitor. A capture is attached. That was the maximum waveform amplitude I could get. The Marker with '2' shows the ground level of the signal, thus it swings +/- 0.85V.

Have been trying to work out, sensibly if electrolytics can get damaged in an AC coupling config, they should not. The voltage drop across them is a few mV at best, it depend on the impedance and current flows.

Technical details aside, glad you've fixed your Amiga's audio. What games you going to play to test it?

Ian

[jimbob]

Originally Posted by Stedy

Hi,

Nice soldering job on the Ceramic capacitors, very neat.

A while ago I measured the audio output of the A600, there is no DC bias, that is why there is an AC coupling capacitor. A capture is attached. That was the maximum waveform amplitude I could get. The Marker with '2' shows the ground level of the signal, thus it swings +/- 0.85V.

Have been trying to work out, sensibly if electrolytics can get damaged in an AC coupling config, they should not. The voltage drop across them is a few mV at best, it depend on the impedance and current flows.

Technical details aside, glad you've fixed your Amiga's audio. What games you going to play to test it?

Ian

Thanks Stedy, the audio on these A600s was fine just trying these out while doing assorted maintenance on a few A600s. I have mostly been playing Uridium 2 and Desert Strike while testing stuff out .

Yes, DC is not present on the audio outputs. The DC bias is present before the 22uF caps in the circuit. It is of course removed by the cap. This means that the signal never swings the cap voltage below 0V. I'm pretty sure I didn't make a mistake measuring this. On the scope I could see the 1Khz from the the test program. If I switched from AC to DC coupling on the scope input the waveform shifts up ~2.2V.

So the caps never have voltage across them in reverse to their polarity. The only thing I can think of to cause a problem would be exceeding the ripple current rating of the caps by connecting to an amp input with low enough impedance. I don't know how likely this is.