Feedback capacitors for Naim power amplifier.

[diybry]

thanks Pete, good info
Bry

 

[Arkless Electronics]

Having ordered a couple of 100uF wet tants, I now understand that Naim used two 47uf in parallel to reduce the chance of failure.
CAn anybody please advise what the implications of failure are? Totalling my speakers is not a preferred option!
thanks

Tants usually fail short circuit so putting two in parallel will double the chance of failure!
Failure would likely mean a volt or two of DC fed to the speaker. Not enough to harm them.

 

[chrisn]

Just to add my experience a couple of days ago:-

The dc offset has always high on one of my NAPA boards at 180mv but I judged it as not damaging and it sounded OK
I retested it after being reminded by this thread and it was 500mv ie potentially damaging
To me this usually means I have to find someone to repair it not easy

I took the capacitors out and tested them, one of the paralleled pair of 47uf feedback was dead (zero capacitance) surprise

Without much hope I replaced the pair with AVX FFB and hey presto offset 10mv!

 

[diybry]

Tants usually fail short circuit so putting two in parallel will double the chance of failure!
Failure would likely mean a volt or two of DC fed to the speaker. Not enough to harm them.

Interesting- I had not thought that through... Any idea why Naim changed from a single to parallelled caps then?

 

[Arkless Electronics]

Just to add my experience a couple of days ago:-

The dc offset has always high on one of my NAPA boards at 180mv but I judged it as not damaging and it sounded OK
I retested it after being reminded by this thread and it was 500mv ie potentially damaging
To me this usually means I have to find someone to repair it not easy

I took the capacitors out and tested them, one of the paralleled pair of 47uf feedback was dead (zero capacitance) surprise

Without much hope I replaced the pair with AVX FFB and hey presto offset 10mv!

Interesting- I had not thought that through... Any idea why Naim changed from a single to parallelled caps then?

The offset is from the same issue that would result in a Volt or two if they failed short. In this case they are leaky....

Although it is on record that I'm no fan of Naim amps, I can of course repair them and all other makes... So unless you're outside of the UK you've found someone to do repairs

There is no good technical reason, in this context, as to why two have been used in parallel. It may well be down to two smaller ones being cheaper than one big 'un! Especially if they already stock large quantities of the smaller ones.

 

[MJS]

The 100uF were leakier than 2 x 47uF in parallel. And cheaper, yes.

 

[Mike P]

For anyone interested....

I've had a couple of 220uF wet tants on output coupling duties in my CD player for a few weeks now and whilst they were certainly OK I didn't find them anything special. I've just reverted back to a pair of 2.2uF polycarbonate caps which I prefer.

Perhaps some wet tants are better than others? Or what's good for an amplifier feedback position isn't necessarily good for signal path coupling?

 

[BobMaximus]

Or what's good for an amplifier feedback position isn't necessarily good for signal path coupling?

Correct.

Perhaps some wet tants are better than others?

You mean you think they are better. Restated: "Perhaps some wet tants are different from others?". That would be undeniable, Mike!

 

[martinab2]

Although it is on record that I'm no fan of Naim amps, I can of course repair them and all other makes... So unless you're outside of the UK you've found someone to do repairs

Unfortunately as an 'unbeliever' any repairs you do to Naim amps will not sound as good. I can help here - although I don't repair amps (other than my own) I am a Naim fan and can do the incantation and final blessing (including a signed certificate) if you send me the serial number and £49.99.

 

[337alant]

The offset is from the same issue that would result in a Volt or two if they failed short. In this case they are leaky....

Although it is on record that I'm no fan of Naim amps, I can of course repair them and all other makes... So unless you're outside of the UK you've found someone to do repairs

There is no good technical reason, in this context, as to why two have been used in parallel. It may well be down to two smaller ones being cheaper than one big 'un! Especially if they already stock large quantities of the smaller ones.

I use an 82uf wet tantalum and I get approx 35-40mv offset and 0.0 ESR ( measured today) after a couple of years of use which to me is of no concern whatsoever.
For the record I have had solid tantalum bead caps as used by Naim fail short and also give very high ESR but I have never had a mil spec we tant fail short
The requirement of Military spec capacitor is it is manufactured to a high reliability and low failure rate factor, thats why they are very expensive new.

Alan

 

[Arkless Electronics]

I'd not use a tant in audio personally as they are about the worst type. They had a period of popularity around the time Naim gear came out and several companies used them, but they were later shown to be not a good idea and so pretty much everyone but Naim stopped using them...

 

[337alant]

Jez in general I agree I wouldn't use a tantalum bead in an audio signal path but then you are faced with the dilemma what works better?
In this feed back application I have tried many different types of cap , the solid aluminium BC128 is better than the tantalum bead. I tried many types of electrolytic and film types even big polypropylene giving them weeks of trial period but they had their plus and minus with treble mid and bass performance in this particular application but when you put in a wet tantalum it all just comes together and sounds superb so love or hate tants in this application they work and I haven't found anything better ?
Alan

 

[Fatmarley]

The only wet tant I've tried are those funny looking Russian, green ones. Thought they were ok but much prefer the low leakage solid type (cant remember the model number off hand). They beat any electrolytic or film type I've tried in the feedback position of my Onix oa21.