AC coupling on headphones amps?

G'day all, this touches on an earlier thread of mine, dealing with dual op amps (OPA2134). I make routine use of one of my DIY 'Silicon Chip' design stereo headphone amplifiers.

Recently one of them developed a circuit fault (a faulty LED) along with another 'annoying' problem that has happened several times previously, that of destroyed headphones caused by high level DC transients (heard as a loud 'click'), caused by momentarily interruptions on the split rail power supply.

I've recently hopefully solved the problem by installing 100 uf 50 v working bipolar electrolytic capacitors on the outputs, thereby breaking the 'destructive' DC path. Everything appears to work as well as previously with no apparent quality degradation. So I ask the question, why isn't AC (capacitor) output coupling routinely used thereby prevently 'accidental' stereo headphone destruction? Regards, Felix.
Last edited:
G'day mate, no the circuit is fine, an OPA2134 driving an BD139/ BD140 output stage per channel. Here is the circuit. Regards, Felix.


  • img063.jpg
    72.5 KB · Views: 403


2006-09-08 2:04 pm
I have no metaphysical qualms about using blocking capacitors, but 100µ seems a bit thin.

This puts the corner frequency at 50Hz, which seems definitely highish.

I would settle for 1000µ at least, and if you know the polarity, you can use a standard cap of even higher capacity without problems.

Some people on this forum will probably challenge this opinion, but I advise you make the decision for yourself, after proper evaluation.
G'day mate, yes I'm aware of the 50 Hz corner frequency but since that amp gets used mostly as a monitor amp for recording work, that corner frequency is fine.

Anything that stops precious (and expensive) stereo headphones from being destroyed is ok by me. When you destroy an expensive set of Sennheisers in a 'click', that's not a good feeling! Regards, Felix.
AC coupling

Most single ended and Darlington voltage amplifier type headphone amplifiers typically use a 1000 UF series capacitor. But the Silicon Chip design is a small Class A amplifier, and remember it is originally used for speakers that could handle it. You may need a silent jack or delay relay like many professional and high quality systems. That is just the way it is, and many lower end (Peavey) pro components snap and pop if not used with care.


2002-06-06 4:43 am
Electrolytic caps in the signal path work better if they have a dc bias on them. I'm never happy even with 'non-polar' electrolytics as they are just two back to back elco's.

So headphone amps with single supplies and using an electrolytic capacitor on the output is a better idea I think. They are biased at 1/2 the supply voltage. The single supply JLH headphone amp particularly is a very good headphone amp and compares very favourably with the no output capacitor +/- supply version. It also doesn't pop during switch on and switch off.

With no capacitor at the output you will need a mV Dc sensor and relay to protect the phones at some extra expense. The Electrolytic cap however is cheaper and more "reliable" with only one component doing the job.
Under fault conditions, one of the output devices can drive about 400mA into a 32R headphone, not quite enough to blow the 500mA fuse, but more than 5W dissipation.

I never understood why headphone amplifiers are "one size fits all" for 32R to 600R types.
A more realistic +/- 8V supply would limit damage.