[aaron-xp]

This is probably a very easy question (I've searched, but couldn't find answers).

How can one build a passive pre-amp, such that the RCA outs of the CDP goes into this passive preamp, which is only made of a single potentiometer? Then, I'm thinking of having 2 RCA outs from this "passive preamp".

I intend to recable a pair of headphones (something cheap) to be terminated with 2 male RCA ends. It will then be used together with this "passive preamp". I've tried the headphones straight out of the RCA outputs on the CDP, and found it a little too loud.

SUMMARY:

CDP (RCA out) --> Passive Preamp (RCA out) --> recabled headphones with RCA termination

Thanks for the help. And sorry if it sounds rather confusing. I have no idea how else to state this.

 

[amb]

Passive preamps are usually intended to feed a speaker power amp directly, not headphones. Using a passive preamp with headphones is akin to having an inline volume pot with your headphone cable, except that the typical passive preamp's pot resistance is much higher than optimal for direct use with the headphones. If you reduce the pot value, then it (along with the headphones) will unduly load your CDP's outputs severely. Not to mention that an inline pot in the headphone wire is not good for sound quality.

You should use a proper headphone amp with a volume control to drive your headphones.

 

[NelsonVandal]

If the voltage output is enough, and you want the best possible sound without colortation you should go for a buffer with volume control. The best buffer I can think of is JISBOS that can be found on Headwize.com: http://headwize.com/ubb/showpage.php?fnum=3&tid=6711
There are simpler open loop buffers (that could be battery powered) like the "Sijosae ClassAB", LISAIII or PPA v2 buffers, but they require perfect transistor matching not to cause a high DC-offset, and they're open loop and not a perfect match for low impedance phones.

 

[kvant]

Quote:

Originally Posted by NelsonVandal /img/forum/go_quote.gif ... "Sijosae ClassAB", LISAIII or PPA v2 buffers ... not a perfect match for low impedance phones.

Could you please explain this in more detail? I don't quite see it. Thanks.

 

[NelsonVandal]

What we call buffers are current amplifiers. They don't amplify the voltage = gain 1. They are a link between a low current high impedance output (like line out of a CDP or DAP, or a low current opamp) and speakers/phones (that require a high current low impedance output). Take a look at this schematic, the "classic" Sijosae AB buffer, very simple.

You can add constant current sources (in this case JFET's) like in the "Monofied" version below. This improves the sound.

You can add emitter resistors to the input transistors like used in most of the better "diamond buffers" like LISAIII below. Another tweak in LISA are the constant current diodes from rails to emitters on the output transistors.
http://www.esnips.com/doc/2927d82c-4...SPECIFICATIONS
This is the PPA v2 schematic
http://tangentsoft.net/audio/ppa/amp...schematic8.pdf
Just imagine these amps without input opamps, feeding the signal from the volume control to the input of the buffer.

 

[NelsonVandal]

Quote:

Originally Posted by kvant /img/forum/go_quote.gif Could you please explain this in more detail? I don't quite see it. Thanks.

Like I said there are two problems with these open loop buffers - they have a relative high output impedance that will cause crosstalk especially with low impedance phones, i.e. there will be "leakage" from left to right and right to left, and the sound will become a bit more mono, like a crossfeed http://en.wikipedia.org/wiki/Audio_s...le_performance. There will also be a less than desired damping factor, and there will be less control of the headphone membrane http://en.wikipedia.org/wiki/Damping_factor.

Since they don't use feedback the measured distortion will be very high.

Even with these faults, the sound is very good. Very warm, detailed, intimate, uncolored etc.

The other problem is DC offset that will require good matching of transistors and probably a little bit of resistor adjustments. My LISAIII buffers only measure 1 mV of DC-offset from zero to max volume.

 

[kvant]

I see. However,...

Quote:

Originally Posted by NelsonVandal /img/forum/go_quote.gif a relative high output impedance that will cause crosstalk

This would not be an issue in a two-channel configuration with a real dual-rail power supply.

Quote:

There will also be a less than desired damping factor, and there will be less control of the headphone membrane.

On the other hand, some designs (SOHA, original Millett's hybrid) use output resistors of values quite a bit larger than some 1 or 2 ohms the diamonds might have. Maybe a low damping factor is not such a big deal. Or maybe it is. Then there is the 120 ohm "standard" mentioned e.g. in http://www.head-fi.org/forums/showthread.php?t=216180

Quote:

Since they don't use feedback the measured distortion will be very high.

I guess this depends on personal definition of "very high". With as much class-A bias as people use in PPA it is likely to be rather low.

 

[NelsonVandal]

I think the sound of the above mention buffers beat any opamp, I was just pointing out some theoretical problems. I think everybody should try them sometime. They are cheap and easy to DIY on breadboard.

I still think there is an audible crosstalk problem, both with low and high impedance phones.