ZEN-like headphones amp

I've built a scaled-down version of the Zen amp to feed my Grado headphones.
The HexFets employed are the usual IRF610, both for the active side and for the current source.
The bias current I chose is 300mA, the power supply voltage about 30Vdc, for the maximum compatibility with Hi-Z headphones.
I've obtained best results taking advantage of Nelson's patent 5,710,522, this allows to halve the distortion figure and to increase the drive caapability: on 32 Ohms load I've measured over 8Vp-p without any apparent compression, as soon as possble I'll perform a standard measuring set.
The sound is beautiful, more natural (notwithstanding Grados..) than the RA-1, which is not that champion people believe, at least in my opinion...
The input impedance can be set between 13KOhm ( freq. respnse as far as 80KHz) and 40KOhm ( upper -3dB limit 26KHz).
If someone would desire the schematic, I'll send it to him gladly.
Hi... I'm very interested in headphones and headphone amp for almost a year now... I have build the szekeres from headwize.com in five times now.. active Iscource.. parralel resistors, non inductive resistors, MKT bypass, low Iq, high Iq ..tried them all.. and guest what >>>> The most simple version sounds best...

anyways..... You do not need any voltage gain for a headphone, so why build a ZEN type? The szekeres from headwize.com build with five parallel metal film 100ohm resistors, irf610, Iq at 200mA, good MKT bypass , GRADO headphone, sounds fantastic...

I've also designed a SOZ headphone amp which is nearly compleet..but since the szekeres sounds so good I lost interest and the project stalled...I'm now desinging a Aleph version.. that could drive a loudspeaker too....

anyways...I would be vey thankfull if you would send me your design..

tschrama, what do you use as source for your headamp?
is there really no need at all for voltage gain?
i wanna use it for my pc.
i don´t think the output from a soundcard is as high as lineout from a cd-player for example.
thought about the variant with an opamp in front of the fet-stage.
(BTW: the best sounding headamp i built so far is an OPA627 buffer by BUF634.see BUF634 datasheet)

thanks in anticipation


2001-06-01 4:53 pm
tortello, I'd like to see your schematic. Can you email it or post it somewhere? Thanks.
I've just completed an amp with a Szekeres-type headphone amp and am quite pleased with it, but I have to chase a small hum problem, similar to what JoeBob has described in another forum. The amp includes a op-amp in front with a gain of (I think) about 3.
i know this is off topic but how do you distribute your audio signals to the headphone amp etc.
for example a power amp with one stereo-in that has an additional headphone amplifier and a vu-meter.
i have finished all of these boards and wanna combine them in one case.
any recommendations for opamps and schematics for distribution would be cool.(only found one application in the TL074 datasheet)


2001-06-01 4:53 pm
My new box includes a balanced line input stage (INA2137), then runs to a power amp (The ESP project 3a) and a headphone amp (the Szekeres). There is a separate volume control for each of these two amps, but I don't do anything special for audio distribution, just run them in parallel. There are independent power supplies for the power amp and the line input & headphone amp, but they are in a separate enclosure.
If the input impedances of your three circuits are high enough, you can just connect them together. You may, however, want to buffer them through an op-amp or other circuit. The OPA627 has a good reputation, but make sure it's stable at unity gain before you wire it up that way.
Here is the schematic of my headphone amplifier, a simple circuit based on the Zen idea, with the exploitation of the Pass Labs patent # 5,710,522.
This increases a lot some circuit features, such as distortion and driving capability:thanks again for your open-mind, Nelson!
With the schematic values I obtain over 8Vpp onto a 22Ohm load (!), with about 1% of T.H.D., but what really matters is that with my Grado 325 I obtain a very good sound, far better than with the RA-1.
The gain is 3.2, or about 10dB, the input impedance about 40KOhm, without the volume pot.

Fine settings:
R4 should be set to obtain the requested drain bias current: I chose 300mA;
R15 should be set to obtain about 1/2 B + on TP1;
R8 should be set to obtain i (R10) = i (R12) in dynamic conditions, for example
with a 1KHz sinewave at the input.

The pass [Pass?!:)] -band is about 4Hz-25KHz: if you want to extend this, you should change R9 into 68K, R14 into 10K and R15 into 56K, the input impedance decreases to about 10KHz: the values must be adjusted as above.
Feel free to change values, bias current and P.S. voltage.

I like its sound very much, I'll be very glad if you will build it and tell me what do you think, thanks to all for this!

P.S. The image is not so good, I know...: if anyone wants a better .pdf file, please let me know.


  • tortello's_amp.jpg
    52.4 KB · Views: 15,954
HI all..

my set-up:
a Philips cd723 cd-player , cheap but ok.. modified the ouput stage by a opa2132 .. this gives me so much voltage out (2V rms??) that I use a standart 10K+1K voltage devider infront of the headphone amp.. it is still too loud (grado sr-60).. now I also use a 24bit/196KHz crystal evaluation board for the digital out .. this goes to my second headphone amp .. it has a 12step atenuader... beginning with 10k+1K and then downwards.. I gues a 10K+220Ohm would be a good start...

so you see I need NO voltage gain .. I attenuate by 20dB or more...

A third headphone amp I use as refeence.. some how it always sounds best.. all headphone amps are szekeres .. 2 earlier prototype are dismantled..

I also have two buf634 and two buf04 around.. gonna try them too ...
Dear Paulb

Hum problem....:

the original szekeres does not use a decoupling cap for the MOSFETs bias circuit.. that could very well reduce the hum......

also 200mA DC with 2mA AC though a 0.1 Ohm ground lead causes enough hum to be audible so carefull gronding is nesscery..

you could also put the MOSFET inside you opamp feedback loop... an opa134 could be stable at a gain of 3 .. or use a compensation cap around the feedback loop to give it a 100KHz roll all .. should be more then enough... This is what I want to try if I had just a litle more time for things. :-(

Thanks, tschrama, I've been pretty darn careful with my grounds but I may have messed something up. The hum is too low to see on my scope, but it's audible at zero volume setting. I haven't had a chance to look more closely at it yet, too busy listening to the power amp section.
Just a quick update on my hum problem.
I found that it was the ripple coming in to the cap multiplier that caused the problem. While the multiplier did a very effective job of removing the ripple, the input to it was close enough to the headphone amp board (2 inches or so) that it induced ripple into the amp.
The power supply is in a separate enclosure, so I added a simple RC (4R7, 1000 uF) there. Voila! The ripple was cut down by about a factor of 10, and it softened the edges a bit. The hum is now inaudible.