Hybrid headphone amp with 12AU7/ECC82 IRF510 and LM317

Rodeodave

Member
Paid Member
2004-12-26 9:38 pm
in the alps
Yesterday evening I built a hybrid headphone amplifier based on a design published over at DIY Audio Projects: NP-100v12: DIY 12AU7 (ECC82) Tube / IRF510 MOSFET Headphone Amplifier.

It's the first headphone amp I've built, and also the first time I've built something with tubes. This may interfere with my perception, but I still think it's a very nice amplifier.
I used what I found in my various junk boxes, hence the odd values. The input caps are Nippon Chemi-Con FNX-HS 1uF 100V foil capacitors from a discarded power supply, in case someone is interested. All other caps are Panasonic FC, the resistors are from the Vishay MRS25 series (copper leads, magnetic endcaps, 1%, 0.6W).

Here's a few numbers I've pulled from the circuit (with no signal present; DMM):

V_supply=12.6V from a benchtop linear PSU (that I rescued from a container)
I_supply=0.38A (reading from display)

Left channel:
V_gate=10.90V
V_source=7.06V
V_cathode=0.063V

Right channel:
V_gate=11.01V
V_source=7.24V
V_cathode=0.059V

The tube's heaters are wires in series, and the center tap measures 6.30V to ground.

Quite interestingly the circuit simulation spit out values surprisingly close to reality:
I_supply=0.32A
V_g=11.05V
V_s=7.31V
V_c=0.059V

At the moment the tube is a Valvo ECC82 that measured the following (Eico 666 with 12AU7 settings from 667 master settings chart): 119%/106%
I'm guessing this imbalance is at least partially responsible for the different L/R voltage levels.

I'm listening through my Sony MDR-EX71SL (Mr. Linkwitz seems to like them: Reference earphones, almost at the bottom), the source is a Lenovo S10e with FLAC and it's ALC269 audio chip (which Ilike). The amp sounds a bit distorted, but not unpleasant. The bass is incredible, the sound as a whole very involving.
The simulation suggests the distortion is mainly 2nd harmonic (surprise...), with the 2nd being 40dB down, and the 3rd 70db down.

Nevertheless a fun, cheap and quick project. Gonna roll me a tube now...
Attached you'll find my actual schematic and some pics.
 

Attachments

  • hybrid_headphone.jpg
    hybrid_headphone.jpg
    90.1 KB · Views: 1,452
  • 04112010158.jpg
    04112010158.jpg
    675.4 KB · Views: 1,376
  • 04112010155.jpg
    04112010155.jpg
    452.4 KB · Views: 1,225
  • 04112010153.jpg
    04112010153.jpg
    600.2 KB · Views: 1,200

Rodeodave

Member
Paid Member
2004-12-26 9:38 pm
in the alps
I went ahead and built me a high-Z version of the modified Linkwitz crossfeed filter (article here: HeadWize - Project: An Acoustic Simulator for Headphone Amplifiers by Chu Moy), something I have wanted to do for a long time now. And I'm glad I did, because turned out quite nicely and it's really sweet.
At first it doesn't do much - I even used the "Chromatic Scale" track from Stereophile's Test CD3 to verify that it actually does do something audible, which it does - but after some time the brain seems to appreciate the crossfeed and eases up. Bliss.
Again i used parts I had laying around. The Wima FKP1 seem to work really well, maybe I'll use them more often in future projects. Wiring the 16 resistors and 10 capacitors was quite a task (I blame the drinking), but at around 3am the circuit was ready and, quite surprisingly, worked.

After some sleep I was being annoyed by having to use the benchtop PSU, and not wanting to mess with mains voltage this morning, I dug out an old wall wart powersupply and noticed something funny. The thing is so old it seemingly is a linear supply and has a big transformer inside. It put out almost 20V (unloaded) of what looked quite like DC, even though it was rated for [email protected] Under load it still exceeded 16V but now had dozens of mV sawtoothy ripple on it.
Copying from the Pearl Two PSU (http://www.passdiy.com/pdf/PEARL 2.pdf) I added a 1R WW resistor, a 2200uF cap, a 7812 regulator bypassed with a 0.47uF foil cap each at in/output and fed the latter via another 1R WW resistor to the circuit.
Works like a charm, and passing ~0.5A the regulator settles at ~50°C. It also (still) sounds excellent.

Very nice project all in all, but the next project will have to be at a higher voltage...
 

Attachments

  • xfeed_amp_ wide_comp.jpg
    xfeed_amp_ wide_comp.jpg
    170.6 KB · Views: 317
  • crossfeed_bottom_comp.jpg
    crossfeed_bottom_comp.jpg
    225.6 KB · Views: 230
  • crossfeed_detail_comp.jpg
    crossfeed_detail_comp.jpg
    147.2 KB · Views: 1,074
  • 12V_reg.jpg
    12V_reg.jpg
    28.1 KB · Views: 475
  • 12V_reg_comp.jpg
    12V_reg_comp.jpg
    140.3 KB · Views: 400
  • xfeed_amp_det_comp.jpg
    xfeed_amp_det_comp.jpg
    177.8 KB · Views: 262