The Objective2 (O2) Headphone Amp DIY Project

My O2 has passed testing and if officially finished! Thanks to RS, Jokener, olli1324 and to everyone who helped with advices.

O2 works great, except 23V AC supply is really too much for it, voltage regulators got very hot very fast. But it's not regulators I'm worried for, it is C4 and, especially, C5 which are boiled by the regs and are rated at only 85C. Perhaps, I'll need to switch those for 105C caps?
 
My O2 has passed testing and if officially finished! Thanks to RS, Jokener, olli1324 and to everyone who helped with advices.

O2 works great, except 23V AC supply is really too much for it, voltage regulators got very hot very fast. But it's not regulators I'm worried for, it is C4 and, especially, C5 which are boiled by the regs and are rated at only 85C. Perhaps, I'll need to switch those for 105C caps?

that's too high.. try 18V output AC supply
 
I'd like to, but 23V is the best I could find.

And the very first listening inpression: no audible difference compared to Sony CDP-XE510's headphone output. Either my headphones or my ears are probably not good enough to tell the difference. These are the headphones, if anyone's wondering: http://www.german-maestro-shop.de/e.../Shops/61902279/Products/"GMP 8.35 D Monitor"

And the second impression: for Red Book source the gain is too high. I'm currently using recommended gain, which, I believe, is 2.5x, and it is about right only when I engage line level regulator of my CDP, dropping the level by -20dB. Looks like I need gain < 1X.

P. S. O2 really is very easy to assemble even for a noob like me (not without some help from the community, I admit). And I only got 4 spare parts left, can you believe that? :)
 
Last edited:
got mine all finished and tested a few days ago and have been enjoying the extra juice desperately needed with fostex planars. thanks for all your hard work in developing and offering the design to for the diyers. the testing steps are particularly good and the board layout is well laid out for easy soldering. It was quick and the info was clear. a suggestions for future versions is some consideration for rf interference. i have to position all mobile devices over a meter away to mitigate the issue
 
It was quick and the info was clear. a suggestions for future versions is some consideration for rf interference. i have to position all mobile devices over a meter away to mitigate the issue

Just curious - are you using the B2-080 case or other metal box, and if so did you run a ground wire from the case to the ground pin of the input jack? Also see if the interference sounds worse when you move the phone near the input cable from the source, the output cable to the headphones, or the power suspply input cable. Try moving the phone near the edges of the front and back panels and see if that makes it worse. Are those two panels metal? Also do you know what type of radio service the phone uses, like GSM? Or alternatively just the model of phone and which carrier.

RocketScientist has an input RF filter in the circuit, R7/R3 and C11/C12, but the RF may be finding another way in.

Thanks! :)
 
@Alexium & Idjoel2000, I would suggest the resistor modification a few pages back to let the regulators run a bit cooler. The regulators are rated for 35 volts input and they should be comfortably under that even with the high transformer voltage and no resistors. I would go with the 15 ohm resistors--possibly even 18 ohm if you don't expect to every work the amp hard into a low impedance headphones.

The case temp of the regs should still be under 85C (even though the internal die temp might get that high under some conditions) and the caps far under 85C. So I wouldn't worry about the caps unless you plan to leave the amp on, and working hard, 24 hours a day 7 days a week.

It sounds like you have very sensitive headphones so that means the O2 will not be working hard at all. Removing the batteries also lets the regulators run cooler. 1X gain is easy, you can just clip, or heat and lift, one end of one pair of the gain resistors. Or if you have sockets, just pull them out. If you never plan to buy any voltage hungry headphones, and have a Redbook source, I would clip the 274 ohm resistors so you're left with 1X and 2.5X gain.

In terms of sound quality the O2 really shines with either headphones that require a low source impedance (such as balanced armature IEMs or certain full size headphones that need more resonance control) or with headphones that many sources simply cannot drive to suitable peak levels without current and/or voltage clipping or other heavy distortion. With typical dynamic sensitive headphones your old source might have been doing an good enough job.

@algfrimur, agdr is correct, the B2-080 case with metal panels should make your O2 fairly immune to cellular interference (RFI). I have to get my phone within a few inches (perhaps 10 cm) to hear some noises but it will depend on the phone, wiring, headphones, etc. If you still have problems, you could try making the 274 ohm input resistors 470 ohm or even 1K ohm and/or adding external clamp on ferrite suppressors to the input cable at the O2 end of the cable. As a last resort, in addition to those mods, a clamp on ferrite on the headphone cable near the plug should solve the problem.

But, overall, the O2 in its full metal case should be less prone to RF issues compared to many headphone amps--especially all the ones without an input filter or poor PCB layouts with excessive loop area.
 
It sounds like you have very sensitive headphones
Not at all! With the gain of 2.5 I have more or less the same impression (sound pressure level-wise) of GMP 8.35D, old 30 bucks-worth Koss R/80 and $12 Koss KSC75. I believe this is usual sensitivity, maybe slightly above average, but not outstanding. I just don't need high SPL for in-doors use.
Thank you again for the advice, I'll perform the resistor mod. I'd like to add that the regulators are heating at idle, I didn't feel them getting any more hot when listening to the music, probably due to low power output needed.

P. S. It is so great you have designed PCB with switchable gain! I didn't really expect any audible improvement over XE510 (to my unskilled ear), because I have compared it to $200 (and well regarded) external soundcard EMU 0404, and found it's DAC+HPA indistinguishable from my XE510's DAC+HDA (and the owner of 0404, who has a more sharp ear, backed my impression up). The main for building O2 is to use it with DAC I'm planning to build, and the thing is that DAC will probably have output level as low as 0.5V, and having gain switch allows me to accomodate O2 for both my CDP and DAC. Even better, I plan to experiment with different DAC output stages, and I can build two different circuits and accomodate O2 for their different output level, which should allow me to switch between them faster so I can compare them more closely!
 
Last edited:
I'm finding that the suggested knobs are quite difficult to get on the pot shaft. I don't know if I'm just not pushing hard enough but I feel like the pot might dislodge from the vias if I push too hard.

Maybe some heat softens them... I have the same problem.

Brgds

BTW: I can confirm that the new BOM with new values for R9 and R25 works for another brand of batteries as well - Fisher 270mAh NiMH Accu. This completes my testing of the former issue.
 
Last edited:
Yeah, the knobs can be a tight fit. Heating, and lightly "lubricating" the knob by tossing it in a glass of hot soapy water might help. Or, better still, slide the board out of the case (with the front panel) and put your thumb behind the body of the pot and with the finger of that same hand press the knob on. That takes the pressure off the pins and PCB.

@Alexium, the 0404 has a really poor headphone output but you may know that already. I have the 0404 and will be publishing some measurements of its DAC as part of my series coming up on USB pro audio interfaces.

If you mean you plan to try blind level matched switching that is a very good idea. You can make a 60 hz sine wave test track with Audacity (free) and use a cheap DMM to read the AC voltage playing back the track. That allows match levels accurately between different signal paths for fair A/B switching. If you let someone else do the switching so you don't know which is which you might be surprised when testing output stages ;)
 
the 0404 has a really poor headphone output
No, It never occured to us (me and 0404's owner). Perhaps, it's only poor with some tough phones?

If you let someone else do the switching so you don't know which is which you might be surprised when testing output stages ;)
Yep, I meant exactly that :) I'm a bit sceptical about my ability to tell the difference, though. Cheap Sony XE 510 is the top of sound quality inprovement I can hear, I'm afraid. I do, however, hear a difference between different headphones (I've heard a few in $100-$200 range), and I'm not quite happy with the current ones, so having a decent amp will be great when I'll find some better phones to test.
 
@regal, the board is designed to slip into a few different versions of the Box Enclosures cases. But there are also 3 mounting holes provided. The holes, and their precise coordinates, are shown in the O2 documentation PDF. They're also visible in several of the pictures. 3 holes are plenty to secure the board with stand-offs in any application.

@alexium, the 0404 (and every other pro interface I've tested so far including the O202, M-Audio, Tascam, Focusrite, A.R.T., Roland and Behringer) headphone output has a relatively high impedance. That causes all sorts of problems with many headphones. See: Headphone Amplifier Impedance

Benchmark Media also just published a new paper that provides additional objective evidence why higher output impedances are bad: Sonic Advantages Of Low Impedance Amps
 
Last edited:
@regal, the board is designed to slip into a few different versions of the Box Enclosures cases. But there are also 3 mounting holes provided. The holes, and their precise coordinates, are shown in the O2 documentation PDF. They're also visible in several of the pictures. 3 holes are plenty to secure the board with stand-offs in any application.
[/URL]

Thanks I see those are standoff holes, don't know how I missed that. Thanks
 
Benchmark Media also just published a new paper that provides additional objective evidence why higher output impedances are bad: Sonic Advantages Of Low Impedance Amps

I just read that. It's informative. I was speculating about the part where they mentioned that the HD650's performed better (less distortion) than the MDR-V6's even when the damping factor was the same. I suspect that's because the Sennheisers are requiring more voltage, likely bringing the opamps voltage output into the sweetspot for lowest distortion (pushing the signal closer to the knee).

I've been wondering how to go about trying to get the O2 to feed my headphones at the ideal voltage for best performance of the opamps without making my ears go deaf. It's just for an excercise in achievement, really. Probably too off topic for this forum.

Good read. Thanks, RS.