[JacobLee89]

  There's a new build of the ODA over at DIYAUDIO if anyone is interested:
 
http://www.diyaudio.com/forums/headphone-systems/229934-version-o2-desktop-amp-oda-74.html
 
You can follow this persons build to see if you think you might want to tackle the project!
 
Alex

 
I'll need to save up a bit, grab a whole spool of solder wick, and a PCB holder or some sort.
Then I'll tackle this!

 

[HotIce]

I am not sure if anyone tried this, but it should be possible to replace U5 and U6 with with MC7815 and MC7915 respectively, and change R1 and R2 to 390 Ohm, to have a 6V P2P boost.
The 20V AC supply should be good, or eventually bump it to 22..24V AC.
The NJM4556 can take up to +-18V, so 15V should be safe.
I would not use it in that configuration to feed high demanding low Z HPs (as the Pd for it tops 800mW for the SIP8 version), but for HPs with higher Z and hard to drive, similarly to the "there is no replacement for displacement", there is no replacement for voltage

 

[adydula]

Have u ever looked at AGDR's O2 booster board?
 
Alex

 

[HotIce]

Reading right now.
What I was proposing is much less involved, and requires $2 of parts and dis-soldering two easy TO-220 ICs, and two resistors.

 

[mcandmar]

Low impedance phones need current not voltage.   The mods you have listed are fairly easy to do but there is a gotcha to do with the current handling limits of the IC's.  I believe this fundamental issue was what started the O2 Booster project in the first place.

 

[HotIce]

Which is exactly what I stated (FWIW, Z == Impedance):

"I would not use it in that configuration to feed high demanding low Z HPs (as the Pd for it tops 800mW for the SIP8 version), but for HPs with higher Z and hard to drive"

Ie, a 600 Ohm HP, at about 13V peak, would ask for about 20mA, or about 10mA per buffer.
A 300 Ohm, about 20mA per buffer, well within the IC specs.

 

[castleofargh]

the list of headphones needing 13V is rather short I would guess. there are already not so many headphones that will need more than a basic O2's voltage to reach some 115 or 120db. and I would guess that those limited by the O2 are mainly so because of an overall power limit like orthos where sensitivity and impedance are both rather low.

 

[adydula]

Hey I took the liberty to talk to the O2 Booster design engineer and here are his comments for your read:
 
Interesting!  You are right, that exact issue is covered on the 4th to the last page in the O2 Booster Board build instructions here

https://drive.google.com/folderview?id=0B67cJELZW-i8ejFUWGFNLWdIWVU&usp=sharing

under the "O2 Headphone Amplifier +/-15Vdc Power Supply Upgrade" heading.

 In addition to replacing the regulators there are a couple of more parts that have to change (for voltage rating).  And the big thing - whether he leaves the NJM4556A's in or replaces them with the Booster Board - no less than 600 ohm headphones if using the NJM455As or no less than 300 ohm if using the Booster Board. 

The issue here is the amount of power dissipated by the output chips, as that poster notes.  Even though both chips (NJM4556A or the LME49600 buffers on the booster board) are rated for +/-15Vdc rails, the math still has to be done tp make sure the chips won't overheat with a given load at that voltage level.   The higher the chip supply voltage the more chip power dissipation for a given ohm load and given output voltage swing.

Remember that even with the original O2 +/-12Vdc rails there is a problem with the NJM4556A's overheating with the combination of a 16 ohm load, low sensitivity headphones, higher volume levels, and music with high sustained musical peaks.
 
Fyi... an thanks to AGDR.
 
Alex

 

[diamondears]

Is NWAVGUY interested in EVEN improving this design? Or its potential maximized already?

 

[PDVJAM]

  Is NWAVGUY interested in EVEN improving this design? Or its potential maximized already?

Well, no one knows who is he and where he is now

 

[PDVJAM]

I also have a question. I'm planning to buy o2+odac for the following set of headphones: CIEM with 16ohm and sensitivity 111db, hd650, audeze lcd-2 and hifiman he-6. Would it be enough to have gain settings 1x (for CIEM)/3x (for all other) in this case?

 

[headwhacker]

  I also have a question. I'm planning to buy o2+odac for the following set of headphones: CIEM with 16ohm and sensitivity 111db, hd650, audeze lcd-2 and hifiman he-6. Would it be enough to have gain settings 1x (for CIEM)/3x (for all other) in this case?

 
HE-6 will be too much for O2 to handle. HE-560 is borderline maximimum O2 can drive. At 1x O2+ODAC may be too load for the CIEM. But you can reduce the volume on your PC to compensate.
 
HD-650 and LCD-2 should be easy for O2 at 3X. JDS Labs allows an option for 1X/3.5X gain.

 

[Thad-E-Ginathom]

  Is NWAVGUY interested in EVEN improving this design? Or its potential maximized already?

 
NwAvGuy doesn't "exist" any longer.

 

[upstateguy]

  are people moving to something else because it's an upgrade? or simply because they realized a little late that  transparent wasn't their idea of good sound?
<snip>

 
Agree with you completely.
 
It's very hard to obtain the level of resolution you get from an O2 with a tube rig.
 
What you will get with a tube rig is various amounts of coloration, which is why people roll tubes.
 
Another approach might be to start with the highest resolution you can get, and then if needed, color to taste with EQ.

 

[adydula]

now thats a great idea!
 
Alex