O2 AMP + ODAC | Page 242 | Headphone Reviews and Discussion - Head-Fi.org

Peti · Nov 13, 2014 at 5:18 PM

dark_wizzie said:
HD800 is super easy to power when you place it next to the HE6.

I wasn't talking about powering the HD800 but to find a proper amp (coloration, etc). But I want pure FIDELITY so the O2/ODAC is the perfect choice for me if I ever buy an HD800 again.

Rajikaru · Nov 14, 2014 at 1:57 AM

solrighal said:
I think if I ever do buy another pair of expensive headphones it'll be the HD 800. I'm a bit concerned that they'd be overly bright compared to the HD 650. Reports do seem to suggest that to be the case. I also might need to buy another amp which I'd be loathe to do.

For an amp and DAC released in 2011 and 2012, with the amount of controversy and discussion it is still generating today...

Other amps will come and go, but I'm keeping mine.

Two decades from now, it is either going to so ubiquitous that practically everyone will have one, or it's going to be a collector's item.

Mjolnir125 · Nov 14, 2014 at 4:02 AM

rajikaru said:
This is my Mayflower O2+ODAC. I just put in a couple of hours of listening last night, using it to drive the 90 dB/mW Hifiman HE-560. I was satisfied at the 3-4 o'clock position on the volume knob, and I listen loud.

For exceptionally quiet tracks, the volume may have to be turned up higher.

How does it sound? Like the track being played through the sound signature of the headphone. Bass is present and has impact.

A tip for driving less sensitive headphones (90 dB/mW range): The gain settings have to be optimized.

The common mistake when trying to drive a less sensitive headphone with the O2 is to switch to high gain, and if the source output voltage exceeds a certain amount (see below), and the amp is going to clip and sound harsh or broken.

To get the most output on AC power (which is the only option for an O2+ODAC anyway), 7 divided by the output of your source is the optimum gain. If the number exceeds 7, it's going to clip eventually as the volume gets turned up. For example:

FiiO X5 - 1.5 V output, the optimum gain on the O2 would be 7/1.5=4.6666 which translates to a practical gain of 4 or 4.5.
Most DACs output at 2V, so 7/2 = 3.5 optimal gain.

The default gain settings for 2.5 and 6.5 were made to account for both battery and AC usage with a variety of source input voltages:
2-2.5V desktop DACs (use 2.5 gain)
.5-.7V portable sources (use 6.5 gain)

This is to the best of my understanding. Hopefully JDS or Mayflower will let me know if something isn't right.

Hope this helps!

Yup, the high gain setting is not for low sensitivity headphones, but rather for quiet sources using the line in. Using the high gain with the ODAC will clip if your source is also set to maximum volume and you are listening to a song that is close to the max level. Given that I only need to be at the 3 O'clock position with both my HD600s and K702's I can't imagine what would ever need more than 50% of the ODAC's output (unless you are listening to some incredibly quiet high DR stuff).

headwhacker · Nov 14, 2014 at 4:26 AM

mjolnir125 said:
Yup, the high gain setting is not for low sensitivity headphones, but rather for quiet sources using the line in. Using the high gain with the ODAC will clip if your source is also set to maximum volume and you are listening to a song that is close to the max level. Given that I only need to be at the 3 O'clock position with both my HD600s and K702's I can't imagine what would ever need more than 50% of the ODAC's output (unless you are listening to some incredibly quiet high DR stuff).

The 6.5x gain was there for sources with 1 Vrms (on AC power) or below. 2.5x for anything else up to around 2.5 Vrms. This is to accomodate a wide range of sources. I believe most of the complaint about O2 clipping and not sounding right is due to source/gain mismatch.

I have a DX90 @ 1.7 Vrms and O2 at 2.5x gain drives my T1 very well with the volume knob between 1 to 3 o'clock position.

I have an O2+ODAC coming from JDS Labs with 1x and 3.5x gain. HE-6 is not of my wish list so I got myself covered on the DAC/Amp part for a long time. I hope

castleofargh · Nov 14, 2014 at 10:53 AM

headwhacker said:
mjolnir125 said:

Yup, the high gain setting is not for low sensitivity headphones, but rather for quiet sources using the line in. Using the high gain with the ODAC will clip if your source is also set to maximum volume and you are listening to a song that is close to the max level. Given that I only need to be at the 3 O'clock position with both my HD600s and K702's I can't imagine what would ever need more than 50% of the ODAC's output (unless you are listening to some incredibly quiet high DR stuff).

Click to expand...

The 6.5x gain was there for sources with 1 Vrms (on AC power) or below. 2.5x for anything else up to around 2.5 Vrms. This is to accomodate a wide range of sources. I believe most of the complaint about O2 clipping and not sounding right is due to source/gain mismatch.

I have a DX90 @ 1.7 Vrms and O2 at 2.5x gain drives my T1 very well with the volume knob between 1 to 3 o'clock position.

I have an O2+ODAC coming from JDS Labs with 1x and 3.5x gain. HE-6 is not of my wish list so I got myself covered on the DAC/Amp part for a long time. I hope

not a surprise really. more than half the people I talk to that are already into audio, so more than half of a very precise population target, have no clue about how to use gain settings and go with the usual moaaaarrrrr is better idea that leads to really messed up things. like clipping when it shouldn't, or people crying over the very huge channel imbalance while using sensitive IEMs on high gain or simply the fact that the volume level is harder to set...
I'm very sure a lot of the dissatisfied users with the O2 were themselves the cause of the problem. that and those who expect an amp to sound "different". those guys will never like anything transparent obviously.

I think that having the X1 gain is a good start for 90% of headphones/IEMs. then 2.5 or 6times the gain should be picked in accordance to our harder to drive headphone.
that way I see very little reason to be dissatisfied.

headwhacker · Nov 14, 2014 at 11:03 AM

castleofargh said:
not a surprise really. more than half the people I talk to that are already into audio, so more than half of a very precise population target, have no clue about how to use gain settings and go with the usual moaaaarrrrr is better idea that leads to really messed up things. like clipping when it shouldn't, or people crying over the very huge channel imbalance while using sensitive IEMs on high gain or simply the fact that the volume level is harder to set...
I'm very sure a lot of the dissatisfied users with the O2 were themselves the cause of the problem. that and those who expect an amp to sound "different". those guys will never like anything transparent obviously.

I think that having the X1 gain is a good start for 90% of headphones/IEMs. then 2.5 or 6times the gain should be picked in accordance to our harder to drive headphone.
that way I see very little reason to be dissatisfied.

It's a shame really. A lot of people insist that high gain always sound better even for sensitive iems. Then complain if they hear a hiss.

jseaber · Nov 14, 2014 at 12:28 PM

rajikaru said:
To get the most output on AC power (which is the only option for an O2+ODAC anyway), 7 divided by the output of your source is the optimum gain. If the number exceeds 7, it's going to clip eventually as the volume gets turned up. For example:

FiiO X5 - 1.5 V output, the optimum gain on the O2 would be 7/1.5=4.6666 which translates to a practical gain of 4 or 4.5.
Most DACs output at 2V, so 7/2 = 3.5 optimal gain.

The default gain settings for 2.5 and 6.5 were made to account for both battery and AC usage with a variety of source input voltages:
2-2.5V desktop DACs (use 2.5 gain)
.5-.7V portable sources (use 6.5 gain)

This is to the best of my understanding. Hopefully JDS or Mayflower will let me know if something isn't right.

Hope this helps!

Best summary I've seen in a long time.

Default gain of our custom O2+ODACs is 1.0/3.5x, which is optimal as long as the O2+ODAC is driven by its own DAC. It makes sense to increase gain only if you plan to listen from from a weaker external source, or if the majority of your music is recorded well below 0dB.

adydula · Nov 14, 2014 at 4:03 PM

Yup that makes perfect sense...

Alex

bada bing · Nov 15, 2014 at 12:04 AM

rajikaru said:
The common mistake when trying to drive a less sensitive headphone with the O2 is to switch to high gain, and if the source output voltage exceeds a certain amount (see below), and the amp is going to clip and sound harsh or broken.

To get the most output on AC power (which is the only option for an O2+ODAC anyway), 7 divided by the output of your source is the optimum gain. If the number exceeds 7, it's going to clip eventually as the volume gets turned up. For example:

FiiO X5 - 1.5 V output, the optimum gain on the O2 would be 7/1.5=4.6666 which translates to a practical gain of 4 or 4.5.
Most DACs output at 2V, so 7/2 = 3.5 optimal gain.

The default gain settings for 2.5 and 6.5 were made to account for both battery and AC usage with a variety of source input voltages:
2-2.5V desktop DACs (use 2.5 gain)
.5-.7V portable sources (use 6.5 gain)

This is to the best of my understanding. Hopefully JDS or Mayflower will let me know if something isn't right.

Hope this helps!

jseaber said:
Best summary I've seen in a long time.

Default gain of our custom O2+ODACs is 1.0/3.5x, which is optimal as long as the O2+ODAC is driven by its own DAC. It makes sense to increase gain only if you plan to listen from from a weaker external source, or if the majority of your music is recorded well below 0dB.

Actually, this is a bit confusing. The O2 has very unusual clipping behavior and clips *only* based on the level of the input. The O2's volume control has no effect at all on clipping behavior. The 7v / the output voltage level of the source is a useful rule of thumb, but brings up a consequence of the clipping behavior - the maximum power output is produced without clipping at only one input level. You can not adjust the O2 to maximum power output below clipping by using the O2's volume pot. This is different that just about every other amp on the market. Not important unless the discussion is about "hard to drive" headphones. The O2 has only a moderate output on paper, but in practical real world use it is even less. Recorded material doesn't stick very well to the 0 db standard recording level, which makes it impossible to set the gain at a level that allows the O2 to produce it's maximum rated unclipped output power with a typical random selection of recorded material. A workaround would be to use digital software volume control in conjunction with the gain settings, but that's somewhat problematic.

All this isn't important until the discussion turns to "hard to drive" headphones, but the discussion seems to go there somewhat often. The O2's design objectives were "numbers first" and as a consequence the full rated power of the amp is less usable in the real world. Understanding the unusual design trade offs that were made to generate the last bit of measurement benefit is necessary for max power usage.

Rajikaru · Nov 15, 2014 at 1:22 AM

jseaber said:
Best summary I've seen in a long time.

Default gain of our custom O2+ODACs is 1.0/3.5x, which is optimal as long as the O2+ODAC is driven by its own DAC. It makes sense to increase gain only if you plan to listen from from a weaker external source, or if the majority of your music is recorded well below 0dB.

Thanks!

bada bing said:
Actually, this is a bit confusing. The O2 has very unusual clipping behavior and clips *only* based on the level of the input. The O2's volume control has no effect at all on clipping behavior.

Understood. Edited my original post.

HotIce · Nov 15, 2014 at 12:42 PM

I wonder if anyone ever tried to replicate the measures posted online in terms of power output.
One thing which struck me, is how an amp rated for 630+mW on 33 ohms, which would require over 130mA of RMS (assuming power figures in the most noted web sites are RMS) current, would work with the original NJM4556AD current stage opamp, which is rated at 70mA peak.

jseaber · Nov 15, 2014 at 12:55 PM

hotice said:
I wonder if anyone ever tried to replicate the measures posted online in terms of power output.
One thing which struck me, is how an amp rated for 630+mW on 33 ohms, which would require over 130mA of RMS (assuming power figures in the most noted web sites are RMS) current, would work with the original NJM4556AD current stage opamp, which is rated at 70mA peak.

We've replicated every test and verified each specification by published NwAvGuy on Objective2 and ODAC. All measurements are legitimate, if not slightly conservative.

Objective2 uses two, parallel NJM4556 ICs per channel. Available current handling is doubled to 2*70 = 140mA.

HotIce · Nov 15, 2014 at 3:30 PM

jseaber said:
We've replicated every test and verified each specification by published NwAvGuy on Objective2 and ODAC. All measurements are legitimate, if not slightly conservative.

Objective2 uses two, parallel NJM4556 ICs per channel. Available current handling is doubled to 2*70 = 140mA.

I see, thanks. I finally been able to find a schematic which is bigger than 200x200 pixels

So it is 630mW peak on 33 ohms, right?

diamondears · Nov 15, 2014 at 8:04 PM

jseaber said:
We've replicated every test and verified each specification by published NwAvGuy on Objective2 and ODAC. All measurements are legitimate, if not slightly conservative.

Objective2 uses two, parallel NJM4556 ICs per channel. Available current handling is doubled to 2*70 = 140mA.

Hello there John.

Is there a plan to make a much more portable O2? Is that even possible with the license and present design?

Also, is there an existing portable amp out there whose design is based or similar to the O2 and ODAC?

Cheers.

jseaber · Nov 15, 2014 at 10:16 PM

hotice said:
I see, thanks. I finally been able to find a schematic which is bigger than 200x200 pixels

So it is 630mW peak on 33 ohms, right?

O2 produces 613mW @ 33 ohms.

diamondears said:
Hello there John.

Is there a plan to make a much more portable O2? Is that even possible with the license and present design?

Also, is there an existing portable amp out there whose design is based or similar to the O2 and ODAC?

Cheers.

Modifying O2 breaks the license. Even disregarding the license, it makes no engineering sense to modify O2. A portable (or desktop) amp should be designed from the ground up with desired features and performance in mind. It's quite possible to design an amp with equal or greater performance than O2. We've done so several times in the past two years (sorry, cannot share R&D!).

Battery capacity is the greatest limitation when designing a portable amp. The market has shifted dramatically since 2011, and I think there's a strange obsession now with max output power. Sure, it's important to adequately power headphones, but most portable users have high sensitivity sets which are easily driven by an amp with much less power than O2:

[SPL Chart Data]

We could design a portable amp that blows away O2 in terms of max output power, but I think doing so is unnecessary. Run-time for 1W of output power would be incredibly short, even with a large Li-Ion battery; run time would be fine when driving efficient headphones and/or listening at moderate volumes. Again, portable users typically do not need more so much power. Driving truly demanding headphones necessitates AC power.

Latest Thread Images

Peti

500+ Head-Fier

Rajikaru

100+ Head-Fier

Mjolnir125

100+ Head-Fier

headwhacker

Headphoneus Supremus

castleofargh

Sound Science Forum Moderator

headwhacker

Headphoneus Supremus

jseaber

Member of the Trade: JDS Labs

adydula

Headphoneus Supremus

bada bing

100+ Head-Fier

Rajikaru

100+ Head-Fier

HotIce

100+ Head-Fier

jseaber

Member of the Trade: JDS Labs

HotIce

100+ Head-Fier

diamondears

Headphoneus Supremus

jseaber

Member of the Trade: JDS Labs

Users who are viewing this thread