O2: High Performance Portable/Home Amp

[khaos974]

Originally Posted by Satellite_6

I like the idea of a pure desktop amp. NwAvGuy has already said that the original version would be OK as a transportable amp. Do you plan to use headphones that are harder to drive than 300 ohm Senns on the go? If not you should be good!

Anyway:

Nope, the hardest thing I have to drive is an LCD-2, the "always more headroom" philosophy is just an interesting design exercise , to what extent can one push power while keeping superb performance across the board?

My personal opinion is that if getting more headroom without harming performance in other areas is possible, it should be done even at the price of needing more expensive components. After all, people in need of performance for cheap already have the O2 and I think no matter how nwavguy splurges with expensive parts, it will remain reasonable compared the the price of high end headphones.

It's also a convenience isssue, I have a few CDs where the max volume levels is at -20 dBFS with an average at -40 dBFS (what the heck was the mastering engineer thinking ), if I had such a track with the O2, I'd have to switch to the high gain mode to get enough volume.

However, this is not "the Wire" bis, it's O2 desktop edition, I don't know how much tweaking the O2 design allows for.

[CrowdAds]

[khaos974]

Originally Posted by Limp

You could always use a more regular portable amplifier (like the O2) when at the café, and/or maybe use some headphones that do not require an unhealthy amount of power.
Often it is easier to adapt your own requirements to the facts of the world, than to demand the facts of the world to adapt to your requirements.
/zen

Of course I don't NEED an uber powerful amp but pride of possession is still one reason why someone buys/builds something. And "can drive anything, in the most inimaginable conditions" is one heck of a motivation for a selling argument. But it's an easy argument when no measure is provided, but when everything is provided like for nwavguy's projects, it's a lot harder to back up.

[Satellite_6]

Originally Posted by khaos974

Of course I don't NEED an uber powerful amp but pride of possession is still one reason why someone buys/builds something. And "can drive anything, in the most inimaginable conditions" is one heck of a motivation for a selling argument. But it's an easy argument when no measure is provided, but when everything is provided like for nwavguy's projects, it's a lot harder to back up.

Hmmm. . . It might shut up the people on Head-Fi, who at the moment are basically saying "it's no good for every ridiculous situation I can possibly think of, therefore it sucks." Then again, they'll probably whine about non-existent issues no matter what.

I should really be asleep. . .

I wish I had LCD-2's. I'm jelly.

[inarc]

It's also telling how they attempt to shift the blame from their nonstandard sources to the O2.

[mikeaj]

Well let's be realistic. If you already own some sources, certainly you would want an amp that plays nice with all of them. Regardless of which is "standard" (and there's not really a standard), they may already own some hot DACs, and they don't yet own an O2. Of course the O2 does better with regards to input clipping than a whole bunch of portable amps at the default gain levels, and it will have no problems after a small adjustment to 1X or other gain.

So the argument goes that not all users can be bothered to figure out if their sources are too hot, and adjust the O2 gain? You'd expect any DIYer to figure it out for themselves, given all the warnings and explanations in the articles. Maybe the amp detractors are conceding that the O2 may become popular with others.

However, I kind of agree from a usability standpoint that in general you shouldn't allow the possibility for devices to be run improperly through user error. The fringe 1% source case is maybe like 10% among audiophiles? Who knows. Then again, what with the babying that is required with tube amps, the required caution by the user with many amps and headphones not to blow the drivers or their eardrums by setting the volume too high, and amps that damage headphones by turning on or off, avoiding potential input clipping in fringe cases doesn't sound so onerous.

[xnor]

Originally Posted by mikeaj

Regardless of which is "standard" (and there's not really a standard), they may already own some hot DACs, and they don't yet own an O2.

There are common nominal line levels, -10 dBV for consumer and +4 dBu for pro (US) audio which are 0.316 and 1.228 Vrms respectively.

[NwAvGuy]

Lots to cover already... here are some answers/comments...

Originally Posted by khaos974

Would it be possible to design an amp with superlative performance on AC and the same level of performance as the original O2 on battery? <snip>

Anyway, I think that "high end" gear should be designed to provide more headroom and be usable in more extreme cases that don't often happen in normal situation (such as driving those apogee speakers whose impedance went down to 1 ohm).

There are lots of trade offs in audio design. In this case, as I said, power consumption and the rail voltage are fairly significant. If I know the rail voltage is always say +/-12 volts I can optimize the gain structure so the input will never overload. But if it's sometimes going to be +/- 8 volts on low batteries that's a huge difference in terms of voltage swing. So now I need to seriously compromise that optimal gain structure to avoid input stage overload when the amp runs from battery. And if I just put a caution in the documentation my critics will whine that's too confusing.

To the second point, my designs are the opposite of cost-is-little-object high-end. The whole idea is to offer 95% of high-end performance for less money than others are charging for their half-baked entry level amps. And that means cutting out the useless fat. It's the needless excess that drives up the price. I try to spend money on things that actually matter not things that only will make some feel better about the design. Many are not going to agree with that approach, preferring overkill instead, and that's their choice. They can spend more but they won't get better sound.

Originally Posted by khaos974

Nope, the hardest thing I have to drive is an LCD-2, the "always more headroom" philosophy is just an interesting design exercise , to what extent can one push power while keeping superb performance across the board?

My personal opinion is that if getting more headroom without harming performance in other areas is possible, it should be done even at the price of needing more expensive components. After all, people in need of performance for cheap already have the O2 and I think no matter how nwavguy splurges with expensive parts, it will remain reasonable compared the the price of high end headphones.

It's also a convenience isssue, I have a few CDs where the max volume levels is at -20 dBFS with an average at -40 dBFS (what the heck was the mastering engineer thinking ), if I had such a track with the O2, I'd have to switch to the high gain mode to get enough volume.

However, this is not "the Wire" bis, it's O2 desktop edition, I don't know how much tweaking the O2 design allows for.

It is possible to design in more power/headroom/etc. to a point without compromising anything but the size/price/power consumption/etc. But why? I came up with some worst case headphones, and short of the K1000 which aren't really headphones, nobody has disputed or come up with anything worse. More power just means more risk of someone accidentally damaging their expensive heapdhones. The O2 can flawlessly drive your LCD-2s while running on battery. More headroom/power wouldn't help the sound.

The quiet recording/excess gain issue is a classic problem with any headphone amp. At least the O2 has a gain switch making it much less of an issue. Personally I think it's worse only using the first third of the volume control's range with 95% of my music just to have an extra 12 dB of gain in reserve for the 5% of quiet stuff. Using a PC/laptop as the source you can add the extra gain in the digital domain. You can also normalize poorly recorded/ripped recordings, use playback gain, etc. Excess gain in an amp degrades performance, increases the risk of headphone damage, creates channel balance problems, make volume adjustments "touchy", etc. I realize not everyone will agree, but a gain switch and/or the digital solutions mention above, seem like the best compromise.

The "how much tweaking" issue is partly how much time I want to sink into the desktop amp. The further it deviates from the O2, the more work it's going to be and the longer it will take. You also have to be realistic about what matters. The O2's performance is already well past the point of diminishing returns. It has a few areas that can be improved--like the gain structure.

The idea is to spend my time improving things that will make a difference and not give into all the usual snake oil mythology. I admit I could add several useless things that would make the desktop amp more popular with the subjective crowd. But doing so just adds to the price, and in the end, hurts everyone by further draining their wallets for stuff that won't make it sound or work any better. I described my philosophy as "Harmless Excess" in the 2nd article. If it's easy, harmless, and cheap to include some audiophile excess I consider it.

Originally Posted by mikeaj

So the argument goes that not all users can be bothered to figure out if their sources are too hot, and adjust the O2 gain? You'd expect any DIYer to figure it out for themselves, given all the warnings and explanations in the articles. Maybe the amp detractors are conceding that the O2 may become popular with others.

However, I kind of agree from a usability standpoint that in general you shouldn't allow the possibility for devices to be run improperly through user error. The fringe 1% source case is maybe like 10% among audiophiles? Who knows. Then again, what with the babying that is required with tube amps, the required caution by the user with many amps and headphones not to blow the drivers or their eardrums by setting the volume too high, and amps that damage headphones by turning on or off, avoiding potential input clipping in fringe cases doesn't sound so onerous.

Thanks Mike. Very well put. Everything is a trade off. The Mini3 uses just one battery to be smaller but that requires a virtual ground which creates lots of problems and limits the output to barely enough to properly drive HD600s. Headphones like the AKG K701 and LCD-2 are off the list. The Mini3 also has audible noise with my IEMs because of where the volume control is in the circuit while the O2 does not. So I made different trade offs than AMB and, personally, I think they're worth it. The O2 is also about half the price.

And yeah, the boutique headphone amp market is full of designs that require their users to be aware of all sorts of things. The Mini3 can blow up if it's output is even briefly shorted and hates low impedance loads. Some tube amps produce scary amounts of voltage. Traditional 2 battery Cmoys require constant vigilance on the battery level to avoid headphone damaging DC.

It's worth noting even FiiO's desktop amp, the E9, overloads at 2.1 V RMS of input (0.6% THD). There's a graph in my review. It can't handle the HRT MS II let alone the Audio-GD DAC my critics are waving around. The $100 FiiO E7 input overloads at only 1.2 volts. So I thought I was on solid ground with the O2 as I don't see anyone bashing the E9 for input clipping at only 2.1 volts.

Ultimately this is how the breed is improved. My critics are already well out on a limb with the O2. The idea with the desktop amp is to take their tree away completely when it comes to objective criticisms.

[NwAvGuy]

Originally Posted by Satellite_6

Can we call the desktop version the:

Anti-H34d-F1 Amp

I assume that would avoid any copyright issues?

I'm open to suggestions... but I'm leaning towards "ODA" - aka the Objective Desktop Amp. It's short, easy to type, pronounceable, and nothing resembling a headphone amp comes up doing a Google search. It also keeps the name in the "O Family"

[khaos974]

Concerning classical music, revising the gain structure to allow for no input clipping even with high gain is a great idea (for the ODA), it much more convenient, the end user would have to switch gain when they listen to strangely mastered albums.

And yes, that's laziness to some extent, but ergonomics is always a goal to achieve, is it not?

[NwAvGuy]

Originally Posted by khaos974

Concerning classical music, revising the gain structure to allow for no input clipping even with high gain is a great idea (for the ODA), it much more convenient, the end user would have to switch gain when they listen to strangely mastered albums.

And yes, that's laziness to some extent, but ergonomics is always a goal to achieve, is it not?

OK, I think we're talking about two slightly different things. I agree the input stage overload issue is worth fixing and with improving ergonomics. But the user is still left with the choice of how much gain the amp should have. If you set the gain so you can use at least half or two thirds of the volume range with typical recordings there won't be enough surplus gain left over for those really quiet recordings. So all the same trade offs still apply even if the input stage never clips. I should probably make this more clear in the O2 article.

[khaos974]

Originally Posted by NwAvGuy

OK, I think we're talking about two slightly different things. I agree the input stage overload issue is worth fixing and with improving ergonomics. But the user is still left with the choice of how much gain the amp should have. If you set the gain so you can use at least half or two thirds of the volume range with typical recordings there won't be enough surplus gain left over for those really quiet recordings. So all the same trade offs still apply even if the input stage never clips. I should probably make this more clear in the O2 article.

Currently, high gain (say 6X) + 2 Vrms source = input clipping, the problem is some recording are mastered very softly, thus a high gain is needed even with a standard 2 Vrms source.

For people wanting convenience, I was mostly think that with no input clipping one could set a relatively high gain as the default gain, then they would end up with the pot at 9 o'clock for pop music and 2 o'clock for recordings mastered at a low level.

For people who care less about convenience, they could switch gain to avoid using the potentiometer in it non optimal zone.

PS: If you did the measures with the 1.1 PCB with a the origin starting at a very low voltage instead of 0V, it may be a good idea to upload them.

[Satellite_6]

Originally Posted by NwAvGuy

I'm open to suggestions... but I'm leaning towards "ODA" - aka the Objective Desktop Amp. It's short, easy to type, pronounceable, and nothing resembling a headphone amp comes up doing a Google search. It also keeps the name in the "O Family"

That would be more sensible. . . anyone like my idea tho? xD

A 3V source has been found, oh my.

[maverickronin]

Originally Posted by NwAvGuy

I'm open to suggestions... but I'm leaning towards "ODA" - aka the Objective Desktop Amp. It's short, easy to type, pronounceable, and nothing resembling a headphone amp comes up doing a Google search. It also keeps the name in the "O Family"

How about the Ozone since ozone is O3?

[NwAvGuy]

Originally Posted by maverickronin

How about the Ozone since ozone is O3?

Hmmm... I like it except for one small detail... You should know how I feel about amps with a "3" in the name.

[maverickronin]

Did the uDAC2 get a pass...

If it's just called the Ozone there's no "3" in the actual name, only the explanation for it.

[Cain]

As much as I like the name "Ozone", and the ingenious resoning behind it, I would save it for later installments in the O series (maybe amp + DAC?). Renaming it now would give the idea of a "new" O2, or at the very least an improved version, what would be, in a way, the same as saying that there was something wrong with the original design, just because it won't work with every source out there. IMHO, just name it O2 Desktop, to avoid dissociating this version from the original.

[Satellite_6]

If not my awesome name. . .

The ODA or the O2' or O2 MK II.

[NwAvGuy]

Good points all, thanks. It might partly depend on how different it ends up from the current O2. It might be different enough to deserve a different name. Just to be clear, this won't just be the same board with the batteries removed, connectors moved around and the input overload issue addressed. There will be other worthwhile upgrades as well.

[Satellite_6]

Thanks for your effort!! I guess my questions about the gain were kind of pointless if I wait for this version.

[NwAvGuy]

For anyone who wants a portable amp its still good to understand how the O2 works. But considering FiiO can sell tens of thousands of amps that overload more easily than the O2, including their desktop amp, it's worth keeping it all in perspective.