[Steve Eddy]

Quote:

Don't take the bait, Steve!

Bait? That was the most delicious brownie I've ever eaten! Who in their right mind would use something so scrumptious as bait?
 
se
 
 
 

 

[Achmedisdead]

Quote:

 
 
 
 

+1
 
 

 

[The Monkey]

Quote:

The O2 is as suitable as a desktop amp as it is as a portable/transportable because of it's performance - low output impedance (0.58 ohms), gobs of power (relative to it's size) and very respectable distortion levels. You can't do better within the audible range.
 

 
Well, it purports to be for now.  But I agree with the premise.  So for those people who are limiting comparisons to portable amps or dismissing hot sources because no portables are that hot--knock it off.

 

[amb]

Quote:

Quote:

Originally Posted by amb 
 
For 1Vrms into 32 ohms, the power being delivered is a bit over 31mW.  To get 1mW into 32 ohms, you need only 0.18Vrms. Doing a conversion from one to the other. the PX100's sensitivity becomes 99dB/mW.  I mention all this to show that it's not the most efficient portable headphones compared to many others.  Yet it is an immensely popular model so it's appropriate to use as an example.

 
Where do you get 0.18 V from? You need 1 volt RMS at 32 ohms to get 31 mW.

 
Yes, P = 1V ^2 / 32 = 31.3mW.  Isn't that exactly what I said (bold green above)?
The 0.18Vrms is how much voltage is needed to deliver 1mW into 32 ohms (bold blue above).  0.18V ^2 / 32 = 0.001W = 1mW.
In my original post I said that I calculated the dB/mW figure in order for the PX100's sensitivity to be directly comparable to other headphones.
 
Please read carefully first before acting as if you found a mistake.  I don't make such posts haphazardly.
 
Quote:

Quote:

How much gain is needed?  You must take into consideration the sensitivity of the headphones, the output level of the source, and add enough headroom to account for recordings that are low in volume so that you don't run out of volume pot travel.  Many recordings are almost constantly pegged near the 0dBFS line and dynamically compressed.  These recordings will sound very loud (and bad).  But there are also recordings that are recorded very low in level (some from the late 70s and 80s for example), maybe 10-15dB below 0dBFS.  Thus, it is prudent to select an amp gain level above and beyond what's needed to amplify a 0dBFS signal to compensate.

 

The target number for the O2 is 110 dB peaks which is seriously loud. The PX100 needs 4 dB less than 1 V RMS to hit 110 db SPL. The math is antilog(-4/20) = 0.63. 1 V RMS * 0.63 = 0.63 V  RMS.
 
With an 0.5 V RMS source the O2 would need at least .63/.5 = 1.3X gain to achieve 110 dB SPL peaks. A gain of 2X would provide some extra volume control range.
 
To give an example of how loud 110 dB really is that's about the Mini3's limit driving the HD600.

 
That's all fine, but you fail to notice what I wrote in bold and take into account the 15dB headroom I added...  The point is not to play above 110dB, but to have enough gain to compensate for quiet recordings.  It would be incorrect to design the gain structure of your system based purely on 0dBFS numbers.  Music recordings are not a full-scale amplitude sine wave.
 
Quote:

...
Even with a 0.5 V RMS source 5X gain is ridiculous with the PX100.

 
Yes, 0.5Vx5 = 2.5Vrms would be more than loud on the PX100, but there is a volume control to let you regulate the actual volume.  Let me repeat yet another time in case it still hasn't sunk in, the point of the gain is not to actually play at loud levels with a 0dBFS signal, but to allow quiet recordings to reach loud levels without max-ing out the volume pot and left with no more room to adjust.
 
Quote:

You forced me to get some books out to disprove your claims, and I don't appreciate that at all. I've enough on my plate as it is without having to devout hours of my time to this forum thread. In the future I'd suggest people look at claims that seem backed by math and physics with suspicion - unless you can verify the numbers for yourself, don't trust them to be accurate.

 
We forced you?

 

 

[b0ck3n]

People seem to also forget that the O2 has the same distortion levels from 0 to max volume, meaning that as long as you've got a volume control on your source or DAC you can make sure to stay within the limits of the amp clipping very easily. The misconception that the O2 clips before getting loud enough is just that, a misconception.

 

[b0ck3n]

Quote:

 
Yes, P = 1V ^2 / 32 = 31.3mW.  Isn't that exactly what I said (bold green above)?
The 0.18Vrms is how much voltage is needed to deliver 1mW into 32 ohms (bold blue above).  0.18V ^2 / 32 = 0.001W = 1mW.
In my original post I said that I calculated the dB/mW figure in order for the PX100's sensitivity to be directly comparable to other headphones.
 
Please read carefully first before acting as if you found a mistake.  I don't make such posts haphazardly.
 
 
That's all fine, but you fail to notice what I wrote in bold and take into account the 15dB headroom I added...  The point is not to play above 110dB, but to have enough gain to compensate for quiet recordings.  It would be incorrect to design the gain structure of your system based purely on 0dBFS numbers.  Music recordings are not a full-scale amplitude sine wave.
 
 
Yes, 0.5Vx5 = 2.5Vrms would be more than loud on the PX100, but there is a volume control to let you regulate the actual volume.  Let me repeat yet another time in case it still hasn't sunk in, the point of the gain is not to actually play at loud levels with a 0dBFS signal, but to allow quiet recordings to reach loud levels without max-ing out the volume pot and left with no more room to adjust.
 
 
We forced you?

 

Read my last post: max out the O2 volume and use the volume control on your source or DAC. The O2 will play louder than the Mini3 so you really don't have to worry about headroom.

 
 

 

[amb]

Quote:

People seem to also forget that the O2 has the same distortion levels from 0 to max volume, meaning that as long as you've got a volume control on your source or DAC you can make sure to stay within the limits of the amp clipping very easily. The misconception that the O2 clips before getting loud enough is just that, a misconception.

 
Other than something like the headphone output on portable audio players or software volume control on the computer, how many sources (i.e., CD players, DACs) actually have a volume control on it?  Yeah, Benchmark DAC1 has one that could be optionally enabled, but look at the field of sources currently available, commercial or DIY, you'll find that assuming a source with a volume control is not a good plan.
 
And to play "devil's advocate"... even if the source has a volume control, the moment you use it to turn down the signal volume prior to it entering the amp's input stage, you've reduced some S/N ratio right there.  If you then have to turn up the volume on the amp to make up for the lost volume, then wouldn't that be quite the antithesis of what the designer was trying to do with the volume pot after the output of the gain stage?
 
 
 

 

[Shike]

Add a passive preamplifier like the NHT PVC or set the gain properly for the source you'll be using - problem solved.  This isn't hard or something that can't be worked around unlike certain (and IMO larger) issues with other designs.

 

[Anaxilus]

Quote:

I can't believe that I agree with kwkarth on something.
 
For reference, I am a scientist and skeptical of many audiophool traditions...... but I strongly believe that anyone who thinks an amp is perfectly transparent because of a select and limited group of measurements is wrong. It shows a substantial amount of arrogance, and a strong underestimation (or misunderstanding) of the complexities of both the electronics and the human brain.
 
The amp needs to be ABX'ed, double blinded, some sort of proper listening test. Until then, both the objectivists and the subjectivists are full of it.

Thank goodness.  Well said.  
 

 

[Head Injury]

amb, I was thinking the same thing about source volume. Not many have it, it's hardly a solution at all.
 
Quote:

If you then have to turn up the volume on the amp to make up for the lost volume, then wouldn't that be quite the antithesis of what the designer was trying to do with the volume pot after the output of the gain stage?

Yes, but it's sort of a worst case solution. Not that many source/headphone combinations will have a hot enough line out and need enough gain to cause clipping in the first place. With a 2Vrms source on AC power you can get a gain of what, 3.5x? That's enough to drive the HE-6 to 110dB. The new default of 2.5x gain is enough to handle a 2.8Vrms source, and drive the HE-6 to 107dB. At least according to Anonanimal's awesome spreadsheet.
 
If you're building it for portable use, set a low gain. You won't need the extra power, and you'll give the batteries a breather. Then you won't have to mess with the source volume.

 

[kwkarth]

Quote:

But how do you know how something "actually sounds" given that how something sounds is a highly subjective thing and can vary dramatically among listeners?
 
Some think Julia Roberts is absolutely gorgeous. Others think she's rather ugly. So how does Julia Roberts "actually look"?
 
From the O2's designer's perspective, its objective performance is such that what alterations it does make to the signal are so far below known audible thresholds that by all evidence, it should not impart any sound of its own.
 
se

In the context we find ourselves with respect to this amp, the "objectivity" that is being represented, isn't.  So give me some subjective listening impressions.  At least those may be off discworld.
 

 

[Shike]

Quote:

In the context we find ourselves with respect to this amp, the "objectivity" that is being represented, isn't.

If the measurements are factual I fail to see where you're coming from, if it's accurate to the signal it's accurate to the signal - what else do you want people to say?

 

[kwkarth]

Quote:

If the measurements are factual I fail to see where you're coming from, if it's accurate to the signal it's accurate to the signal - what else do you want people to say?

Tell me all about the measurements you made.
 
 

 

[Currawong]

I think AMB's point is that, even if the amp measures well in the lab, once people start using it in a variety of situations, due to various factors about the source used, they may not be able to use in in a manner that gets the best results, both measurable and audible, so the measurements may be moot. 
 
This has come because I think nwavguy has pitched it in a "Fight the man!" manner, which appeals to some people here, suggesting that for $35 (or whatever the total is) you can have an amp that will be good enough for everything. Now the people who know what they are talking about have come in and started telling it like it is, some of you don't want to hear it.  I suggest some of you should ditch the battle mindset, as anyone building this will need to know as much as they can and any issues found now will only benefit everyone in the long run. 

 

[khaos974]

Nice post Curra, this thread has turned too confrontational; anyway, the 1.1 PCB design has been released, default gain is now 2.5.x.