[Monago]

Hello,
 
I'm posting this thread here/the DIY forums because I assumed the knowledge of fundamental amplifier function and general technical know-how would be much higher than the amplifier forums. If this particular thread is inappropriately placed, let me know and I'll post elsewhere.
 
I have a Lovely Cube which I've recently discovered sounds vastly better with 0db of gain than any of the other optional settings (10db, 18db, and 20db). I purchased the class A 3.49mA bias mod with OPA627 opamp modification when buying my lovely cube, as I heard the SQ improvements were well worth the cost. With the class A bias modification installed in my amplifier at 0db, I'm hearing a wider, deeper sound-stage with better instrument separation and 'air'. At 10db, the amp becomes muddy, closed in, and looses all redeeming qualities over the headphone out on my source with the class a bias modification.
 
My question, is whether I'm hearing oscillations because of my amp being at 0db; it being potentially impossible that my OPA627 is stable with the class a bias mod on 0db  (at 3.49mA); or if I am indeed hearing a stable opamp at 0db of gain.
 
I've read that with a class A bias modification, you need higher gain to make the chip stable; however my confusion comes from the chip sounding better at 0db in class a than any other gain setting. Any clarification here would be greatly appreciated.
 
Oh, and my headphones are the HD650's if that makes any difference.
 
Thanks!
 
Edit: Changed to reflect proper gain settings.

 

[tomb]

I'm not familiar with that amp, but based on your description, I suspect that something else is going on besides the gain differences.  I've not heard any suggestion that Class A bias needs gain in the circuit to remain stable.  High Class A output buffers are always zero gain and the Class A bias has little to do with whether it oscillates.
 
As for oscillation itself and what it sounds like - generally speaking, it's characterized by a lack of bass.  You may also experience a "false" sense of increased highs (lots of tizziness?).  Eventually, you'll realize that the highs are not "better" in that case, though.  Your description doesn't really sound like either one.
 
I'm a bit surprised that you're using 300 ohm Sennheisers and are fiddling around with zero gain.  Those two items usually don't play together well - the Sennheisers usually need a reasonable amount of gain to provide enough voltage for a decent up and down swing of the music signal.
 
What could be happening is that the 10 gain setting is simply too much for the amp with Sennheisers as a load and you actually have some clipping going on.  Is there a way you could try something like a 5 gain?

 

[nikongod]

I like buffers with 300ohm headphones. Most headphones really. Within the limits set by your source's output voltage a simple buffer can be part of a reallllllly nice setup. Its really just a question of how loud you want the amp to go. A buffer certainly wont make the HD650 hit 120db from a "standard" 2v CD player, but SPL around 100db seems right. Whatever the SPL its plenty loud if you like to listen quiet  
 
I think that he is stating gain in dB. 0dBgain=voltage gain of 1
10dB=voltage gain of 3.2
 
It is slightly possible that you are hearing the differences between different parts of the potentiometer. This is a slightly Scientific, slightly Wild A** Guess, so take it for what little its worth. 

 

[Mad Max]

tomb's right; I think that something is amiss in his amplifier if the higher gain modes somehow sound much worse, which they should not at all.
There is no 5 gain setting by default, only 0, 10, 18, and 20.

 

[akiroz]

Sorry for the stupid question, but what is a Lovely Cube?
 
As for the OPA627, you mean 1x gain right?
0 gain means the input signal is multiplied by 0 so there should be no output signal.
This is electrically impossible.
According to the OPA627's Data sheet, it is "Unity-Gain Stable" Meaning it is stable at a gain of 1 or more
 

 

[Avro_Arrow]

Lovely Cube

 

[nikongod]

 
Quote:

Sorry for the stupid question, but what is a Lovely Cube?
 

It is a clone of another amp. You can find a schematic here:
http://www.head-fi.org/t/501046/the-lovely-cube-headphone-amp-lehmann-black-cube-linear-clone/375#post_7246299
 
Im not sure how much one is allowed to actually say about clone amps, so I wont say more. 
 
Everyone is citing dB, but not saying it. f'ing up units will f you up

 

[Monago]

First of all, thanks for the responses everyone!
 
Quote:

I like buffers with 300ohm headphones. Most headphones really. Within the limits set by your source's output voltage a simple buffer can be part of a reallllllly nice setup. Its really just a question of how loud you want the amp to go. A buffer certainly wont make the HD650 hit 120db from a "standard" 2v CD player, but SPL around 100db seems right. Whatever the SPL its plenty loud if you like to listen quiet  
 
I think that he is stating gain in dB. 0dBgain=voltage gain of 1
10dB=voltage gain of 3.2
 
It is slightly possible that you are hearing the differences between different parts of the potentiometer. This is a slightly Scientific, slightly Wild A** Guess, so take it for what little its worth. 

 

I did indeed mean db; I apologize for the miscommunication. Unfortunately, that just made answering my post much more of a chore for anyone willing to provide input. As for your potentiometer theory, I've been able to easily reach 70% volume (regular levels for 0db of gain) with some albums on 10db of gain, and didn't hear any discernible changes to the sound from the regular 10-12 o'clock position. The overall sound remains the same at each respective gain level regardless of volume level/potentiometer position, so this definitely isn't the issue.
 
 
Quote:

I'm not familiar with that amp, but based on your description, I suspect that something else is going on besides the gain differences.  I've not heard any suggestion that Class A bias needs gain in the circuit to remain stable.  High Class A output buffers are always zero gain and the Class A bias has little to do with whether it oscillates.
 
As for oscillation itself and what it sounds like - generally speaking, it's characterized by a lack of bass.  You may also experience a "false" sense of increased highs (lots of tizziness?).  Eventually, you'll realize that the highs are not "better" in that case, though.  Your description doesn't really sound like either one.
 
I'm a bit surprised that you're using 300 ohm Sennheisers and are fiddling around with zero gain.  Those two items usually don't play together well - the Sennheisers usually need a reasonable amount of gain to provide enough voltage for a decent up and down swing of the music signal.
 
What could be happening is that the 10 gain setting is simply too much for the amp with Sennheisers as a load and you actually have some clipping going on.  Is there a way you could try something like a 5 gain?

Your descriptions of oscillation most decidedly does not fit the sound I hear with my class A bias modification in place. The bass, in fact, becomes even more prevalent along with every other instrument, which seems to give an overall feeling of being "closer" to the music, so I suppose oscillations aren't the issue. 
 
However, you mention my lack of gain is something to be concerned about. I can't comment on the technicalities of my headphone or it's requirement for "more voltage" with higher gains, but only that I experienced nothing but sub-par sound quality when reaching said higher gains. Now, when I describe the sound as "sub-par", I only mean sub-par in contrast to 0 db. If I hadn't decided to flick it down to 0 db (something that seemed completely counter-intuitive for better SQ based on everything I've read in forums), I would have been happy with 10db or 18db of gain. But with time to kill, I started messing around for the hell of it, and found 0db to be flat-out better; slightly more of everything that wasn't present at 10db, 18db, or 20db of gain - detail, separation, etc.  Edit: So I'm not sure what to make of that, but I suppose the question remains: 
 
Should I be worried? (Maybe even about my ears 

)

 

[tomb]

Looking at the schematic kindly referenced by nikongod, it looks like the amp has a pretty decent output buffer and plenty of supply voltage, which is counter to a suspicion of clipping.  Thanks also to nikongod for clearing up the units issue.

 
I still wonder if clipping is still the issue.  Here's a question out of left field, but could you be using an unusually strong source - maybe even another amp or the headphone out of a CD player/digital player? 

 

[nikongod]

 
Quote:

I did indeed mean db; I apologize for the miscommunication. Unfortunately, that just made answering my post much more of a chore for anyone willing to provide input. As for your potentiometer theory, I've been able to easily reach 70% volume (regular levels for 0db of gain) with some albums on 10db of gain, and didn't hear any discernible changes to the sound from the regular 10-12 o'clock position. The overall sound remains the same at each respective gain level regardless of volume level/potentiometer position, so this definitely isn't the issue.

 
Thanks for confirming that its not a problem in the pot. Sometimes it happens. 
 
The area between 50% and 70% spin is my favorite part of the pot, BTW. It usually gives you the best matching the pot can do, and still enough room to spin it up some more for quiet recordings. 
 
Quote:

 
However, you mention my lack of gain is something to be concerned about. I can't comment on the technicalities of my headphone or it's requirement for "more voltage" with higher gains, but only that I experienced nothing but sub-par sound quality when reaching said higher gains. Now, when I describe the sound as "sub-par", I only mean sub-par in contrast to 0 db. If I hadn't decided to flick it down to 0 db (something that seemed completely counter-intuitive for better SQ based on everything I've read in forums), I would have been happy with 10 or 18 gain. But with time to kill, I started messing around for the hell of it, and found 0 gain to be flat-out better; slightly more of everything that wasn't present at 10, 18, or 20 db of gain - detail, separation, etc.  Edit: So I'm not sure what to make of that, but I suppose the question remains: 

So it sounds like you are saying that the high gain settings are pretty good, but the low gain/buffer setting is better. You are concerned about that because it is the opposite of everything you have read elsewhere. 
 
Ok, here are some reasons that the low gain setting is usually better than higher. 
The bandwidth of an op amp decreases as gain increases. It may still be quite a lot at gain of 10=20dB, but why have less bandwidth when you could have more for free?
The random noise produced by the op amp and the parts around it increase as noise increase. Why have more noise when you could have less for free? 
The channel-to-channel matching of potentiometers is worse at low volume levels (just above the mute position). Again - better channel matching is better and free!
With low gain you have lots of control over the volume level - if you want to make a very small change of volume its very easy with low gain. With high gain this can be somewhat tricky.
There are a couple more for amps without global feedback, but this should be enough
 
There are 2 reasons that the higher gain setting is better.
You really dont have enough gain - the volume control is absolutely maxed out and you need more. 
The op amp is not stable at lower gain. 
 
I think that the biggest misconception about gain is that more gain=more power. This is totally wrong. Gain has absolutely nothing to do with power, but there is a "feeling" of power associated with just gently tapping the volume knob and having the sound go from calm to OMFGloud. Compared to an amp where you have to spin the knob way up high to get to your listening level it is easy to think this. The common correlation (perhaps even subconsciously) is to an automobile where you need to practically floor a low-powered car to merge onto the highway VS just tapping the gas and poof your going 65 in a very powerful car. 
 
The above misconception & preference for more gain spans pretty much the whole audio industry. 
 
Quote:

Should I be worried? (Maybe even about my ears 

)

In my totally not-medically qualified opinion - Only if you listen really loud. Get an SPL meter and try to keep the average in the 60-70dB range and you should be totally fine. OSHA says 80dB, or even slightly higher for reduced time, but why tempt fate?

 

[Monago]

Firstly, thanks again to everyone responding! Greatly appreciated!
 
Quote:

Looking at the schematic kindly referenced by nikongod, it looks like the amp has a pretty decent output buffer and plenty of supply voltage, which is counter to a suspicion of clipping.  Thanks also to nikongod for clearing up the units issue.

 
I still wonder if clipping is still the issue.  Here's a question out of left field, but could you be using an unusually strong source - maybe even another amp or the headphone out of a CD player/digital player? 

The output from my sound card shouldn't be any more than what you'd experience from any other run-of-the-mill dac or CD-player. Additionally, the only amplification being provided 'down the line', is from the aforementioned Lovely Cube - so again, unless there's something internally malfunctioning or amiss with my amplifier that's causing clipping, I'm doubting that could be the cause of the sound differences I'm hearing. 
 
And it's a relief to know my amplifier has plenty of voltage 

. I was getting a bit worried there for a moment.

 
Quote:

 
 
Thanks for confirming that its not a problem in the pot. Sometimes it happens. 
 
The area between 50% and 70% spin is my favorite part of the pot, BTW. It usually gives you the best matching the pot can do, and still enough room to spin it up some more for quiet recordings. 
 

So it sounds like you are saying that the high gain settings are pretty good, but the low gain/buffer setting is better. You are concerned about that because it is the opposite of everything you have read elsewhere. 
 
Ok, here are some reasons that the low gain setting is usually better than higher. 
The bandwidth of an op amp decreases as gain increases. It may still be quite a lot at gain of 10=20dB, but why have less bandwidth when you could have more for free?
The random noise produced by the op amp and the parts around it increase as noise increase. Why have more noise when you could have less for free? 
The channel-to-channel matching of potentiometers is worse at low volume levels (just above the mute position). Again - better channel matching is better and free!
With low gain you have lots of control over the volume level - if you want to make a very small change of volume its very easy with low gain. With high gain this can be somewhat tricky.
There are a couple more for amps without global feedback, but this should be enough
 
There are 2 reasons that the higher gain setting is better.
You really dont have enough gain - the volume control is absolutely maxed out and you need more. 
The op amp is not stable at lower gain. 
 
I think that the biggest misconception about gain is that more gain=more power. This is totally wrong. Gain has absolutely nothing to do with power, but there is a "feeling" of power associated with just gently tapping the volume knob and having the sound go from calm to OMFGloud. Compared to an amp where you have to spin the knob way up high to get to your listening level it is easy to think this. The common correlation (perhaps even subconsciously) is to an automobile where you need to practically floor a low-powered car to merge onto the highway VS just tapping the gas and poof your going 65 in a very powerful car. 
 
The above misconception & preference for more gain spans pretty much the whole audio industry. 
 

In my totally not-medically qualified opinion - Only if you listen really loud. Get an SPL meter and try to keep the average in the 60-70dB range and you should be totally fine. OSHA says 80dB, or even slightly higher for reduced time, but why tempt fate?

Fantastic response, nikongod!! Thank you for taking the time to so thoroughly explain! That literally answers every question I had about the differences in sound between gain settings.
 
It's interesting that you point out higher gain/higher volume being subconsciously perceived as more powerful, as that was my thought process after finding the lower gain setting to be more resolving. It would be hard for one not to feel there's "more power" behind a sound when so little a movement on the potentiometer is needed to change volume levels, just as "quality interconnects" supposedly change the sound to become "more robust and full" because they look bigger and are more expensive. The expectation of a desired result is often more powerful than what is not objectively the case, and like anything that can be open to some interpretation (like audio), pseudoscience and superstition become common practice so long as belief in something not immediately, objectively provable remains more powerful than the belief in hard, scientific fact. Gawd - it's downright incredible how much of that is floating around the audiophile community. Anyway, I'll end this post here before I get into trouble...
 
Thanks again for the clarification, nikongod. Truly excellent post!