[AstralStorm]

I'd be quite suprised if O2 sounded even a bit better with OPA627. That part would actually be inferior for IEM use due to higher current noise, similar voltage noise. It'd be "meh" for headphones.
Changing the capacitor could've reduced ESR pushing the compensation a bit up in frequency, however this shouldn't have any audible effect, unless you've managed to make O2 unstable.
Which is possible, OPA627 is faster and less stable than NE5532, it's also worse at driving capacitive loads, as typical.
OPA627 is further from rails, so you're wasting quite a bit of gain capability. In case you run out, it clips pretty horribly.
It's a part for the gain stage of an old-style analog scope and not audio.
 
Summary: It might sound different, but that would mean it's introducing distortion - most likely ringing or oscillation. Not to mention forcing you to pay through the nose.

 

[ddcpitt]

Might I suggest the C421 for portable and the upcoming ODA (Objective Desktop Amplifier) for home use?  The ODA will take the objective design of the O2, and improve upon it, making it even better for home use.  Hopefully the design if finalized by late summer and will be released sometime during the fall (hopefully as a DIY and pre-assembled by places such as JDS Labs).

 

[stv014]

Quote:

Originally Posted by AstralStorm /img/forum/go_quote.gif
 
Rather than the use of RMAA, I'm more worried that -57 dB crosstalk is the best your measurement setup can do...

 
In my opinion, -57 dB, especially if it was measured at 10 kHz (so it is probably better at lower frequencies), is low enough not to be audible.

 

[stv014]

Quote:

With upgrade version O2 : with op amp 627 and black gate resistor , O2 sounding much better

 
I would recommend reading the article "Op Amp Measurements" from August 2011 at the blog that must not be named. The measurements there show that using expensive op amps in the O2, especially at low gain, is most likely a waste of money, if not worse.

 

[ClieOS]

In my opinion, -57 dB, especially if it was measured at 10 kHz (so it is probably better at lower frequencies), is low enough not to be audible.

I would say the same as well. I actually got lower number from RMAA on both C421, around -68dB or so. Not nearly as good when compared to O2 and such, but still decent enough.

 

[stv014]

Quote:

I would say the same as well. I actually got lower number from RMAA on both C421, around -68dB or so.

 
How did you test it exactly ?
Actually, the O2 is "only" -65 dB, but that is likely a reliable measurement, and also with a 15 Ohm load (lower impedance loads generally increase crosstalk and distortion).

 

[ClieOS]

How did you test it exactly ?
Actually, the O2 is "only" -65 dB, but that is likely a reliable measurement, and also with a 15 Ohm load (lower impedance loads generally increase crosstalk and distortion).

That's the dark secret of RMAA. It should only be used to make relative measurement (against a known standard) because the number it puts out vary depends on the tools (i.e. sound card) you used for measurement and even different configuration. That's why quoting RMAA generally doesn't make it very helpful when you are comparing RMAA result from two different setups.

 

[Raines]

ODA + c-421 seems good idea

 

[AstralStorm]

Unsuprisingly, Audiotrak Prodigy Cube is more than enough of an amplifier for low impendances.
Beatifully sharp sound on 0 Ohm output (big jack), about as extended as E17.
(Maybe more, but I can't do A/B, need to finally send the FiiO devices for a repair)
 
Of course it won't drive 300 and 600 Ohm headphones - not enough power for that.
 
The sound is pretty suprising, since it uses the supposedly "warm" OPA2134. I hear nothing warm about that, correct (if not perfect) detailing and 0 noise instead.
If someone likes muffled sound with their IEMs, they're free to use the 32 Ohm minijack.

 

[stv014]

Quote:

That's the dark secret of RMAA. It should only be used to make relative measurement (against a known standard) because the number it puts out vary depends on the tools (i.e. sound card) you used for measurement and even different configuration. That's why quoting RMAA generally doesn't make it very helpful when you are comparing RMAA result from two different setups.

 
The most common problem with using sound cards for measuring crosstalk is that sound cards usually have unbalanced inputs and outputs referenced to a common ground (i.e. basically the metal chassis of the PC). Since with low impedance headphone loads much of the crosstalk occurs through the ground wire, the sound card loopback bypasses it, and the measured value is therefore better than it really should be. Also, for whatever reason, the crosstalk value (and the noise/dynamic range as well, by the way) printed by RMAA seems to be simply wrong; it does not agree with the value read from a real time FFT analysis while playing a test tone on only one channel under the same conditions as with RMAA.
For reasonably accurate crosstalk measurement, one method is to simply use a DMM that works at the tested frequency on the resistor load (it does not allow for measuring very low crosstalk with most inexpensive DMMs, but if it is less than -60 dB, then it is not much of an issue anyway). For PC based measurement, software other than RMAA might be worth trying, and either two separate machines for playback and recording (I did not test this), or differential input for recording (I tried this with a DIY differential amplifier, and it did make a useful difference in some cases).

 

[dannyd123]

is the c421 a better version than the cmoybb v2.03?

 

[tzjin]

You could say that, but it's a completely different amplifier. A Ferrari is basically a better version of the Camry. Both cars, but completely different performance levels. Read here: http://www.headfonia.com/flyin-high-jdslabs-c421/
 
While Mike and L enjoy the 2227 opamp, many on Head-fi enjoy the 8620. Just something to think about.

 

[Ultrainferno]

Quote:

You could say that, but it's a completely different amplifier. A Ferrari is basically a better version of the Camry. Both cars, but completely different performance levels. Read here: http://www.headfonia.com/flyin-high-jdslabs-c421/
 
While Mike and L enjoy the 2227 opamp, many on Head-fi enjoy the 8620. Just something to think about.

 
Many on Head-fi also prefer the 2227. Just something to think about. 

 

[dannyd123]

what is the difference between the 2227 and the 8620 versions?

 

[ClieOS]

what is the difference between the 2227 and the 8620 versions?

I reviewed both versions here: http://www.head-fi.org/t/615766/impression-jds-labs-c421-opa2227-vs-ad8620-featuring-cmoybb-o2-t5-and-uha-4