The Opamp thread
Sep 14, 2009 at 9:57 AM Post #751 of 7,383
Quote:

Originally Posted by SpudHarris /img/forum/go_quote.gif
My bad, sorry been such a long time since I rolled with the P3
redface.gif
....You can use duals in the centre configuration, it was obviously the buffer positions I was thinking of. I usually use singles in L/R & G because I use Class A adapters to squeeze a little more from them which you can't do with duals.



Ahh what does a class A adapter actually do? I've heard some biased output or something does that have to do with class A opamps?
 
Sep 14, 2009 at 10:29 AM Post #752 of 7,383
Yeah, by shorting pins 6 and 7 with a resistor you bias to Class A. It's subtle but there are improvements if you have the right gear to reproduce them i.e. good cables, source, phones. Any improvement is a plus and when you get to a certain point even small ones are so worth it.
 
Sep 14, 2009 at 3:04 PM Post #753 of 7,383
Does biasing a chip opamp to class A make it use much more power and generate much more heat? or is that just with discrete class A?
 
Sep 14, 2009 at 3:47 PM Post #756 of 7,383
Quote:

Originally Posted by SpudHarris /img/forum/go_quote.gif
Yeah, by shorting pins 6 and 7 with a resistor you bias to Class A. It's subtle but there are improvements if you have the right gear to reproduce them i.e. good cables, source, phones. Any improvement is a plus and when you get to a certain point even small ones are so worth it.


what resistor value should be used?
 
Sep 14, 2009 at 4:02 PM Post #757 of 7,383
so what do you guys think of the JRC4580 anyway? maybe it's time for me to change my mind
evil_smiley.gif


that's the JRC4580, 0.0005% THD+N :


LME49720, THD+N rate of 0.00003% :


so yes, it's more distorted! but the LME49720 is nowhere near as beefy to drive headphones?! and to amplify a signal for headphones....you have to distort it anyway
frown.gif


I was pretty surprised by how distorted an Aria Arietta sounds...yikes, hated it! and most ppl swear by it
confused.gif


the line-out on the ST/STX seems to offer enough gain to drive headphones, but it's clearly not the case on the HD2...yet its HP out seems to sound better than the ST/STX HP out anyway, that had a very shrill upper spectrum to me(and to another friend too, using a dx1k).

I guess aiming at very low THD is nice and all...but someone's gotta feed the phones
redface.gif


I was about to roll the THS4032 on the line-out, but I'll keep the 3xLME49722 + JRC4580 on the HP out a bit I think...the sound is more distorted on HP out but my cd1k(same drivers as cd3k) is pretty bass shy, it takes power to make the low end sing I'm afraid
happy_face1.gif
 
Sep 14, 2009 at 5:18 PM Post #759 of 7,383
JRC4580 has higher output current, which makes it better for driving a load. AD8397 is one of the better chips for driving a headphone directly, far superior to JRC4580, IMO.
 
Sep 14, 2009 at 6:26 PM Post #760 of 7,383
oh ok, but it's soldered and it's not even SOIC8..so I'd rather not touch it
biggrin.gif


anyway, it seems to add an "analog" touch that's a nice contrast w/ the LME49722 clinical sound I think...something my previous soundcards were lacking, the mids are a lot less fatiguing on this card anyway(even on the line-out).

and I wonder if I could try the 3<>5 bypass trick in the AK4396 buffer op-amps now that I use the JRC4580 on the HP out
evil_smiley.gif


when I've first put the LME49722...they were crackling a bit, but they have stopped now?! I'll try the THS4032 tomorrow, and the OPA211 later on
smily_headphones1.gif
 
Sep 15, 2009 at 1:08 AM Post #761 of 7,383
Quote:

Originally Posted by SpudHarris /img/forum/go_quote.gif
I haven't noticed any increase in heat.


You stated your bias resistor will only work with a single channel. Is there asimilar modification for a dual channel opamp?
If not, maybe I will build a dual DIP8 adapter for two single channels biased to Class A. You have me curious to check this out a bit further..
 
Sep 15, 2009 at 2:06 AM Post #762 of 7,383
humm, I had forgotten about these 2 links I had been given on a french forum ages ago:

Google Translate

Opamp Design

NJM4580, distortion rate 0.00040% (-108dB)...he's not such a bad guy after all
smily_headphones1.gif


plenty beefy to drive headphones, and a nice contrast w/ the AK4396+2*LME49722! it gives that "tuby" sound ppl always talk about, a complete opposite of the ear-splitting sound of some other soundcards I used to own in the past. human ears were not meant to listen to raw sinewaves
redface.gif


I can't believe how bassy it makes my cd1k, yet very detailed...just not harsh AT ALL
darthsmile.gif


The Stranglers - Golden Brown is a perfect example, PRAT is high yet the harpsichord doesn't sound uber-bright whatsoever.
 
Sep 15, 2009 at 2:26 AM Post #763 of 7,383
Sep 15, 2009 at 6:50 AM Post #764 of 7,383
Quote:

Originally Posted by ROBSCIX /img/forum/go_quote.gif
You stated your bias resistor will only work with a single channel. Is there asimilar modification for a dual channel opamp?
If not, maybe I will build a dual DIP8 adapter for two single channels biased to Class A. You have me curious to check this out a bit further..



According to ''HiFlight'' (My OpAmp Guru) this can't be acheived. He's the guy who let me into this little secret + OBCA (Output Bypassed Class A) which will only work with AD744's plus one other OpAmp (NE5534 I think).

This is Class A Only Adapter.

P1010115.jpg



This is OBCA (Output Bypassed Class A) AD744's

P1010119.jpg
 
Sep 15, 2009 at 3:30 PM Post #765 of 7,383
Very helpful thread, Thank you for everybody.
Being inspired about opamp rolling I requested LT website for a samples. And now they asking me for some questions. I honestly don't know what should I answer to then.
smily_headphones1.gif
Here is questions:
Thank you considering Linear Technology Products in your design.

We request you to provide us the following details

Project name -
Application –
Brief program Description –
Part Number/s-
Quantity/Board-
Prototype Date/ Proto Qty -
Annual Qty -
Mass production date -
Production Location -
End Customer name - If it is available.
Design Location with complete Address -
Design Engineer Name & contact Details -

Could someone help me please, coz my own imagination wouldn't enough to answer it. Any though will really appreciate.
Thanks.
 

Users who are viewing this thread

  • D1N0
Back
Top