National LME49600 Reference Design Project - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems

Headphone Systems Everything to do with Headphones

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 2nd April 2011, 02:01 PM   #21
diyAudio Member
 
RocketScientist's Avatar
 
Join Date: Mar 2008
Location: Northwest
Quote:
Originally Posted by Ferrari View Post
RocketScientist, how about adding a third channel (a copy of the L/R channel, with unity gain) to create a 3-channel headphone amp?
The 3-channel topology keeps headphone return currents from disturbing the signal ground.
Btw, count me in as interested.
A third channel might be interesting to at least test. I know the M Headphone Amplifier ) uses one and it seems to measure well and has a loyal following.

My personal belief is with proper PCB layout and grounding you shouldn't need a third channel. If a 200 watt power amp with ground currents 100 times higher can have really low distortion, it's comparatively easy to manage the ground currents in a headphone amp.

Some claim a third channel also provides a more "symmetrical" drive for the headphones, lowers crosstalk, etc. But I haven't seen any proof of those supposed benefits. There's also a downside. In terms of noise performance a 3 channel amp will potentially be worse. It's also significantly less efficient, produces more heat, costs more, requires a bigger PCB, and potentially a bigger enclosure.

It would be fun to test the claims but I'm curious what others think? Is it worth testing? It might be possible to have an optional 3rd channel on a PCB that was about 50% larger with some jumpers, etc.

Or I could possibly provide some pads/jumpers on a 2 channel PCB making it easy to use a 2nd PCB only half populated as a ground channel. But that might somewhat compromise the grounding layout.

Ultimately, the effects of ground currents, crosstalk, etc, are measurable. So the Prism dScope could help make the choice easier. Does anyone know if there are comparable measurements anywhere for a 2 ch vs 3 ch design?
__________________
http://nwavguy.com - Personal non-commercial audio blog
  Reply With Quote
Old 2nd April 2011, 03:26 PM   #22
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
I've explained the engineering analysis of "3-channel" amplifiers as advocated by some at Head-Fi in several posts there - I have given reasons that the claimed "advantages" don't work as advertised

there is a distinction between "3-channel" and "active gnd" - "active gnd" is sometimes used to split a single supply such a a battery or wall wart - but proper wiring connects input signal gnd and load gnd together for fewer introduced errors than the "3-channel" scheme

Quote:
Originally Posted by jcx View Post
The Pimeta’s "3-channel" topology is fundamentally flawed - most technical reasons given for it are simply not true or can be bettered by proper layout

"active supply splitter" can be useful when wanting to use single V supply - the single active gnd should be used for input, feedback and load gnd

but "3-channel" puts the gnd “buffer” amp in series with the load - with the input and feedback grounds separated from the load gnd - this means all of the error of the "output gnd" buffer is in series with the headphones - giving distortion and "output gnd" impedance crosstalk that is worse than with a good "passive" gnd of heavy conductor with signal input gnd, feedback gnd "star" connected with the load common ground pin of the TRS connector


A "active gnd" showing star at output - can better "3-channel" performance:
Click the image to open in full size.
...
I would go for 4-pin heaqphone cable connectors, separating R/L load gnd paths (TRS "single point" contact can have 10 mOhm common gnd R in the jack/plug interface)
dual polarity supply, then "dual mono" and/or bridged/balanced output for real engineering advantages in an "ultimate" headphone amp

the BUF and LME (and TPA6120 - I don't know why peole get stuck on only condidering unity gain buffers in the loop) are so good that full Class A output operation is probably of no audible advantage but it does ease the layout considerations of keeping the "half-wave" Class AB +,- supply currents from giving some nonlinear common impedance couping on the board

I've also shown how to bias op amps for Class A output when they can handle the heat
High loop Gain Composite Op Amp Circuits
http://www.head-fi.org/forum/thread/...397-class-a/15

at one Head-Fi gathering some who swore they could hear differences in op amps couldn't tell when my parallel TPA6120 output amp was biased Class A at ~220 mA vs the few mA when the bias V was turned off

Last edited by jcx; 2nd April 2011 at 03:56 PM.
  Reply With Quote
Old 2nd April 2011, 05:03 PM   #23
diyAudio Member
 
RocketScientist's Avatar
 
Join Date: Mar 2008
Location: Northwest
Thanks JCX. That pretty much sums up most of my thoughts on a third channel. But I was trying to keep an open mind and/or wondered if I might be missing something.

I'm all for innovation and creativity but analog class A/B audio amplifier design is relatively mature. Lots of things have been tried over many decades but few gain long term traction. Once we achieved sufficient speed, linearity, bandwidth and direct coupling, the rest has mostl been refinement of already well proven conventional circuit topologies and the semiconductors themselves.

There have been various fancy bias schemes alleging Class-A performance with Class AB efficiency, "smart" (switched multi-rail, etc.) power supplies, "diamond bridge" distortion canceling circuits, MOSFET outputs, feed-forward, low TIM designs, IGBT's, and many more new twists. But surprisingly few of these "innovations" stayed around and/or ever gained much acceptance. Many remind me of the 100 MPG carburetors and other things that were supposed to magically quadruple the efficiency of internal combustion engines but never seemed to survive any sort of objective testing.

Is there a preferred PCB mount 4 pin headphone connector? Assuming there's enough room along one edge, it would be easy to provide for that. I was already thinking of both 1/4" and 3.5mm to avoid external adapters.

Does anyone have thoughts regarding the power supply alternatives on the previous page?
__________________
http://nwavguy.com - Personal non-commercial audio blog
  Reply With Quote
Old 2nd April 2011, 05:56 PM   #24
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
I would probably cast a vote for option #2 for the PS, include something on the board with the option of not populating it and instead using something external.

I had mentioned a Jung SR or clone in that "wire" thread, but there are so many good supplies out there and everyone has their favorite. I'm building up an AMB Sigma22 right now for another project that I may wind up re-purposing for this board.

Just my two cents.
  Reply With Quote
Old 2nd April 2011, 09:09 PM   #25
Luke is offline Luke  New Zealand
diyAudio Member
 
Join Date: Sep 2001
Location: Wellington NZ
Send a message via AIM to Luke
Quote:
Originally Posted by RocketScientist View Post
A third channel might be interesting to at least test. I know the M Headphone Amplifier ) uses one and it seems to measure well and has a loyal following.

My personal belief is with proper PCB layout and grounding you shouldn't need a third channel. If a 200 watt power amp with ground currents 100 times higher can have really low distortion, it's comparatively easy to manage the ground currents in a headphone amp.

Some claim a third channel also provides a more "symmetrical" drive for the headphones, lowers crosstalk, etc. But I haven't seen any proof of those supposed benefits. There's also a downside. In terms of noise performance a 3 channel amp will potentially be worse. It's also significantly less efficient, produces more heat, costs more, requires a bigger PCB, and potentially a bigger enclosure.

It would be fun to test the claims but I'm curious what others think? Is it worth testing? It might be possible to have an optional 3rd channel on a PCB that was about 50% larger with some jumpers, etc.

Or I could possibly provide some pads/jumpers on a 2 channel PCB making it easy to use a 2nd PCB only half populated as a ground channel. But that might somewhat compromise the grounding layout.

Ultimately, the effects of ground currents, crosstalk, etc, are measurable. So the Prism dScope could help make the choice easier. Does anyone know if there are comparable measurements anywhere for a 2 ch vs 3 ch design?
Hi Rocketscientist,

I have an M3 and while it does some things really well, Im not sure I like it compared to my F4. But I really need more time with it before I make any decisions. But the reason Im posting here is I like the sound of your project and have always wondered about the ground channel. I personally would be interested in having one that can be be bypassed at the flick of a switch cos I cant see how they can possibly help, but then again Ive been wrong many times in audio.
__________________
If you give a man a fish he will eat for a day. But if you teach a man to fish he will buy an ugly hat. And if you talk about fish to a starving man then you are a consultant. Dilbert
  Reply With Quote
Old 2nd April 2011, 09:29 PM   #26
diyAudio Member
 
Join Date: Aug 2009
Location: Sacramento
Hello,
4 pin output preference? If the L&R grounds are separate that means that 99 + % of the stock headphones will need to be rewired. My preference would be a mini 4-pin XLR.
Power supply preference is flexibility. The Op-Amps offer high PSRR to start with and 317 / 1085 type regulator will work well however others prefer dark as the night shunt regulators. I like gel cells for this type of application.
DT
All just for fun!
  Reply With Quote
Old 2nd April 2011, 10:21 PM   #27
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
another point for using a CFA with gain as the outupt buffer is that you can then subregulate the input op amp ps withou losing any Vswing at the output - the sub regualtion is much more effective because the input op amp doesn't have any significant dynamic current demand
since the op amp psrr is the op amp excess loop gain you get ridiculus psrr in the composite amp
  Reply With Quote
Old 3rd April 2011, 03:15 AM   #28
diyAudio Member
 
RocketScientist's Avatar
 
Join Date: Mar 2008
Location: Northwest
Quote:
Originally Posted by jcx View Post
another point for using a CFA with gain as the outupt buffer is that you can then subregulate the input op amp ps withou losing any Vswing at the output - the sub regualtion is much more effective because the input op amp doesn't have any significant dynamic current demand
since the op amp psrr is the op amp excess loop gain you get ridiculus psrr in the composite amp
Yes, good point. Perhaps I should include jumper pads and an easy to cut trace in the power supply rails between the gain and output stage? That would allow someone to easily split the supplies as you suggest. Because of the high PSRR it's likely overkill, but some like overkill

In a big audio power amp you want the output stage to swing as close to the rails as possible for maximum efficiency. So having either gain in the output stage, or higher rails for the earlier stages is a very good idea. But, in a headphone amp, burning a few more tenths of a watt in the output stage isn't a big dea. And, with 15 volt rails, there's more output swing than I can imagine just about anyone needing.

I know you're a huge fan of the TP6120 and it looks to be an impressive part. But I have one serious reservation. TI says it may need to dissipate upwards of 2 watts of power. And DIY'ers don't have any good way to properly solder the "hidden" heatsink pad underneath the device without reflow equipment and solder paste. So it's much less DIY-friendly compared to the simple 5 pin tabbed package of the LME49600.

I realize you came up with a solution to this problem mounting the parts upside down, re-bending the pins, etc., but it's not a very practical DIY solution for most. My goal is a relatively low cost, simple, easy to build solution with world class performance. So elaborate heat sinks, mechanical work, and/or tricky soldering are not ideal.

If I thought the TP6120 would yield significantly better performance I'd be more open to it. But I'm not sure it will? It has higher peak current capability, and can have gain, but do you think it would yield lower overall total distortion than the LME49600 design?
__________________
http://nwavguy.com - Personal non-commercial audio blog
  Reply With Quote
Old 3rd April 2011, 03:25 AM   #29
glt is offline glt  United States
diyAudio Member
 
Join Date: Oct 2004
Quote:
Originally Posted by RocketScientist View Post
... My goal is a relatively low cost, simple, easy to build solution with world class performance. ...
That's my type of project :-)
__________________
www.hifiduino.wordpress.com
  Reply With Quote
Old 3rd April 2011, 05:00 PM   #30
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
Default with added gain in the loop there is "extra" distortion reduction

there is no chance that the output stage will limit the performance with the TPA6120 in a multiloop - just look at the distortion spec when operating by itself, providing all of the loop gain

the LME49600 distortion # is for the buffer operating inside the LME49710 55 MHz GBW feedback loop

apply the same factor of excess loop gain for all audio frequencies to the TPA6120 distortion plots and the output distortion will not be measurable
since the LME47910/4562 is nearly a pure integrator just divide the 55 MHz GBW product by frequency and by amplifier closed loop gain to get the excess loop gain (or feedback factor) output stage distortion reduction at any frequency
for Av 10, 1KHz the output distortion is reduced by ~5000x

but you can also add loop gain with the TPA6120 unlike a buffer - use modified integrator feedback around the TPA6120 so its gain can be >100x over all audio frequencies
then the input op amp then only swings ~ 100 mV at audio frequencies
2nd order distortion decreases in percentage 1:1 with level, third order by the square of level - higher order distortions follow the n-1 power law too so they disappear even more quickly
so with added loop gain from the TPA6120 the input op amp is operating at >100x increased linearity due to the reduced signal level it has to provide - this is in addition to the ~1/F reduction in output stage distortion which doesn't "care" where in the loop the excess gain is distributed



as for heat sinking the TPA6120 should do ~ 1W without the power pad connection - quite enough in Class AB if you've managed to coordinate your ps V and load for most headphones

Perander's QRV07 project uses the TPA6120 - a power pad soldering comment is here: Re: TPA6120 headphone amplifier power supply capacitance for the output THD specified - Audio Amplifiers Forum - Audio Amplifiers - TI E2E Community

I would try using one large hole instead of the via grid for hand soldering - say ~ 1/2 the area of the power pad - then you can get solder, soldering iron tip in the hole from the backside

for just 1 or 2 TPA the upside down approach could use much smaller video chip coolers - a spring clip giving a clamping force of ~20x the weight of the tiny Al heatsink shouldn't bend the pcb out of shape

Last edited by jcx; 3rd April 2011 at 05:07 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
National LME49810TB/LME49830TB UltraHighFidelity HighPower Amplifier Reference Design KLe Chip Amps 75 21st October 2013 12:47 PM
Help with Dayton Reference 2-way Project coldcut Multi-Way 10 23rd May 2010 01:07 PM
How about a 3 inch reference project? type Full Range 52 13th October 2007 04:17 AM


New To Site? Need Help?

All times are GMT. The time now is 01:22 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2