Different grounding, different sound quality or what can be the problem? - 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 28th June 2013, 03:03 PM   #1
diyAudio Member
 
Join Date: Sep 2011
Default Different grounding, different sound quality or what can be the problem?

This is the first and the second version of my amp
First:
Click the image to open in full size.
Second:
Click the image to open in full size.
Like you can see, I route wrong the volume pot in the first version, min-max = max-min and wrong channel too, i only know that after i finish soldering the first board. After that i do version 2 which i add 2 capacitors for power supply and different grounding. I think version 2 is more optimized at grounding because there are more dedicated ground traces but in fact when i listen, ver1 sound clearly better: clearer, bigger sound stage, ver 2 mid sound more close in, more mid bass. I think the different may be cause of the different perception of left and right ear although when using ver1, i turn my headphone around to get correct channels (my headphone has symmetric design) and ver1 still sound better. Then i cut the traces to the volume pot and re-route it to make the channel correct in ver1 and listen to it, still better than ver2 but then i solder the cut traces to take it back to original, one channel sound like ver2: less volume, more muddy, the other channel is clearer, sound bigger.
So the different between ver1 and ver2:
-Different grounding: ver1 star ground point at common 9v battery point, ver2 star ground point at one possitive leg of the filter capacitors.
-0.6mm signal traces with ver1, 0.5mm with ver2
Everything else is the same (i already add the filter capacitor for ver1)
So can you guy suggest me what can be the problem here?
  Reply With Quote
Old 29th June 2013, 07:29 AM   #2
marce is offline marce  United Kingdom
diyAudio Member
 
Join Date: Jun 2007
Location: Blackburn, Lancs
Because version 1 has better return current paths.
  Reply With Quote
Old 29th June 2013, 08:57 AM   #3
peufeu is offline peufeu  France
diyAudio Member
 
Join Date: Mar 2001
Location: Lyon, France
> ver2 star ground point at one possitive leg of the filter capacitors

Wrong star ground point, this node carries a halfwave rectified image of the output voltage...
  Reply With Quote
Old 29th June 2013, 09:06 AM   #4
diyAudio Member
 
Join Date: Sep 2011
Quote:
Originally Posted by marce View Post
Because version 1 has better return current paths.
Quote:
Originally Posted by peufeu View Post
> ver2 star ground point at one possitive leg of the filter capacitors

Wrong star ground point, this node carries a halfwave rectified image of the output voltage...
Thanks for your replies, can you explain more and give me some suggestion?
P/s: The top opamp is none invert amplifier and the two bottom opamp is for current buffer.

Last edited by proid; 29th June 2013 at 09:33 AM.
  Reply With Quote
Old 30th June 2013, 02:22 AM   #5
diyAudio Member
 
Join Date: Sep 2011
This is my new design, does it look any better?
Click the image to open in full size.
  Reply With Quote
Old 30th June 2013, 07:35 AM   #6
marce is offline marce  United Kingdom
diyAudio Member
 
Join Date: Jun 2007
Location: Blackburn, Lancs
Proid,
I am not a fan of the extreme spider leg star ground that o9nly appears on DIY audio designs to this extreme. All other audio and sensitive analogue designs I see and boards I lay out involve a ground plane (there is often more than one depending on the design, and the engineer, joined at a STAR POINT, for power, analogue, digital, though more often than not it is one for low level signals). This has been the case for the 27+ years I have been doing this. The extreme spider legs ground add inductance and limits the signals choice of return path, and forcing all the return current down thin legs is more likely to cause the lower frequency content swamp the lower power high frequency signals. It also does not take into account how signals really flow, being based I believe on the train of electrons running round the circuit...
The following presentation is an excellent primer to how signals actually travel round a circuit.
http://www.x2y.com/filters/TechDay09...%20_JohnWu.pdf

There are also numerous application notes and articles from National Texas etc on precession analogue design.
  Reply With Quote
Old 30th June 2013, 11:27 PM   #7
diyAudio Member
 
Join Date: Nov 2004
Location: Germany
It is certainly beneficial to think of "star" ground as more like a "tree" ground. The star point goes in the trunk.

Obviously using a ground plane is a good option when done right (and it should be considered here, as it seems we're looking at 2 layers). Cost or manufacturing concerns may limit a DIY design to single layer though, and then an idea how to obtain a star grounding/power scheme with low inductance where needed still is useful.

A ground plane is nice because current can basically flow back right underneath the power track it came from - which amounts to low inductance as loop area is minimized. You still have to watch out for current going directly from V+ to V- and try to minimize loop area there (assuming this can not be taken care of by local bypassing).
  Reply With Quote
Old 1st July 2013, 07:39 AM   #8
marce is offline marce  United Kingdom
diyAudio Member
 
Join Date: Jun 2007
Location: Blackburn, Lancs
I am not advocating a ground plane as such, though I do prefer them and they do give the lowest noise analogue layouts, but just a bit less manic on the branching of the "tree ground", again it seems over extreme and adds its own problems, such as increased inductance etc.
I do believe it has risen due to misunderstanding as to how currents flow. When you tell people that electrons travel at slow speeds when conducting electricity (1.6 x 10-4m/s) they are puzzled, this stems from the way current flow is illustrated, with a train track of electrons whizzing round the wires.
  Reply With Quote
Old 1st July 2013, 10:34 AM   #9
peufeu is offline peufeu  France
diyAudio Member
 
Join Date: Mar 2001
Location: Lyon, France
Quote:
Originally Posted by proid View Post
Thanks for your replies, can you explain more and give me some suggestion?
Here's a glorious napkin schematic.

As you can see, if the output stage gets out of class A (even a little) the power rails will carry halfwave rectified currents. I have drawn currents for a class B stage, but you get the idea. A wimpy opamp driving headphones will definitely get out of class A. Those currents carry lots of high harmonics and you do not want to couple them into your signal.

When thinking about layout you must think about loops, current always travels in loops, it doesn't care if you forget to design correct return paths...

So, leaving aside a lot of stuff, if we consider for example coupling via shared impedance, when a current travels in a bit of impedance (say, a ground track), it will create a voltage. If this impedance is shared with the reference of a sensitive node (say, your feedback network "ground" node, or whatever) then a voltage is injected in the sensitive node and you have a coupling problem. There are a lot of other ways for noise to couple inside sensitive spots too (magnetic, capacitive, EM waves, etc).

So.

If you have some coupling from a signal which is a linear image of the input, it isnt' so bad. Say, if 0.1% of the output voltage gets transferred to the ground node of your feedback network, noone will notice unless you're making a microvolt precision instrument which needs to be calibrated very accurately. For an audio device it would change your gain maybe a tiny fraction of a dB, or change the frequency response of your highpass filter, or introduce a little some interchannel crosstalk.

HOWEVER, if you have nonlinear class-B harmonics injected in your ground reference, because the currents in your power rails couple through the input, you will notice, it will sound like crap, it is in fact just like crossover artifacts.

Many advocate putting the ground reference like on this drawing, at least it is a spot that only carries linear current, not distorted ones.

Apart from that I'll second marce : ground plane is very useful, 1 cm of trace is enough to destroy the performance of modern capacitors, even good electrolytics...

sorry if my writing is a bit rough, I'm in a hurry
Attached Images
File Type: jpg hppscan19.jpg (18.7 KB, 84 views)
  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
Quality Control differences = variations in sound quality? KT Class D 3 4th June 2014 12:02 AM
sound quality vs sound quantity. Etocynned Full Range 203 26th May 2012 03:36 AM
Level-dependend Sound Quality by certainly Bass Drivers - very bad sound at low level tiefbassuebertr Subwoofers 11 3rd March 2012 10:54 AM
Level-dependend Sound Quality by certainly Bass Drivers - very bad sound at low level tiefbassuebertr Multi-Way 1 2nd April 2010 04:12 AM
Grounding problem LaZarus Solid State 8 23rd March 2004 12:15 PM


New To Site? Need Help?

All times are GMT. The time now is 03:27 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