High bias class D - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

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 24th August 2010, 12:21 PM   #1
sendler is offline sendler  United States
diyAudio Member
 
Join Date: Aug 2004
Location: ny
Default High bias class D

I decide to start a new thread on this topic as the links are dead on the old one and the old thread title would not draw in new viewers to the topic I want to discuss which is the huge improvement in sound quality of a Class D amp that comes from using output filter coils with values that are much smaller than what would otherwise be dictated by the universally accepted "textbook" slope.
Quote:
Originally Posted by analogspiceman View Post
Hello DNA,

It's been discussed on this forum before, but you are seeing the curious effect on deadtime of the peak-to-peak current of the output inductor current of a class-d amplifier.  As long as the output inductor current ripple is larger than the load current at the moment of switching, its current is in the right direction to force the output stage node voltage to begin slewing to the other rail immediately when both MOSFETs turn off at the beginning of deadtime (this is natural commutation - also known as soft switching).  For larger load currents, slewing doesn't commence until the end of deadtime when the other MOSFET forces the voltage to change.  Thus, unlike a linear class-b amplifier (where deadtime is at zero output) there are two deadtime zones at non zero load current (symmetrically spaced about zero)
.
This is an interesting concept and supports my listening tests where the lowest tolerable coil value (I go as low as 3.5uH on each leg of my TK2050 amps) for the output inductor yields the best sound quality. Everyone blamed the change in sonics on the increased peak at 65KHz lifting the frequency response in the audio band but it is only a .5db increase at 20KHz so I knew that there must be another mechanism at work for the large improvement. So by accepting an out of audio band peak and some loss of efficiency at idle, rather than arbitrarily designing a 2nd order filter with a textbook slope, we will improve the sonics of a class D amp by using a lower inductance series coil in the output filter in a similar way to cranking up the bias of an AB amp further into class A.
.
PS. EMI emissions of the switching frequency are still very low with the smaller Wurth XXL coils at 3.5uH. Much lower than with Ferroxcube toroids wound to 8uH which leak quite a bit. Idle current is increased 25% with the lossy Wurths compared to the stock 20uH coils in the Shure 2X100 amp but the improvement in sonics is worth it.
__________________
Scott
  Reply With Quote
Old 24th August 2010, 01:46 PM   #2
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
With the help of a parametric equalizer you can discover that 0.5dB in the 10khz to 20khz range in fact represent a dramatic change on perceived sound.

Then, what are you tasting? Apples or oranges? You can't be sure if you eat both at the same time.

Non scientific methods yield non reliable results.

High ripple current in the inductor has technical advantages, though.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 24th August 2010, 03:20 PM   #3
diyAudio Member
 
darkfenriz's Avatar
 
Join Date: Jun 2004
Location: Warsaw
The nice thing in decreasing inductance is pushing crossover distortion up to higher powers without compromising efficiency much. This is very unlike class B, where xover dist happens at some miliWatts typically.

Wheather there is a peak in frequency response or not is a total blind guess anyway.

Or do you think an 8ohm speeker behaves anywhere close to 8ohm resistor at 10kHz to 20 kHz?
  Reply With Quote
Old 24th August 2010, 06:11 PM   #4
sendler is offline sendler  United States
diyAudio Member
 
Join Date: Aug 2004
Location: ny
Default EQ

Quote:
Originally Posted by Eva View Post
With the help of a parametric equalizer you can discover that 0.5dB in the 10khz to 20khz range in fact represent a dramatic change on perceived sound.

Then, what are you tasting? Apples or oranges? You can't be sure if you eat both at the same time.

Non scientific methods yield non reliable results.

High ripple current in the inductor has technical advantages, though.
My Audio source for the last 5 years is a direct out modified DEQ2496 so I am quite familiar with the sound of simple frequency response changes. The improvement that is realized with the smaller values of inductance is something different than a simple top octave lift. More information is revealed across the entire audio band to the point where this has become my best sounding amp by far. Better than my former favorite TDA2750/ TIP amp and much better than the ever popular 3886 gainclone.
__________________
Scott
  Reply With Quote
Old 30th August 2010, 03:54 AM   #5
diyAudio Member
 
Join Date: Apr 2010
Even if the DC value of load current is very high (compared to the inductor ripple current), wont the voltage slew to the opposite rail due to the body-diode conduction during the dead-time? Am I making a mistake.

I don't understand how large ripple current can make the sound more pleasing. Aren't we interested only in the low frequency sound say upto 20kHz as anything above that would almost not make any difference to a human ear?

Quote:
Originally Posted by sendler View Post
I decide to start a new thread on this topic as the links are dead on the old one and the old thread title would not draw in new viewers to the topic I want to discuss which is the huge improvement in sound quality of a Class D amp that comes from using output filter coils with values that are much smaller than what would otherwise be dictated by the universally accepted "textbook" slope.

.
This is an interesting concept and supports my listening tests where the lowest tolerable coil value (I go as low as 3.5uH on each leg of my TK2050 amps) for the output inductor yields the best sound quality. Everyone blamed the change in sonics on the increased peak at 65KHz lifting the frequency response in the audio band but it is only a .5db increase at 20KHz so I knew that there must be another mechanism at work for the large improvement. So by accepting an out of audio band peak and some loss of efficiency at idle, rather than arbitrarily designing a 2nd order filter with a textbook slope, we will improve the sonics of a class D amp by using a lower inductance series coil in the output filter in a similar way to cranking up the bias of an AB amp further into class A.
.
PS. EMI emissions of the switching frequency are still very low with the smaller Wurth XXL coils at 3.5uH. Much lower than with Ferroxcube toroids wound to 8uH which leak quite a bit. Idle current is increased 25% with the lossy Wurths compared to the stock 20uH coils in the Shure 2X100 amp but the improvement in sonics is worth it.
  Reply With Quote
Old 30th August 2010, 05:51 PM   #6
diyAudio Member
 
Join Date: Jul 2010
Quote:
Originally Posted by terminator12 View Post
Even if the DC value of load current is very high (compared to the inductor ripple current), wont the voltage slew to the opposite rail due to the body-diode conduction during the dead-time? Am I making a mistake.

I don't understand how large ripple current can make the sound more pleasing. Aren't we interested only in the low frequency sound say upto 20kHz as anything above that would almost not make any difference to a human ear?
Very interested by this topic. I tweaked seriously a TK2050 kit from SURE and got it to a very satisfying point, especially compared to what it sound out of the box with the low cost passive components used. All caps are no low ESR ones, all possible values where changed to film type instead of chemical type, output inductors are high current types from WURTH, the power supply is a toroidal transformer with 10000 F of filtering made with 1000F caps in parallel etc...I also removed the polarity protection input diodes in serie with the power supply.
I stayed with the recommend output filter value of 10 H for the inductor and 470 nF for the caps. The listening impression bothers me on SOME CDs. Let me explain : all drums have less impact, all cymbals sound too loud. Feels like there is really too much treble going out of these amps, on the other hand, despite their excessive level, they sound far much more metalic and therefore realistic than on other class AB amps from the market.
The theory show a theoretical flat response up to 20 Khz and even above with those filter values. But listening is in contradiction. Too brigh.

Anyone can help me to understand ? That really pisses me off as the amp is capabl of so much magic on the rest of the audio range, but that excess of treble really is a problem as it colours too much the sound ! How does lowering the inductor value help. It must be increasing the cutoff frequency and most probably adds noise that should be rejected by it, doesn't it ?
  Reply With Quote
Old 30th August 2010, 07:15 PM   #7
diyAudio Member
 
darkfenriz's Avatar
 
Join Date: Jun 2004
Location: Warsaw
Speaker's not a resistor, as simple as that.
  Reply With Quote
Old 31st August 2010, 01:35 PM   #8
diyAudio Member
 
Join Date: Jul 2010
Quote:
Originally Posted by darkfenriz View Post
Speaker's not a resistor, as simple as that.
I'm aware of this But the TK2050 has a feedback. So like any other amplifier, the output voltage is input x gain.
My Cabasse speakers have a very flat response at constant input voltage from 20Hz up to 20 Khz, whatever the impedance change involved.
So I'm wondering how I can hear so my treble when the sampe speakers don't generate them with other linear amplifiers with feedback just like this Tripath.
  Reply With Quote
Old 31st August 2010, 03:22 PM   #9
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
Quote:
But the TK2050 has a feedback.
Are you sure? After the filter?
  Reply With Quote
Old 31st August 2010, 03:27 PM   #10
diyAudio Member
 
darkfenriz's Avatar
 
Join Date: Jun 2004
Location: Warsaw
LC filter is outside of feedback loop, so quite possibly the filter amplifies above 10kHz on real speakers.
  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
How to bias mosfet in Class-A Anchan Solid State 18 7th August 2010 10:39 AM
High bias point. jnb Tubes / Valves 4 7th June 2007 03:39 PM
Class A speakers, as in BIAS Pan Multi-Way 3 10th February 2005 01:28 AM
Fets for high bias Zen4 Buhl Pass Labs 0 9th December 2002 10:39 AM
how high to bias? eLarson Pass Labs 2 10th November 2002 01:00 PM


New To Site? Need Help?

All times are GMT. The time now is 12:21 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