low phase delay through filterless 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 30th June 2014, 02:31 PM   #1
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Default low phase delay through filterless Class D?

I don't know much about Class D other than the basic principles, so I'm mostly ignorant about current 'chips' and advances.

I'd like to know how Class D amplifiers compare with, say, Class A amplifiers, in terms of total phase delay through the amplifier - are they better or worse or about the same. About the same is just fine but I had been under the impression that the filter needed on the output of a Class D amplifier produced unwelcome phase delays and that they might be speaker dependent too. So recently I came across these 'spread spectrum' options for Class D amplifiers which are 'filterless' - I assume it means they radiate all over the place but in such a way that it isn't objectionable or is this technology restricted to boom-boxes with short speaker cables ??

So without any output filters, do these 'filterless' amplifiers have good phase response over audio range ???
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 30th June 2014, 06:39 PM   #2
xrk971 is offline xrk971  United States
diyAudio Member
 
xrk971's Avatar
 
Join Date: Oct 2012
Location: Virginia
The filter bank on a properly designed class D amp, even though it has an inductor and capacitors in the output path between the IC and the speaker, should not affect the phase of the audio signal too much. The phase altering effects of the speaker itself, or cross-over, or the room reflections are much more significant. The inductor and capacitor together form a high pass circuit designed to heavily attenuate the switching frequency.

Take for example the Texas Instruments TPA3116D2, the current class-D wunderamp: it has a 10 uH coil coupled with a 680 nF capacitor low pass filter. It also has a 1 nF cap directly tied to ground plus a 3.3 ohm R and 10 nF cap in series which then tie the output to ground for an RF/EMI filter.

If you look at the AC transfer characteristics of these items in a circuit simulator program (TINA for example, a nice fully featured free SPICE program from TI), you can see the amplitude and phase that these coils and caps induce on the AC signal going from the amp to the speaker.

Here I have a circuit representing the signal path from the amp IC output through the recommended filter and driving an 8 ohm speaker.

The resulting transfer function is pretty flat but has a +1dB gain at 20 kHz and -10 deg phase angle at 20 kHz, and otherwise flat. The filter is designed to block the 400 kHz or higher switching signal. The attenuation is -34dB at 400 kHz.

So unless you are worried about a 10 deg phase difference at 20 kHz, I would say that the class D topology is actually very phase linear. Running without the coil is not going to buy you much other than spraying the world with unintended RF noise.
Attached Images
File Type: png Class-D-Filter-Schematic.png (18.0 KB, 92 views)
File Type: png Class-D-Filter-Bode.png (41.5 KB, 90 views)
  Reply With Quote
Old 30th June 2014, 06:48 PM   #3
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Wow, I didn't realize they were so good. In fact it looks as if a gnf loop could be employed up to quite a reasonable frequency if desired.

(I've never heard a hi-fi Class D amplifier yet, only what they put into consumer devices like my iMac)
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 1st July 2014, 07:36 AM   #4
diyAudio Member
 
Join Date: Sep 2004
Location: Napier, Hawkes Bay
Quote:
Originally Posted by xrk971 View Post
Running without the coil is not going to buy you much other than spraying the world with unintended RF noise.
Filterless = approx 80% reduction in IMD. There's a good TI white paper on this.

Cheers,
Mike
  Reply With Quote
Old 1st July 2014, 09:18 AM   #5
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Filterless Class D gets around EMC regulations by spreading more RF muck over a much bigger area, so each piece of land gets a thinner layer of muck. Personally, I would ban them by rewriting the EMC regulations to have a lower threshold for wideband noise transmitters. Before too long the HF and VHF spectrum will become unusable in most towns and cities - there are already concerns about the effect on aircraft comms when overflying them!
  Reply With Quote
Old 1st July 2014, 09:24 AM   #6
diyAudio Member
 
Join Date: Apr 2010
Location: Brisbane
Quote:
Originally Posted by Bigun View Post
I don't know much about Class D other than the basic principles, so I'm mostly ignorant about current 'chips' and advances.

I'd like to know how Class D amplifiers compare with, say, Class A amplifiers, in terms of total phase delay through the amplifier - are they better or worse or about the same. About the same is just fine but I had been under the impression that the filter needed on the output of a Class D amplifier produced unwelcome phase delays and that they might be speaker dependent too. So recently I came across these 'spread spectrum' options for Class D amplifiers which are 'filterless' - I assume it means they radiate all over the place but in such a way that it isn't objectionable or is this technology restricted to boom-boxes with short speaker cables ??

So without any output filters, do these 'filterless' amplifiers have good phase response over audio range ???
Any delay won't be audible as long it's linear and constant and not distorted over the audible band ( in theory ) as long as the original signal is not present ( in real time ) in the amplified signal .
An example is a high quality audio signal transmitted by Bluetooth . The receiver receives the signal delayed , but there's little or no distortion , The delay will be visible if you use a Bluetooth wireless headphone on your TV But the sound quality is as good .

Cheers ,

Rens

Not really on topic , just a teaser but it makes sense ?
__________________
If you measure, you know; if you guess, you don't
  Reply With Quote
Old 1st July 2014, 01:35 PM   #7
diyAudio Member
 
doctormord's Avatar
 
Join Date: Oct 2007
Quote:
Originally Posted by xrk971 View Post
Take for example the Texas Instruments TPA3116D2, the current class-D wunderamp: it has a 10 uH coil coupled with a 680 nF capacitor low pass filter. It also has a 1 nF cap directly tied to ground plus a 3.3 ohm R and 10 nF cap in series which then tie the output to ground for an RF/EMI filter.
Filters should not be considered without load-impedance. Performance of filter is a function of frequency as the (real) loads impedance/parasitics will change with frequency as well.
__________________
http://www.360customs.de
  Reply With Quote
Old 1st July 2014, 02:26 PM   #8
xrk971 is offline xrk971  United States
diyAudio Member
 
xrk971's Avatar
 
Join Date: Oct 2012
Location: Virginia
Quote:
Originally Posted by doctormord View Post
Filters should not be considered without load-impedance. Performance of filter is a function of frequency as the (real) loads impedance/parasitics will change with frequency as well.
Yes, I thought of this, but I am not sure if I can put in a real dynamic driver impedance simulator into a SPICE circuit? I suppose adding at least Le of drive make some difference but really best way is to measure electrical phase in the actual system with a woofer tester or the like.
  Reply With Quote
Old 1st July 2014, 03:32 PM   #9
diyAudio Member
 
Join Date: Apr 2010
Do not forget also, if the feedback is from after the LC, the behavior gets improved (like UCD)
  Reply With Quote
Old 1st July 2014, 06:05 PM   #10
gmarsh is offline gmarsh  Canada
diyAudio Member
 
gmarsh's Avatar
 
Join Date: Apr 2004
Location: Halifax, NS, Canada
Quote:
Originally Posted by DF96 View Post
Filterless Class D gets around EMC regulations by spreading more RF muck over a much bigger area, so each piece of land gets a thinner layer of muck. Personally, I would ban them by rewriting the EMC regulations to have a lower threshold for wideband noise transmitters. Before too long the HF and VHF spectrum will become unusable in most towns and cities - there are already concerns about the effect on aircraft comms when overflying them!
A ham operator I know, in his 70's and in the hobby for decades, laments the days when the HF spectrum wasn't full of crap and you could pull in people from all over the world with a cheap radio. Personally I find I can't pick up AM radio anymore - I can pick it up clear as day on the HF rig with my outdoor antenna, but on a consumer radio it's unlistenable because of the crap everything puts out now. I refuse to design anything filterless, because there's enough crap ******* over the RF spectrum as it is, and the world certainly doesn't need more.

If audiophilia is the goal, make your class D amp run at a higher switching frequency and up the output LC corner frequency to minimize the peaking/phase shift. Or use a UcD topology which puts the output LC in the loop - set the LC and feedback for a higher switching frequency if you really insist on chasing numbers. You'll need smaller FETs that can switch faster and/or beefier gate drive, and you'll have more switching/conduction loss which will make the amp less efficient, but you'll get closer to the "flat to whatever hz" that you're chasing.

Just don't be an ******* and make a mess of the spectrum just because you want your amp to have that extra 'sparkle' or whatever.
  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
Filterless Class D gpapag Class D 37 21st October 2013 09:47 PM
Filterless Class D Switching Distortion carlturner Class D 11 21st April 2012 04:26 PM
Filterless class-d fzaad Class D 4 29th December 2010 10:27 PM
+/-12V 5A, 100W filterless class D newvirus2008 Class D 20 13th September 2010 02:04 PM
Filterless subwoofer Class-D amp Workhorse Class D 11 13th March 2009 08:11 PM


New To Site? Need Help?

All times are GMT. The time now is 09:40 AM.


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