Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.tech
Randy Yates Randy Yates is offline
external usenet poster
 
Posts: 839
Default Class D Output Filtering

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?
--
Randy Yates % "My Shangri-la has gone away, fading like
Digital Signal Labs % the Beatles on 'Hey Jude'"
%
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO
  #2   Report Post  
Posted to rec.audio.tech
Don Pearce[_3_] Don Pearce[_3_] is offline
external usenet poster
 
Posts: 2,417
Default Class D Output Filtering

On Wed, 29 Dec 2010 12:05:12 -0500, Randy Yates
wrote:

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?


Because there are EMI regulations that seriously frown upon high power
transmitters that broadcast a signal every 40kHz from DC to light.

d
  #3   Report Post  
Posted to rec.audio.tech
Randy Yates Randy Yates is offline
external usenet poster
 
Posts: 839
Default Class D Output Filtering

On 12/29/2010 12:44 PM, Don Pearce wrote:
On Wed, 29 Dec 2010 12:05:12 -0500, Randy
wrote:

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?


Because there are EMI regulations that seriously frown upon high power
transmitters that broadcast a signal every 40kHz from DC to light.

d


Apart from EMI, is there a reason?
--
Randy Yates % "My Shangri-la has gone away, fading like
Digital Signal Labs % the Beatles on 'Hey Jude'"
%
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO
  #4   Report Post  
Posted to rec.audio.tech
Don Pearce[_3_] Don Pearce[_3_] is offline
external usenet poster
 
Posts: 2,417
Default Class D Output Filtering

On Wed, 29 Dec 2010 12:45:49 -0500, Randy Yates
wrote:

On 12/29/2010 12:44 PM, Don Pearce wrote:
On Wed, 29 Dec 2010 12:05:12 -0500, Randy
wrote:

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?


Because there are EMI regulations that seriously frown upon high power
transmitters that broadcast a signal every 40kHz from DC to light.

d


Apart from EMI, is there a reason?


Yes. That high frequency square wave would dissipate maximum power
into any tweeter connected to it - even without any audio signal
present. It has to be suppressed.

d
  #5   Report Post  
Posted to rec.audio.tech
Randy Yates Randy Yates is offline
external usenet poster
 
Posts: 839
Default Class D Output Filtering

On 12/29/2010 12:49 PM, Don Pearce wrote:
On Wed, 29 Dec 2010 12:45:49 -0500, Randy
wrote:

On 12/29/2010 12:44 PM, Don Pearce wrote:
On Wed, 29 Dec 2010 12:05:12 -0500, Randy
wrote:

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?

Because there are EMI regulations that seriously frown upon high power
transmitters that broadcast a signal every 40kHz from DC to light.

d


Apart from EMI, is there a reason?


Yes. That high frequency square wave would dissipate maximum power
into any tweeter connected to it - even without any audio signal
present. It has to be suppressed.

d


Are you saying that a tweeter will have a significant mechanical
response at some 2 MHz (e.g., the SSM2301)? Otherwise, the current
and voltage will be 90 degrees out of phase and no power (or very
little) will be dissipated.
--
Randy Yates % "My Shangri-la has gone away, fading like
Digital Signal Labs % the Beatles on 'Hey Jude'"
%
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO


  #6   Report Post  
Posted to rec.audio.tech
Randy Yates Randy Yates is offline
external usenet poster
 
Posts: 839
Default Class D Output Filtering

On 12/29/2010 01:08 PM, Dick Pierce wrote:
Randy Yates wrote:
Yes. That high frequency square wave would dissipate maximum power
into any tweeter connected to it - even without any audio signal
present. It has to be suppressed.


Are you saying that a tweeter will have a significant mechanical
response at some 2 MHz (e.g., the SSM2301)? Otherwise, the current
and voltage will be 90 degrees out of phase and no power (or very
little) will be dissipated.


No, it won't. It will be more like 45 degrees for quite some
way above 20 kHz.

Another issue is whether distributed capacitance can have an
effect at these frequencies. I'm not usre I have any better
an answer for that than anyone else, but my suspicion is the
effect is probably not significant.


Thanks Don/Dick for this information.
--
Randy Yates % "My Shangri-la has gone away, fading like
Digital Signal Labs % the Beatles on 'Hey Jude'"
%
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO
  #7   Report Post  
Posted to rec.audio.tech
Kevin McMurtrie[_3_] Kevin McMurtrie[_3_] is offline
external usenet poster
 
Posts: 14
Default Class D Output Filtering

In article ,
Randy Yates wrote:

An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?


Self-contained systems, like cellphones, don't always have filtering on
Class D amps. In those cases the speaker is guaranteed to have short
wires, high impedance, low capacitance, and not be paired with any kind
tonal or impedance compensation.
--
I will not see posts or email from Google because I must filter them as spam
  #8   Report Post  
Posted to rec.audio.tech
g g is offline
external usenet poster
 
Posts: 111
Default Class D Output Filtering

On Dec 29 2010, 12:05*pm, Randy Yates wrote:
An open-circuit, class D amplifier output has extremely sharp
edges due to the switching of the bridge. The expectation is that,
somehow, the output is averaged over at least one cycle of the switching
period to form a continuous, analog voltage.

If this output is connected to a speaker with a mechanical lowpass
characteristic, then the speaker's mass will effectively average
the "digital" output into a continous position.

Thus my question is, when connecting a class D output to a speaker,
why do we need filtering at all? That is, why wouldn't the mechanical
lowpass filtering of the speaker be sufficient?
--
Randy Yates * * * * * * * * * * *% "My Shangri-la has gone away, fading like
Digital Signal Labs * * * * * * *% *the Beatles on 'Hey Jude'"
* * *%http://www.digitalsignallabs.com% 'Shangri-La', *A New World Record*, ELO


I was reading about the NXP mono amp on a FTM flyer. Suppose the
output frequencies
mixed and caused byproducts ?? They would need some delays.

greg
Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Class D power amplifier---what about class-D CD? New_idea Pro Audio 91 October 16th 09 03:38 AM
proper line filtering for switchmode PSU+class-d amps? David Gravereaux Pro Audio 6 June 28th 09 10:02 PM
Any blind listening tests on Class A vs Class B amps? Don Pearce Tech 18 October 28th 05 05:44 PM
Class D full range/Class T w/Tripath Ivan Lopez Car Audio 11 August 16th 04 02:28 PM
Neve Class A (1272) Output Question Crumb Pro Audio 2 July 7th 04 05:49 PM


All times are GMT +1. The time now is 04:00 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"