Phase Shift or Polarity Inversion?? - 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 17th April 2013, 09:54 PM   #1
diyAudio Member
 
Join Date: Jan 2013
Default Phase Shift or Polarity Inversion??

Hi all,

In Class-D amplifier, the transistors are in Push-Pull configuration. To drive the transistors using the PWM signal, the signal given at either transistor is inverted of the other signal.

The question is:
From most of the references I read, why they normally used 180 degree phase shift?? If PWM phase is shifted, it is not exactly inverted. Both signals which drive the transistors are not inverted to each other, but only delay.

Could somebody explain this to me. TQ.

Last edited by pYrana13; 17th April 2013 at 10:04 PM.
  Reply With Quote
Old 17th April 2013, 10:08 PM   #2
diyAudio Member
 
sofaspud's Avatar
 
Join Date: Nov 2005
Location: San Antonio
If phase is shifted 180, it is exactly inverted.
__________________
It is error only, and not truth, that shrinks from enquiry. - Thomas Paine
  Reply With Quote
Old 17th April 2013, 10:16 PM   #3
diyAudio Member
 
Join Date: May 2011
Location: Silicon Valley
Quote:
Originally Posted by sofaspud View Post
If phase is shifted 180, it is exactly inverted.
However some waveforms are asymmetric; for example, a sawtooth wave is asymmetric. There is no phase shifted sawtooth wave that is equal to an upside-down sawtooth wave.

The output of a trumpet is another asymmetric waveform; see Fig 3.1 of this set of lecture notes. Phase shifting a trumpet wave doesn't give an upside-down trumpet wave.
  Reply With Quote
Old 17th April 2013, 10:31 PM   #4
diyAudio Member
 
Join Date: Jan 2013
Then, why some of the design of Class-D power amplifier use rat-race circuit or any other phase shifter hybrid to invert the PWM input signal, which is clearly not asymmetric?
  Reply With Quote
Old 17th April 2013, 10:56 PM   #5
diyAudio Member
 
vacuphile's Avatar
 
Join Date: Apr 2011
Location: Seaside
Quote:
Originally Posted by transistormarkj View Post
However some waveforms are asymmetric; for example, a sawtooth wave is asymmetric. There is no phase shifted sawtooth wave that is equal to an upside-down sawtooth wave.

The output of a trumpet is another asymmetric waveform; see Fig 3.1 of this set of lecture notes. Phase shifting a trumpet wave doesn't give an upside-down trumpet wave.
I may not understand you correctly, but if I do, I can't agree

An inverted sawtooth, e.g. by using an opamp in the inverting configuration, is exactly that. And the same goes for the output of a trumpet. It's the same signal, upside down.
  Reply With Quote
Old 17th April 2013, 11:21 PM   #6
GoatGuy is online now GoatGuy  United States
diyAudio Member
 
Join Date: Dec 2012
Location: SF Bay Area
Stop thinking of it as a 180 phase shift. It is NOT a 180 phase shift, because it is NOT frequency-dependent, AND ESPECIALLY, it is not time-lagging!!! (which is what a -180 phase shift would have to have, and be a shift in the Z-plane domain)

Textbooks and popular literature very often call the waveform inversion "180 phase shift", but in so doing, they perpetuate the false reality.

We used to say that a "complimentary waveform" is made/created/used/employed. A waveform that is the "mirror image" of the positive waveform, with all wiggles that go up, reflected to go down, and vice-versa.

So, if you think about it like that, an "inverted sawtooth" or even harder to visualize, an "inverted triangle waveform" ... easily has a mirror of itself in the time-domain.

Indeed, ideally, W(t)inverted = -W(t), showing simple sign inversion. In the case of some amplifiers, the "phase splitter" (again, terribly misnamed, which should be called "symmetric signal inverter") not only creates the mirror-image, but also adds a bias to each (which might not be even the same!]. That's what stage-to-stage coupling capacitors are for: to remove the persistent DC/bias component, leaving Inverted = -Uninvderted;

GoatGuy
  Reply With Quote
Old 18th April 2013, 10:25 AM   #7
diyAudio Member
 
Join Date: Apr 2010
1.
Inverted non-sinusodioal periodic waveform generated by an (ideal) opamp inverter: ALL frequencies (fundamental as well as harmonics are 180 out of phase). wave looks mirrored upside down.

2.

time shifted non-sinusodial periodic waveform (shifted by a half period): ONLY the fundamental is 180 out of phase, the harmonics have different phase shifts.
  Reply With Quote
Old 18th April 2013, 12:49 PM   #8
diyAudio Member
 
Join Date: Jan 2013
Then, how could I drive transistors in push-pull configuration with PWM signal? The duty cycle of the pulse in the PWM signal keeps on changing.

Do I need to modulate the signal first with sinusoidal carrier frequency, and only then the phase shift will work like an inverter. Is it the right method to drive class D amplifier or not? I can use the modulated signal and its inversion as shown below to drive both transistors in push-pull?
Attached Images
File Type: jpg PWM and modulated signal.JPG (204.1 KB, 79 views)
  Reply With Quote
Old 18th April 2013, 01:04 PM   #9
diyAudio Member
 
Join Date: Apr 2010
In a half bridge, both transistors must never be ON at the same time, but it is desirable to have dead time as low as possible. So, the goal is to have almost always exactly one of the 2 FETs ON. This can be accomplished by generating a PWM signal for upper FET, then logically inverting this signal and feeding it to lower FET.
  Reply With Quote
Old 28th April 2013, 08:23 PM   #10
diyAudio Member
 
Join Date: Apr 2010
"1.
Inverted non-sinusodioal periodic waveform generated by an (ideal) opamp inverter: ALL frequencies (fundamental as well as harmonics are 180 out of phase). wave looks mirrored upside down.

2.
time shifted non-sinusodial periodic waveform (shifted by a half period): ONLY the fundamental is 180 out of phase, the harmonics have different phase shifts."

-> Since no one corrected my posting above, I ll do it myself: The truth is:

1.
Inverted non-sinusodioal periodic waveform generated by an (ideal) opamp inverter: ALL frequencies (fundamental as well as harmonics) are 180 out of phase in relation to the non-inverted signal. Wave looks mirrored upside down. OK.

2.
time shifted non-sinusodial periodic waveform (shifted by a half period): ONLY the fundamental AS WELL AS THE UNEVEN / ODD HARMONICS are 180 out of phase in relation to the non-shifted signal, the EVEN harmonics will see zero (360 or 360 * 2 or 360 * 3...) phase shift .
  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
within passband phase inversion speakerdoctor Multi-Way 4 5th July 2010 03:57 PM
Passive phase inversion? TequilaKez Parts 4 6th October 2008 01:38 AM
Avoid polarity inversion MaXiZ Everything Else 7 3rd March 2004 05:55 AM
Polarity Inversion circuit ashok Everything Else 17 5th April 2003 11:32 AM


New To Site? Need Help?

All times are GMT. The time now is 04:12 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