Forums

Whats Wrong with the DC300A answer

Started by Phil Allison October 25, 2019
Hi to all the terminally puzzled,

---------------------------------

THE *problem* with the Crown DC300A is simple, but virtually unknown in the industry. 

Any amplifier that has response to DC is similarly questionable as being suitable or safe to use driving loudspeakers or transformers.  

Such amplifiers produce a large DC component when driven into clipping with an unsymmetrical wave (eg speech or singing). Peaks of one polarity passing undistorted while peaks of the other become clipped off = DC offset created. 

With a high powered audio amp, the DC component can reach *25 volts* or more with speech or music.  Enough to bottom a woofer cone in its frame and push the voice coil almost completely out of the magnetic gap. Bad news for the speaker and sound quality - results in severe compression. 

Most amplifiers avoid the issue simply by having a low frequency pole in the feedback network, usually an 100uF or so electro cap to ground - so the amp exhibits only unity gain at DC. Asymmetrical clipping then only causes a small sub sonic signal to appear that tracks the clipping. 
 
See this page by my colleague Rod Elliot, who carefully simulated the situation after I explained the problem to him 16 years ago. 

https://sound-au.com/clipping.htm


FYI: the above plays havoc with speaker muting systems whose drive circuits detect the DC component and open the relay. 



....  Phil 

On Fri, 25 Oct 2019 00:32:27 -0700 (PDT), Phil Allison
<pallison49@gmail.com> wrote:

> >Hi to all the terminally puzzled, > >--------------------------------- > >THE *problem* with the Crown DC300A is simple, but virtually unknown in the industry. > >Any amplifier that has response to DC is similarly questionable as being suitable or safe to use driving loudspeakers or transformers. > >Such amplifiers produce a large DC component when driven into clipping with an unsymmetrical wave (eg speech or singing). Peaks of one polarity passing undistorted while peaks of the other become clipped off = DC offset created. > >With a high powered audio amp, the DC component can reach *25 volts* or more with speech or music. Enough to bottom a woofer cone in its frame and push the voice coil almost completely out of the magnetic gap. Bad news for the speaker and sound quality - results in severe compression. > >Most amplifiers avoid the issue simply by having a low frequency pole in the feedback network, usually an 100uF or so electro cap to ground - so the amp exhibits only unity gain at DC. Asymmetrical clipping then only causes a small sub sonic signal to appear that tracks the clipping. > >See this page by my colleague Rod Elliot, who carefully simulated the situation after I explained the problem to him 16 years ago. > >https://sound-au.com/clipping.htm >
That was the first answer you received. RL
> >FYI: the above plays havoc with speaker muting systems whose drive circuits detect the DC component and open the relay. > > > >.... Phil
On Friday, October 25, 2019 at 1:35:26 PM UTC-4, legg wrote:
> On Fri, 25 Oct 2019 00:32:27 -0700 (PDT), Phil Allison > <pallison49@gmail.com> wrote: > > > > >Hi to all the terminally puzzled, > > > >--------------------------------- > > > >THE *problem* with the Crown DC300A is simple, but virtually unknown in the industry. > > > >Any amplifier that has response to DC is similarly questionable as being suitable or safe to use driving loudspeakers or transformers. > > > >Such amplifiers produce a large DC component when driven into clipping with an unsymmetrical wave (eg speech or singing). Peaks of one polarity passing undistorted while peaks of the other become clipped off = DC offset created. > > > >With a high powered audio amp, the DC component can reach *25 volts* or more with speech or music. Enough to bottom a woofer cone in its frame and push the voice coil almost completely out of the magnetic gap. Bad news for the speaker and sound quality - results in severe compression. > > > >Most amplifiers avoid the issue simply by having a low frequency pole in the feedback network, usually an 100uF or so electro cap to ground - so the amp exhibits only unity gain at DC. Asymmetrical clipping then only causes a small sub sonic signal to appear that tracks the clipping. > > > >See this page by my colleague Rod Elliot, who carefully simulated the situation after I explained the problem to him 16 years ago. > > > >https://sound-au.com/clipping.htm > > > That was the first answer you received.
Not exactly. I referred to the DC coupling as causing DC to damage the speakers, but I didn't consider that this could be the result of clipping an asymmetric signal. Although it seems to me even if the amp were AC coupled at the input or between various stages, if the output stage suffered hard clipping from an asymmetric signal it could still drive DC into the speakers. This does not need to be a result of the entire amp being DC coupled, just the output stage. -- Rick C. - Get 1,000 miles of free Supercharging - Tesla referral code - https://ts.la/richard11209
Rick C wrote:

--------------

> > Not exactly. I referred to the DC coupling as causing DC to damage the speakers, but I didn't consider that this could be the result of clipping an asymmetric signal.
** Yep. Although it seems to me even if the amp were AC coupled at the input or between various stages, if the output stage suffered hard clipping from an asymmetric signal it could still drive DC into the speakers. This does not need to be a result of the entire amp being DC coupled, just the output stage.
>
** Nope. Any low frequency pole inside the amp is enough to stop the effect. Modern Class D amps without a DC servo loop have the same issue. It's an elephant in the room kinda problem. Win should put a chapter in his AofE book about it - and credit me please. ..... Phil
> -- > > Rick C. > > - Get 1,000 miles of free Supercharging > - Tesla referral code - https://ts.la/richard11209
fredag den 25. oktober 2019 kl. 22.16.13 UTC+2 skrev Phil Allison:
> Rick C wrote: > > -------------- > > > > > Not exactly. I referred to the DC coupling as causing DC to damage the speakers, but I didn't consider that this could be the result of clipping an asymmetric signal. > > ** Yep. > > > Although it seems to me even if the amp were AC coupled at the input or between various stages, if the output stage suffered hard clipping from an asymmetric signal it could still drive DC into the speakers. This does not need to be a result of the entire amp being DC coupled, just the output stage. > > > > ** Nope. > > Any low frequency pole inside the amp is enough to stop the effect. > > Modern Class D amps without a DC servo loop have the same issue. >
isn't potentially worse with some class-D? with a half bridge on a split supply and the inductor in the out filter it can pump energy from one rail to the other, potentially increasing the voltage to unsafe levels