Re: Parallel stereo = more output?



meow2222@xxxxxxxxx wrote:

> Fred wrote:
>> GregS wrote:
>> > In article <o1vst19g0dnevqmq5tute0u0e7e934iblq@xxxxxxx>, Steven Goring
>> > <stevegoring@xxxxxxxxx> wrote:
>> > >I have a TDA 2009 audio amp chip that does 10W on each channel.
>> > >
>> > >Is there any reason why I cannot connect the output of both channels
>> > >in parallel to produce a total 20W output?
>> > >
>> > >Obviously, the same signal would be used to drive both inputs, and amp
>> > >output would be mono.
>> > >
>> > >I do not want to use a bridge configuration for design reasons.
>> > >
>> > >Thank you for any advice.
>> > >
>> > >Steven Goring
>> >
>> > I didn't look up the specs, but generally you can't do this.
>> > They have to have identical poles and stuff, but by
>> > adding a little lossy resistance on each output, they
>> > may be stable enough to opperated with the resistor legs
>> > tied together to form the output.
>> >
>> > greg
>>
>> The output of one channel would see, as a load, the very low output
>> impedance of the other channel.
> 
> Only if the 2 amps try to fight each other at some point. Take care
> that the response across the spectrum is the same with both amps and
> things are ok.
> 
> 
>> And when you add series resistors you
>> lose the advantage you are trying to gain by introducing new losses.
> 
> power doubled then -10%, so no, you get 80% more power.
> 
> 
>> Not a good approach.
> 
> It does increase output impedance, and the careless may be tempted to
> do it with unmatched circuits. And if any minor circuit fault changes
> the closed loop characteristics you can end up with 2 chips trying to
> cook each other, hence a few more failures. It works though.
> 
> 
> NT
> 
The original device is two (stereo) amplifiers on one chip.  How much more
well matched do you think you are going to get?  OP may get 41% more usable
voltage but not 41% more usable current; and power dissipation remains as a
limiting factor.
-- 
JosephKK
Gegen dummheit kampfen Die Gotter Selbst, vergebens.  
--Shiller
.


Quantcast