power audio op amp amplifier capacitor

Thread Starter

pager48

Joined Nov 25, 2018
160
Is it ok to use an 0805 MLCC cap for these caps?

Are wire wound aluminum shell resistors ok for R4 power resistor?
 
Last edited:

wayneh

Joined Sep 9, 2010
16,102
I'd use an electrolytic for the 2.2µF cap, just because it would be less expensive than other options.
 

Audioguru

Joined Dec 20, 2007
11,251
A 4 ohm resistor produces NO SOUND!
Use a 4 ohm speaker instead.

The 2.2uF input capacitor value is crazy when it feeds the 100k input bias resistor as shown on the datasheet. The other 2.2uF capacitor feeds the 4.7k resistor properly.
2.2uF into 100k passes earthquake frequencies below 1 Hz. Use a 150nF film capacitor for a cutoff at 11Hz.
 

Audioguru

Joined Dec 20, 2007
11,251
A simple calculation shows that the low cutoff frequency is lower when the 2200uF output capacitor feeds an 8 ohm speaker (-3dB at 20Hz) but the 2200uF feeding 4 ohms has a -3dB cutoff at 40Hz. The datasheet shows that the output power into 8 ohms is 63% of the power into 4 ohms.
Since the output power is less into 8 ohms then the heating is also less but the datasheet does not show it.

I wonder why one engineer at ST Micro used the 2.2uF input capacitor feeding a 100k input bias resistor to pass earthquake frequencies and maybe another engineer shows a 1uF input capacitor feeding a 22k resistor with a cutoff frequency that is 8 times higher??
 

Audioguru

Joined Dec 20, 2007
11,251
The TDA2030A from ST Micro of France-Italy is exactly the same as the TDA2030A from Unisonic of Taiwan. The circuits in their datasheets are also exactly the same.

Frequencies below 20Hz are actually vibrations that you cannot hear and speakers cannot play. The 2.2uF capacitor feeding the 100k resistor passes vibrations down to 0.9Hz (earthquakes and heartbeats) but the feedback and output capacitors cannot pass such low frequencies.
The 2.2uf capacitor feeding the 100k resistor will take 6 seconds to charge and while it is charging the output of the amplifier will be extremely distorted.

Why are you simply heating an 8 ohm resistor with this amplifier? Why doesn't your amplifier drive a woofer speaker?
 

AnalogKid

Joined Aug 1, 2013
8,100
AG - the post #1 schematic shows a 4 ohm speaker. There is no 8 ohm anything anywhere.

TS - R4 (1 ohm) does not have to be a power resistor. It is part of a Zobel network to prevent high frequency oscillation, something all audio power amp chips are prone to. The corner frequency for this network is over 700 kHz.


ak
 

Audioguru

Joined Dec 20, 2007
11,251
Whats a safe power rating for R4? Granted 14w of power output to load.
220nF in series with 1 ohm has half the power (7W) at 727kHz, it has 1.7W at 364kHz, it has 0.4W at 182kHz, it has 0.1W at 91kHz, and it has almost no power at audio frequencies. It simply helps to prevent oscillation at very high frequencies. Use a 1/2W resistor.

Leaving in the 2.2uF cap for low frequency would increase its bandwidth and increase 2-40hz amplification. Its correct those sub 20hz signals are inaudible but they can surely be sensed.
No, the other capacitor values prevent the amplifier from producing extremely low frequencies unless their values are also ridiculously huge.
What speakers can produce sub-audio frequencies anyway?
Most 12" and 15" Dayton Audio woofers from Parts Express in a huge enclosure barely produce 30Hz and not much at lower frequencies.

The RL resistor is a placeholder for an 8ohm primary coil in a loudspeaker it surely does not resemble a resistor symbol.[/QUOTE]
A loudspeaker has resistance and inductance. I think it has some capacitance on one side of its resonance.
 

Audioguru

Joined Dec 20, 2007
11,251
Many large sub-woofers for home theaters drop off at about 30Hz. Smaller (Bose) ones drop off at 100Hz. My home audio system has 8" woofers that drop off at 50Hz but my bass-boost circuit creates a flat response to about 30Hz. I have a new 18" high power woofer (900W) that I have never used. In a huge ported enclosure it can produce 25Hz.

A ceramic and an electrolytic capacitor changes its value when its voltage is changed. The signal changes its voltage which produces value changes and even harmonics distortion at low frequencies. A film capacitor does not. I always use inexpensive polyester (mylar) capacitors since the expensive polypropylene ones sound the same. A ceramic capacitor is also a microphone that produces distortion and acoustical feedback squealing when the sounds vibrate it.
 

danadak

Joined Mar 10, 2018
3,573
C7 is part of a comp network as previously stated. Using an MLCC presents
much lower esr, hence comp network affected.

Shortcut to test is to look at phase margin in time domain. Put the MLCC
in circuit, feed the amp a small signal square wave, say 100 mV or so, and look
at time domain response on scope.



Reduced phase margin = more ringing.


Regards, Dana.
 
Last edited:

Audioguru

Joined Dec 20, 2007
11,251
Which capacitors makes the largest overall difference in sound quality?

The 2200uF capacitor would have to be large to be film if it makes an impact on quality.
That is why most high quality audio amplifiers use a dual polarity supply then the output capacitor is not needed.
 

Audioguru

Joined Dec 20, 2007
11,251
The output capacitor of a single supply 70W amplifier is not a tiny little 100uF surface mount thingy with a high frequency 10Vp-p voltage swing, it is a monster 2200uF to 4700uF with a 67Vp-p voltage swing at a very low frequency.
 

danadak

Joined Mar 10, 2018
3,573
Yes, agree, so where is the data from these manufacturers and how significant is
the modulation of C with V. I am batting zero to find it so far. I am not questioning
others think there is a strong (big enough effect one can discern in audio), I am just
looking for credible data supporting the effect is significant or its a mice nuts effect.

I would think for larger caps the problem, if there, scales with physical size.

And what is the physical mechanism causing this.....the TDK paper seems call out
an effect, no physical explanation.


Regards, Dana.
 

Audioguru

Joined Dec 20, 2007
11,251
I have seen manufacturers graphs of an amplifier IC that has the heating of its output transistors producing distortion its nearby input transistors.
But I have never seen a test of amplifier distortion with and without an electrolytic output capacitor.
 

danadak

Joined Mar 10, 2018
3,573
It would seem simple to drive a cap large signal with a sine, and
look at loaded vs unloaded spectrum for distortion.

At various voltages of course.

Regards, Dana.
 

Audioguru

Joined Dec 20, 2007
11,251
Unloaded, the amplifier and capacitor will produce almost no distortion. When loaded the capacitor reduces the levels of low frequencies and when its capacitance changes with the momentary varying voltage across the capacitor then even harmonics are produced which is distortion.

Odd harmonics would be produced if the capacitor had AC only with no DC across it.
 
Top