Question about THD in LM1876

[laejf]

Hi,

I'm making an audio amplifier using the LM1876. In the datasheet the output power over input voltage is plotted as seen below. I'm planning on using +/-16V as supply and my question is about the two levels of THD in the graph. How do i know which THD I will have, so I know what my output power will be?

[Brianf]

How do i know which THD I will have, so I know what my output power will be?

You are thinking about this the wrong way around. You pick the distortion that you can accept and that determines the power you get.

So with your 16V rails, if you accept 10% THD you can have 20W, if that THD is too high, and you want to keep it at 1% then you can have 15W.

[laejf]

How do i know which THD I will have, so I know what my output power will be?

You are thinking about this the wrong way around. You pick the distortion that you can accept and that determines the power you get.

So with your 16V rails, if you accept 10% THD you can have 20W, if that THD is too high, and you want to keep it at 1% then you can have 15W.

But if i have +/-16V and crank the volume all the way up, I would only have one true output power, so would it be 15 or 20W?

[Gyro]

It is dependent on how far you crank the volume control - that is a useless metric because volume knob setting doesn't tell you anything about the actual signal level that you're putting into the chip. If you crank it all the way up, you will most likely cause the amplifier to clip (try to put out far more than either power level).

The two lines on the graph are just indications of the distortion levels you might expect at different power levels and supply voltages. In reality, there are an infinite number of lines running from 0% THD to 100% (clipped square waves out).

[Brianf]

But if i have +/-16V and crank the volume all the way up, I would only have one true output power, so would it be 15 or 20W?

Read up on 'clipping'.

[Kleinstein]

The output power rating are given with different standards. Some us 1% THT, some use 10 %THD and others (PMPO) don't care about THD at all.
For the full amplifier circuit there is than also the effect of cooling and the size of the filter capacitors, which causes the different ratings for sine and music (peak power).
So there are different defintions for rated power and thus usually different numbers for a single amplifier.

The actual output power level depends on the signal / music and the set volume. Depending on the power the THD will change. So for every supply voltage there would be a THD vs power curve - usually this is also found in the datasheet for some common supply case.
With music the actuall power is usually much lower most of the time and only some peaks may actually need significant power. So even the 10%THD power rating makes sense as this would be only for short bursts, were distortion would actually happen.

[mikerj]

But if i have +/-16V and crank the volume all the way up, I would only have one true output power, so would it be 15 or 20W?

You are still thinking backwards, the harder you drive the amplifier towards it's limit the more distortion you will get.  From this you should be able to see that there is no "one true power", it depends entirely on how much distortion is acceptable to you.  This is why any serious maximum power specification will always be paired with a distortion value, since it pretty meaningless without.

[RoGeorge]

But if i have +/-16V and crank the volume all the way up

If you crank the volume up until you get 15W, then you'll have 1% distortions.
If you crank it up even more, then you'll get 20W but with 10% distortions.

[laejf]

I see what I got wrong now, thank you for the great answers!

[gbaddeley]

The graph shows power into a 4 ohm resistor. Maximum Power into a loudspeaker will be different, and dependent on frequency and waveform.