I must have seen all of these graphs and subconsciously decided to back away slowly :). But force it I must!
Quiescent Supply Current vs Supply Voltage
X axis is supply voltage (v) and Y axis is Quiescent Supply Current. It seems that the larger the power supply you use, the more “parasitic" current that the chip draws in a powered but non-active state.
Note that this graph shows the range for the 4-12V needs but does not go as far up as the 18v that the LM386N-4 can handle.
Power Supply Rejection Ratio vs Frequency
So a quick reminder as to what power supply rejection is - It is the measure of noise from a power supply that the chip can reject. So given a supply voltage of 6V, frequency of 1kHz, a bypass capacitor of 10uf (goes between PIN 7 and ground), and pins 1 and 8 open, you’d end up able to reject 50db of noise.
It seems that this graph shows you the amount of power supply rejection given certain capacitor sizes with the power supply voltage being 6v and the Av (voltage gain) of 26dB.
Peak to Peak Output Voltage Swing vs Supply Voltage
So this graph shows - at the very basic level that as supply voltage increases, output voltage increases.* It’s separated by reaction to different load resistances - for example a speaker that would be 4ohm, 8ohm, or 16 ohms. As you have more load resistance, more power is dropped over the load so you get a more linear response.
I asked Lancelot why they would say “Volts Peak to Peak” since voltage is amplitude in a waveform. And he said that because it’s AC current, the wave goes above and below a certain point. So peak to peak means from + peak to - peak (even if the center isn’t at zero volts). Example below shows a peak to peak voltage of 10v.
I imagine if you used P=IE and E=IR, you’d be able to find the values for the output powers that are listed in the datasheet.
Next! More graphs :)
*EDIT - @mgburr kindly corrected some of my erroneous interpretations:
The max swing references the maximum amount of output capable, not specifically “as supply voltage increases, output voltage increases”. It just means that there is a wider range of swing available, along with what you are using as an output load.
- The example shows with the 4ohm load, you taper off about 3.5Vpp with a 10V supply voltage. At 8ohm it tapers off about 6.5V with a 11V supply. The 16ohm shows enough load resistance to give it a linear output, but you will find it won’t reach 12Vpp with a 12V supply.
This graph is good for helping when designing a circuit when you want a predictable output. So:
- When you’re working with something that is amplitude modulated, the linear graph is efficient.
AM clarification: I asked Mike what he meant by “the linear graph is efficient”: “To be able to ensure your transmitting station doesn’t over modulate, you have to have predictable output power. With known linear gain, you can put in a known range and expect to not put out more than allowed.” The linear graph displays efficient transfer of intended information, without loss or distortion.
- When using something that is at a given voltage, but frequency modulated, then the lower values are easier to work with and keep overall power budget down.
FM clarification: A frequency modulated wave has the same amplitude but different frequencies - hence its name. Because the amplitude is the same across the board, you don’t need to worry about designing an amplifier to work with such broad voltage changes.
In general, there are things you have to consider if you were to make a multistaged amplifier: final output stages use more linear approaches. And because each output affects the next stage’s input, lower load resistances allow you to tune noise reduction.