Class D amplifier design questions

[lemon14899]

I am currently designing an audio amplifier using TI's TPA3123 class-D amplifier, which can deliver about 20W of power into each speaker in a stereo configuration(power supply of 24V).
Some questions:

1. How much current does the amplifier consume at full power(40W total)? The speakers I am using are both 6 ohms passive(they come from an old hifi system I have). The datasheet outlines that the amplifier IC has a quiescent current draw of about 37mA, but says nothing about active current draw.

2. Do I need RC snubbers on the speakers to prevent kickback when the speakers are deactivated? From my understanding, speakers are basically a coil of wire wrapped around an iron core(an inductor) which vibrates, making noise.

3. A bit more complicated, but the datasheet says 1uF capacitors are needed in series on both inputs. However, these are the polarized kind(negative terminal to input). Aren't polarized capacitors supposed to be damaged when you charge them through the negative terminal? Or should I use a NP/bipolar aluminium electrolytic/tantalum instead?

(Note: while my speakers are 6 ohms each, I am designing the amplifier for 8)

[gilshultz]

If you check the manufacturer's application notes and data sheets they will contain an answer to your question. What part of this is related to an Arduino?

[DVDdoug]

@lemon14899

How much current does the amplifier consume at full power(40W total)?

The datasheet has a graph showing efficiency of about 90% at maximum output.
So that would mean about 45W from the power supply. Power can be calculated as Voltage X Current so at 24V that's around 2 Amps.

Note that with regular program material, the average power is about 10% of the peak, so although your power supply should usually be capable of supplying continuous power, the actual power consumption will be less.

1. Do I need RC snubbers on the speakers to prevent kickback when the speakers are deactivated? From my understanding, speakers are basically a coil of wire wrapped around an iron core(an inductor) which vibrates, making noise.

I don't know but whatever the datasheet recommends.

2. A bit more complicated, but the datasheet says 1uF capacitors are needed in series on both inputs. However, these are the polarized kind(negative terminal to input). Aren't polarized capacitors supposed to be damaged when you charge them through the negative terminal?

With a single positive power supply, the input on the chip will have a positive "bias" voltage. It's probably half of the supply voltage and comparatively the input signal is small.

The capacitor should have the positive side to the chip. The capacitor "isolates" the bias, preventing anything on the input from messing-up the bias as well as preventing the bias from leaking "backwards" into the preamp (or whatever is connected).

Or should I use a NP/bipolar aluminium electrolytic/tantalum instead?

You could use a non-polarized film capacitor. Non-polarized electrolytics are a bit of a "hack" and they are usually only used on inexpensive speaker crossovers.

[Grumpy_Mike]

@lemon14899

but says nothing about active current draw.

The active current draw will be given by the voltage output driving the output load. It is simple ohms law. Voltage / load = current

@lemon14899

Do I need RC snubbers on the speakers to prevent kickback

Yes. They are there in the sample schematics in the data sheet. They are the inductor and capacitor shown.

@lemon14899

the datasheet says 1uF capacitors are needed in series on both inputs.

Yes that is right.

@lemon14899

Aren't polarized capacitors supposed to be damaged when you charge them through the negative terminal?

Yes they are. But when you drive an amplifier input you are only supplying current in one direction. The circuit shown in the data sheet is correct.

This sort of amplifier needs a PCB with correct layout for stability. I don't think you can use solderless bread board to try it out.

[Paul_KD7HB]

@lemon14899

How much current does the amplifier consume at full power(40W total)? The speakers I am using are both 6 ohms passive(they come from an old hifi system I have). The datasheet outlines that the amplifier IC has a quiescent current draw of about 37mA, but says nothing about active current draw.

Well, the 40 watts have to come from somewhere, so they come from the power supply. That is 20 watts per channel of peak audio power. So your power supply must be able to supply the peak of 40 watts at 24 volts. Distortion ( class D is already distorted audio) so any less power will be noticeable to you. That would indicate at least a 2 amp DC supply.