Class D audio amplifiers!

[pyrohaz]

Hey guys, i've been reading up on Class D amplifiers and how 555 timers can be used for Pulse width modulation. I decided to see how hard it would be to make a simple class d amplifier using a 555 timer.

I've included a schematic of my circuit I have designed. Oh yeh, one mistake I did make in the schematic was: The speaker i'm using is 8R not 4R

The 555 timer operated at 96KHz, Just below the 100KHz limit of the 555 timer. The first LC Lowpass filter is set at 10.7KHz using a 220uH Inductor and 1uF Tantalum capacitor. The second filter is set at just over 100KHz just incase any harmonics go over the first filter (more a safety precaution than anything). The output capacitor is only 1000uF because I didn't have any higher values at hand. I worked out that the reactance of this capacitor at 50Hz was 3.18R, not a major problem. I evaluated the bass response of this system with a sweeping track from 40Hz to 100Hz and after personally hearing, it sounded as if the peak was ~60Hz. I am using a JVC 8W 4" Hi-Fi Speaker. My power supply is a 16v 2.2A Supply with a 4mH Choke in series and 470uF Capacitor across the supply (After the choke) to lower supply ripple.

Now! What I was wondering:

1. I know that Class D amplifiers are meant to use Mosfets instead of transistors but I didn't have any at hand. Would using Mosfets benefit in any way? After doing a skin test on all the components, the heatsinked transistors were still cool enough to touch continuously and the 741 and 555 timer were the same temperature as when I started (the amplifier had been running for 1 Hour playing some music).

2. Would I benefit from using a TS555 instead of a NE555? I read an Application Note from International Rectifier and it stated that reducing the dead time from 40nS to 15nS reduced THD by 1.92% so since the Rise and fall time of a NE555 is 100nS, the TS555's being 25nS rise and 20nS fall, would this really make a noticable difference?

3. If I swapped out the 741 for a TL0x1 series, would it make much of a difference?

Thanks in advanced guys :)

[Hero999]

Sorry but the circuit you posted will be no more efficient than a class AB amplifier and won't sound anywhere near as good.

[pyrohaz]

Thanks for the reply :), do you know of ways I could make it more efficient? If I swapped to a TS555 and upped the frequency, could this help?

Please help on how I can make it more efficient :D

[audioguru]

Your ordinary old 555 has a total output voltage loss of 4V. The output emitter-followers have a loss of 2V. Then the total losses are 6V and since your supply is 16V then the amplifier is only 62.5% efficient.

A Cmos 555 has an output current that is too low to drive power transistors. Use Mosfets with a Cmos 555 instead.

A little 4" speaker cannot produce deep bass frequencies so it is not hi-fi.

Your filter cuts frequencies above only 10kHz so it is not hi-fi. Also, your filter will not reduce high frequency distortion much.

EDIT:
You are using a lousy old 741 opamp. It is 42 years old and cannot produce frequencies above only 9kHz. It is also not hi-fi.
Your filter will probably cause it to oscillate.

[pyrohaz]

Thanks a bunch dude, you're a real help :) Right, how about if I modify it to:

8" Speaker with tweeter (my dad uses them for his Denon integrated system lol)

Use a TS555 instead (Please can you recommend what MOSFETS I should use? I haven't used mosfets yet, primarily because I just haven't got any!

Replace the 741 with a TL071?

What kind of output filter do you recommend?
Inductors I have:
4mH
1mH
220uH
22uH

For the speaker, I only called it hifi because thats what the sticker says :)

[pyrohaz]

I had a search on my favourite website, eBay! :D And I found:

IRF540
IRF9540

I think the 9540 is the P complement of the 540?

Would these suit for my product because its only ~

[audioguru]

1. I know that Class D amplifiers are meant to use Mosfets instead of transistors but I didn't have any at hand. Would using Mosfets benefit in any way?

Much less voltage loss so much better efficiency.

After doing a skin test on all the components, the heatsinked transistors were still cool enough to touch continuously and the 741 and 555 timer were the same temperature as when I started (the amplifier had been running for 1 Hour playing some music).

the power from the amplifier and the power wasted were very low.

2. Would I benefit from using a TS555 instead of a NE555?

it has no voltage loss and it operates at a higher frequency. Use 500kHz.

I read an Application Note from International Rectifier and it stated that reducing the dead time from 40nS to 15nS reduced THD by 1.92% so since the Rise and fall time of a NE555 is 100nS, the TS555's being 25nS rise and 20nS fall, would this really make a noticable difference?

The Mosfets have a high input capacitance that makes them switch slowlywhen driven from a Cmos low power 555. Then the distortion might be high and the "shoot-through" current might also be high when both Mosfets are turned on at the same time shorting the supply.

3. If I swapped out the 741 for a TL0x1 series, would it make much of a difference?

Yes.
The lousy old 741 opamp is horrible for audio. The TL07x is excellent. They cost the same.

Since your supply is only 16V then you will be lucky for the amplifier to produce only 3.5W into 8 ohms at clipping.

I found IRF540 and IRF9540.

They are fairly old but are fine for your low power amplifier. In Britain you pay 3 times more than I pay in Canada for those Mosfets. Your gasoline (petrol) also costs 3 times more. Does everything costs 3 times more?? Why?? Taxes?? For your Queen??

[Hero999]

There are probably cheaper and better MOSFETs around.

Audioguru hasn't explained how he came to the conclusion about the power output.

It's Ohm's law.

The maximum peak voltage output of a half bridge driver, which is what you have, is always half the maximum voltage swing, which is always less than the supply voltage.

There will always be a voltage loss across each driver transistor, the lower the better.

In this case you've used emitter followers on the output which have a voltage loss of 0.7V unloaded and much higher with a load. Because emitter followers don't have any gain the voltage loss of the driver transistors inside the 555 (see datasheet) will be added to the loss of the output transistors. See Wikipedia for much more information on emitter followers.

The calculation for the peak output power is as follows:
P = V²/R

The RMS power is half the peak power.

[audioguru]

3.5W into 8 ohms is a sine-wave of 15V p-p.
With a 16V supply and half-decent Mosfets that each has a voltage loss of 0.5V then 3.5W into 8 ohms is possible.

Texas Instruments make class-D audio amplifier ICs. Many are bridged (two amplifiers driving each end of a speaker out-of-phase for nearly double the output voltage swing which makes nearly 4 times the power od a single-ended amplifier.
Also many amplifiers can be paralleled for a higher output current.