Noise problems in non discrete class-d amp

[Baldin]

I have been building on my first class-d circuit for some time, and have had really good help by reading the pages here on DIYAudio.com

I started building the amp on a breadboard, and eventually it worked fine. Only problem was a kind of white noise, which was always present.

I thought that a real PCB would cure that, and it almost have, but it is still there to some extent!

Also when playing at very low levels the bass notes have a sound like clipping or metallic (a bit hard to explain in words)! Maybe it’s just the noise that is "modilated" in tune to the music

Also it is very sensitive to touch; If I place a finger on the driver IC (HIP4080), when it is not playing, the noise almost disappears, but placing a finger on one of the caps that controls the PWM freq, does the complete opposite.

I have done some measuring with a scope (Se my homepage ) and nothing really sheds any light on the problem!

I’m pretty sure that the problem is to be found around the PWM oscillator.
Could also be a fundamental grounding problem (Using a ground plane split into two to kind of separate analog from digital).
I'm using a virtual ground for some of the circuit, but as can be seen on my homepage, the "noise" on this ground is not at all alarming 50 mV pp)

Have tried to use NE5532 instead of TL072, but the integrator in the PWM ran pretty hot, and it was difficult to adjust freq. down to below 300 kHz.

The amp is running on a small 16 V DC PSU (until I know it is working properly ).

The amp is intended for a subwoofer, and hope to get it to deliver more than 300 W …. Eventually

The pots for adjusting the dead time is set to 220 k for maximum dead time.


Anyone who has any ideas??

(Se attached diagram)
(Se picture of amp on hompage as well)

[Baldin]

Almost 100 reads and no replays from any of the gurus .... well no response at all!!

Well, I have played around with a few components, ending up with changing the loop resistor around the PWM stage, back to 47k, changing C12 to 1nF and C15 to 680 pF, and pacing a 4p7 ceramic between pin 5 a d 6 on the HIP (slowing the internal comparator down a little). This has almot cured the problem. The brake came form using the 4p7 ceramic on the comperator.

(The switching freq is now around 225 kHz, and the sound is good, though there is still a bit of noise when no music i palying)

Well I read in a thread somewhere that these noise problems maybe seems to be more or less inherent to class-d (switching noise that couples to the rest of the amp through loops etc.)

Then the only real cure is even more carefull design of PCB and use of SMD components.

What about snubbers on all the Mosfets ..... will that help eny, in this respect?

What about freewheeling diods on all the Mosfets?
How do you determin whether these are necessary, and how do you determine the rating in A needed?

Baldin



Sensible Audio

[peranders]

Class D is not trivial as you may have noticed, not many people in the whole world are experts.

Isn't 300 kHz way over the IC's specs? have you read the article in Elektor (rather recently) where they investigated a class D amp made of a HIP IC? The conclusion was: Not very very good PWM IC for audio.

[Baldin]

I'ts running below 250 kHz, which should be no problem for the IC.

No I haven't seen that article .... do you know which exact issue .... then I could download it from their homepage (3.80£)

I'd like to see why this should be such a bad IC for a class-d amp!
.... but I know it's reputation is maybe not the best, though I chose it because of the easy way to adjust deadtime, it's drive capacity, the build in comparator etc.

Mads


Sensible Audio

[witali]

white noise? Probably not symmetry of the feeding the opa-amp and comparer, I faced with this at building UcD on microcircuit

[Baldin]

Hi Witali
Thanks for your answer

What do you mean by symmetry? For the voltage supply?
Please explain a bit more

[Brian Oppegaard]

Try adding more capacitance to your 6V reference ground and omit the opamp buffer.

[Baldin]

Thanks for your answers so far

Have tried out a few things:

Mounting diodes parallel with the gate resistors (1N4148) have helped a bit with the noise, and defenetly with the heat disipation, especially in the LM7812 regulator the the driver IC!

Adjusting the resistors for delay on pin 8 and 9 from 220k to 115k has reduced the noise to a mere humm (not white anymore) which is affected bu tuching the ibput components (more of the same noise)

A 470 uF decoupling close to the Vdd/Vcc of the HIP is helped too.

More decoupling on the virtual ground does not seem to help any. But then again, I havent tried omitting the buffer as you suggested Brian ..... a bit too dificult on the current PCB

All in all, noise is down to almost nothing. Next step will be to raise the supply voltage and se what happens


Have read on Tripath's hompage that the gate resistors for the output FETs should be of at least 1W!!!! Is that really so?

I'm wondering whether the LM7812 used for the driver is actually good enough for this task. What are others experience in this matter??
The use of a 470 uF on the output side of the regulator seems excessive, and will render it's regulating ability somewhat useless! But is it any good with loads in the 250 kHz range??
What are you using for driver PSU??

[Tekko]

No, quarter watt resistors are fine. Atleast on my project 1/4watt gate resistors worked fine.

[witali]

I used resistors SMD size 0603 and and even less
See UcD180 PCB there 0603 resistors