I found another problem , the crackle not only causing by clipping.
My pre-amp have 4 funtion , gain , LPF , Subsonic , Booster.
When the input signal get clip that has DC signal, and across the subsonic.
The signal of DC part will get +/- pulse signal very fast.
I probe the output , when the input get clip , i will get crackle sound.
does anyone has ideal to solve it ???
Or any experience can help me ???
My pre-amp have 4 funtion , gain , LPF , Subsonic , Booster.
When the input signal get clip that has DC signal, and across the subsonic.
The signal of DC part will get +/- pulse signal very fast.
I probe the output , when the input get clip , i will get crackle sound.
does anyone has ideal to solve it ???
Or any experience can help me ???
If you could provide more info on your actual setup, the problem would be easier to trace! 
For example, are you using a mosfet full-bridge, "bridge tied load"? Have you included an LC lowpass filter on the output or are you driving speakers directly? (for example the L core /might/ be saturating and cause thefast voltage rise).
How do you generate the PWM signals? Are you using a digital PCM => PWM converter chip? Or analog to PWM? And are you using feedback from the output or are you running "open loop"?
And, do you get the crackle waveforms also when you are running without a real speaker load, e.g. just a say 5W 110ohm wire-wound resistor? (as an initial quick check to see if this could be a feedback loop stability or interference problem...)
Have you tried the separate battery thing for the IR2110 bootstrap supply?
For example, are you using a mosfet full-bridge, "bridge tied load"? Have you included an LC lowpass filter on the output or are you driving speakers directly? (for example the L core /might/ be saturating and cause thefast voltage rise).
How do you generate the PWM signals? Are you using a digital PCM => PWM converter chip? Or analog to PWM? And are you using feedback from the output or are you running "open loop"?
And, do you get the crackle waveforms also when you are running without a real speaker load, e.g. just a say 5W 110ohm wire-wound resistor? (as an initial quick check to see if this could be a feedback loop stability or interference problem...)
Have you tried the separate battery thing for the IR2110 bootstrap supply?
1.I am usgin half-bridge. Yes, I use LC lowpass filter on the output.
2. I directly convert the analog to PWM signal.
I connect the feedback before LC filter on the mosfet output.
Use a 1K and 200p do a LPF , and use a 47K be the feedback resistor.
Input resistor is 2.2k.
3. I only test it without load. i input a 160Hz sine wave.
It not get crackle when output amplitude get 45v.
When output amplitude higher than 45v , the top of since get some oscilation, Is that called crackle ???
But when i input the signal from cd player.
When the output get about 35v, it start get strange pulse on the signal.
It sounds like crackle. But the waveform not like the oscilation on the top of sine.
4. do you mean use a seperate +15v to the Vb from Vs???
And remove the bootstrape diode from Vcc of IR2110???
2. I directly convert the analog to PWM signal.
I connect the feedback before LC filter on the mosfet output.
Use a 1K and 200p do a LPF , and use a 47K be the feedback resistor.
Input resistor is 2.2k.
3. I only test it without load. i input a 160Hz sine wave.
It not get crackle when output amplitude get 45v.
When output amplitude higher than 45v , the top of since get some oscilation, Is that called crackle ???
But when i input the signal from cd player.
When the output get about 35v, it start get strange pulse on the signal.
It sounds like crackle. But the waveform not like the oscilation on the top of sine.
4. do you mean use a seperate +15v to the Vb from Vs???
And remove the bootstrape diode from Vcc of IR2110???
titanchen86!
You are the only person can tell what does crackle mean, since you said it!
You should define "strange pulse" more properly too (eg. by a picture).
45V doesn't mean anything without knowing the supply voltage.
If it occures when output voltage reaches supply rail, at no loat attached, then it is absolutely normal. Output filter is ringing without load. This cannot cause crackle sound.
P.S.: did you read what I wrote previously?
P.S.2.: You need a dummy load, and you have to measure with it is attached to output, with sine wave.
You are the only person can tell what does crackle mean, since you said it!
You should define "strange pulse" more properly too (eg. by a picture).
45V doesn't mean anything without knowing the supply voltage.
If it occures when output voltage reaches supply rail, at no loat attached, then it is absolutely normal. Output filter is ringing without load. This cannot cause crackle sound.
P.S.: did you read what I wrote previously?
P.S.2.: You need a dummy load, and you have to measure with it is attached to output, with sine wave.
Might your circuit be so noisy in EMI terms that it could disturb CD player operation?
My first high-power SMPS did exactly that. I was naively hard switching 340V over a conducting diode in a point to point DIY layout and I was not even using proper common-mode filtering. After a certain output current value was reached, the CD player started distorting and skipping, and if I increased the current further, it would just stop playing. There was no connection between both devices, except the own mains supply, and they were almost 1,5 meters apart.
Fortunately, I have learnt a lot since then
My first high-power SMPS did exactly that. I was naively hard switching 340V over a conducting diode in a point to point DIY layout and I was not even using proper common-mode filtering. After a certain output current value was reached, the CD player started distorting and skipping, and if I increased the current further, it would just stop playing. There was no connection between both devices, except the own mains supply, and they were almost 1,5 meters apart.
Fortunately, I have learnt a lot since then
OK ,, I will post some picture later... give me a moment to ready.!
My +/-RAIL is 56v. Later i will post the test picture that test with load.
Eva said:Might your circuit be so noisy in EMI terms that it could disturb CD player operation?
My first high-power SMPS did exactly that. I was naively hard switching 340V over a conducting diode in a point to point DIY layout and I was not even using proper common-mode filtering. After a certain output current value was reached, the CD player started distorting and skipping, and if I increased the current further, it would just stop playing. There was no connection between both devices, except the own mains supply, and they were almost 1,5 meters apart.
Fortunately, I have learnt a lot since then
Yes, i do the class-d with a smps .
Because i want to use in the car.
you said " common-mode filtering " ,,, what is that ????
Is that using in smps ???
At last , do you solve the problem???
Yes.
And really I mean a battery with just short wires connected, and a small one. Preferrably with a 2.2ohm series resistor, "just in case". Definitely do not use a mains powered power supply or bench-top supply for this. This is since the entire battery thing has to be capable of jumping around at those e.g. +-50V @ 400kHz (or what your setup uses).
Yes.
Oh and is it one of those self-oscillating designs that you are using? Or a fixed frequency PWM design?
Just thinking, for self-oscillating, have you checked that you don't overvolt the error opamp input, or that the error voltage integrator begins clipping? Aside from that there might be some other issues with the feedback loop too - depends a bit what you actually mean by this "crackle"
At least for the problem I had and which was because of the IR2110 bootstrapping and PCB layout, at higher rail voltages the sound began to "crackle" in a way like it sounds when you have a bad contact somewhere. No clipping, just that sort of scratchy sound output...
Wait a sec, do you mean you feed the class-d amp from an SMPS? Have you checked the output voltage of the SMPS on an oscilloscope while there is 'crackle' on the class-d amp output? Not that the SMPS is overloaded a slightly by your amp, and the SMPS switching frequency and class-d switching frequency difference and phase difference mix stuff into the audio band...
jwagnerhki said:
yes, I use fix frequency modulation.
Yes, I feed the class-d amp from an SMPS , 12v to +/-56v RAIL.
No , i didn't check the output voltage of the smps when crackle.
Do you mean the " strange pulse " is causing by smps is overloaded??
And the hormonic frequency into the audio band ???
My camera get some problem , I will post picture when the camera fixed.
You can't expect us to do your homework. You have to make the "crackle" situation happen, and then you have to use an oscilloscope to spy into each node of the entire circuit (including checking for proper SMPS behaviour) and try to understand what is actually happening.
There are no magic solutions in electronics, only hard work!
Use google to search articles and read on common-mode EMI and filtering. You won't be able to design good switching circuits until you fully understand and apply these concepts. Essentially, common-mode EMI is created by RF voltage differences appearing across PCB traces and planes due to switching pulses and PCB resistance and inductance. That happens particularly in ground traces and planes. PCB size is usually small enough to not allow substantial radiation of these components, but when external wiring is connected to different ends of the PCB showing different ground potentials at RF, the result is like a nice dipole antena connected to a RF generator.
Proper PCB layout and filtering should be employed to attenuate these RF components before they can reach wirings. Components above 1Mhz are particularly harmful, as the higher the frequency the easier it becomes radiated and picked up by other equipment.
BTW: Yes, I learnt to solve these problems and my circuits no longer disturb the working of other equipment.
There are no magic solutions in electronics, only hard work!
Use google to search articles and read on common-mode EMI and filtering. You won't be able to design good switching circuits until you fully understand and apply these concepts. Essentially, common-mode EMI is created by RF voltage differences appearing across PCB traces and planes due to switching pulses and PCB resistance and inductance. That happens particularly in ground traces and planes. PCB size is usually small enough to not allow substantial radiation of these components, but when external wiring is connected to different ends of the PCB showing different ground potentials at RF, the result is like a nice dipole antena connected to a RF generator.
Proper PCB layout and filtering should be employed to attenuate these RF components before they can reach wirings. Components above 1Mhz are particularly harmful, as the higher the frequency the easier it becomes radiated and picked up by other equipment.
BTW: Yes, I learnt to solve these problems and my circuits no longer disturb the working of other equipment.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.