Tekko!
This is not clipping. This is the distortion caused by enormous dead time. You can test this: if you increase input voltage much, output voltage will be increased too, like in the attached picture.
Solution: change to IR2110, or other fast gate-driver! (But then you must do something against cross-conduction!)
This is not clipping. This is the distortion caused by enormous dead time. You can test this: if you increase input voltage much, output voltage will be increased too, like in the attached picture.
Solution: change to IR2110, or other fast gate-driver! (But then you must do something against cross-conduction!)
Attachments
This is just a prototype so i could get some IR2010 again, as well as some LT1016 comparators like titanchen used, build a new one and test again.
Tho i don´t have any 0.1µF bypass caps anywhere except over the 5v rails, so maybe the bootstrap cap (47µF) cannot do its job properly on higher freqs.
Tho i don´t have any 0.1µF bypass caps anywhere except over the 5v rails, so maybe the bootstrap cap (47µF) cannot do its job properly on higher freqs.
Pafi said:
This is also how his triangle wave looked before even having added level shifters or drivers.
Chris!
For this error "triangle" had to have 4 simmetrical discontinuities in a period. I doubt. But dead time couses exactly this.
For this error "triangle" had to have 4 simmetrical discontinuities in a period. I doubt. But dead time couses exactly this.
Why the amp starts distorting and only gives out 10vpp cannot in any way be caused by the deadtime, its not even visible on the op waveform, and it doesent matter what the input freq is.
The modulator is also faaaaaaaar from overmodulation, on max volume the source can do the pwm is only vibrating to maybe 40/60 duty cycle, so it must be something else.
Also the input to the gate driver is not even close to a squarewave, looks more like a capacitor charge/discharge curve, tho the gate outputs are good squares.
The modulator is also faaaaaaaar from overmodulation, on max volume the source can do the pwm is only vibrating to maybe 40/60 duty cycle, so it must be something else.
Also the input to the gate driver is not even close to a squarewave, looks more like a capacitor charge/discharge curve, tho the gate outputs are good squares.
Tekko said:Blah blahToo bad there arent any forums specialized in diy class d.
Too bad google doesn't work for you, and you continue to dismiss the advice given by those who you should pay attention to.
??? What is this forum??? Too bad that you need dosens of schematics of existing circuits, much help, and some weeks to build a simple class D amp. Grow up! I hadn't seen any classD circuit before I designed (design! not randomly connect or copy and paste!) my first one, and I didn't use simulator. Use your brain finally (if you have it) instead of asking and blaming others! And don't lie that your circuit is working well, before you measure anything, because we are not stupid!
Godspeed!
Godspeed!
**** what happen on this forum???? Second bitching thread!!!
Simple question, simple answer Tekko....How can you modulate more than the amplitude of your triangle wave??? Think about this Tekko!
Fredos
www.d-amp.com
Simple question, simple answer Tekko....How can you modulate more than the amplitude of your triangle wave??? Think about this Tekko!
Fredos
www.d-amp.com
Tekko said:
Tekko don't be silly and follow Pafi's advice. There is something wrong with your modulator, the input waveform to the gate driver should be almost perfect square waves too. Also, there is no vodoo here, the steady state output voltage is always the rail voltage multiplied by the duty cycle.
Tekko said:
So then what I trust you've learnt so far:
1. Don't be too quick to convince yourself
2. You at least saved a good bundle so far by doing this on a protoboard of sorts, no 100$+ PCB's this time
3. Still have lots to learn, can't expect people to keep giving you answers to easy questions that you can figure out yourself.
4. Too much time on IRC doesn't pay off
5. That'd I'd throw this in your face when you needed it most 🙂
Eva!
There is a very simple explanation.
"Also the input to the gate driver is not even close to a squarewave, looks more like a capacitor charge/discharge curve"
It can be what it looks like. Level shifter charges slowly the input capacitance of scope cable. Tau= ~100pF*6,8k=0,7 us. He probably doesn't use divider.
There is a same simple explanation of limited output voltage too, but if he doesn't want to know...
PS.: Have the wires arrived?
There is a very simple explanation.
"Also the input to the gate driver is not even close to a squarewave, looks more like a capacitor charge/discharge curve"
It can be what it looks like. Level shifter charges slowly the input capacitance of scope cable. Tau= ~100pF*6,8k=0,7 us. He probably doesn't use divider.
There is a same simple explanation of limited output voltage too, but if he doesn't want to know...
PS.: Have the wires arrived?
- Status
- Not open for further replies.