My very first Class D pwm (switching) amplifier. - Page 17

Jaka Racman · 29th July 2003, 05:06 PM

simulation results

phase_accurate · 30th July 2003, 06:35 AM

Unfortunately the circuit does only damp the resonance because you don't have any feedback at lower frequencies (BTW did you try how it behaves without any load ?)
It can of course be extended so it will provide proper feedback !

Regards

Charles

ssanmor · 30th July 2003, 06:43 AM

Hello all again.
Here are the "definitive" version of my first prototype schematics. If anyone of you wants to have a look at them, any comment will be welcomed.
Meanwhile, I am ending the PCB design.

Thanks in advance.

Sergio

ssanmor · 30th July 2003, 06:49 AM

Oooops! I forgot to remark something about the schematic.
When doing my first tests, I found that, altougth the output of the MAX038 was a perfect square wave, it gets slightly distorted after being amplificated. When I disconnected the output of the opamp U4B from the input of the comparator U5A, it improved, and also when I reduced both the input and feedback resistors, R18 and R10. (they were originally 10K and 22K). What can the cause be? The bandwidth and slew rate of the amplifier seem far higher than needed, aren't they?

Best regards.

ssanmor · 30th July 2003, 06:50 AM

Sorry again. The output of the MAX038 is a TRIANGLE (not square) wave, of course!

phase_accurate · 30th July 2003, 07:23 AM

Hi Sergio

Just try to imagine how much (parasitic) capacitance it takes to form a 100 kHz (or whatever) lowpass with a 22 k resistor !

Whatever you do you have to be always aware that you are building a precision RF circuit used to process audio !

Something that hasn't been discussed so far are the freewheeling diodes that handle the "backward" current caused by the output inductor. Depending on switching frequency you can't always rely on the built in diodes of the mosfets.

Regards

Charles

ssanmor · 30th July 2003, 07:32 AM

Thanks for your fast response time, Charles.

So, if I have understood, the high previous value of 22K in the feedback of the U4B amplifier in conjunction with the parasitic capacitance of the comparator can distort the signal, is that the problem? I reduced it to 2k2 and it improved, but is not by far as perfect as the MAX038 output. Any sugestions?

Best regards

phase_accurate · 30th July 2003, 07:55 AM

Hi Sergio

I was just talking about a possible cause. Be aware that the dv/dt of a triangular may not look very high but it is still a demanding signal.
To avoid unnecessary distortion an opamp must be faster by at least an order of magnitude. The gain-bandwidth product has to be quite high as well: Just try to imagine how much NFB is left at the fundamental frequency of the triangular, let alone it's harmonics !
And don't forget that the lower and upper edge of the triangular represent a very high slew-rate signal.

I forgot to look at your circuit again before starting to post but I will have another look to see, why you use this opamp at all.

Regards

Charles

phase_accurate · 30th July 2003, 08:02 AM

Hi Sergio

I assume it is there only to increase the amplitude of the triangular by a factor of two.
While a large triangular amplitude is a good thing, regarding susceptibility of the modulator to spikes etc, it's linearity is important as well.
So I would suggest to try it with and without this opamp stage and compare the results (the feedback has to be dimensioned accordingly).

Regards

Charles

koldby · 30th July 2003, 08:39 AM

Very nice ssanmor!

I specially like the very straight foreward driver circuit.

It looks like you have made the schematics in Prolte?

Can you upload (or e-mail) the .sch files???

Koldby

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

30th July 2003, 06:35 AM	#162
phase_accurate diyAudio Member Join Date: May 2002 Location: Switzerland	Unfortunately the circuit does only damp the resonance because you don't have any feedback at lower frequencies (BTW did you try how it behaves without any load ?) It can of course be extended so it will provide proper feedback ! Regards Charles

30th July 2003, 06:49 AM	#164
ssanmor diyAudio Member Join Date: Jan 2002 Location: Madrid	Oooops! I forgot to remark something about the schematic. When doing my first tests, I found that, altougth the output of the MAX038 was a perfect square wave, it gets slightly distorted after being amplificated. When I disconnected the output of the opamp U4B from the input of the comparator U5A, it improved, and also when I reduced both the input and feedback resistors, R18 and R10. (they were originally 10K and 22K). What can the cause be? The bandwidth and slew rate of the amplifier seem far higher than needed, aren't they? Best regards.

30th July 2003, 06:50 AM	#165
ssanmor diyAudio Member Join Date: Jan 2002 Location: Madrid	Sorry again. The output of the MAX038 is a TRIANGLE (not square) wave, of course!

30th July 2003, 07:23 AM	#166
phase_accurate diyAudio Member Join Date: May 2002 Location: Switzerland	Hi Sergio Just try to imagine how much (parasitic) capacitance it takes to form a 100 kHz (or whatever) lowpass with a 22 k resistor ! Whatever you do you have to be always aware that you are building a precision RF circuit used to process audio ! Something that hasn't been discussed so far are the freewheeling diodes that handle the "backward" current caused by the output inductor. Depending on switching frequency you can't always rely on the built in diodes of the mosfets. Regards Charles

30th July 2003, 07:32 AM	#167
ssanmor diyAudio Member Join Date: Jan 2002 Location: Madrid	Thanks for your fast response time, Charles. So, if I have understood, the high previous value of 22K in the feedback of the U4B amplifier in conjunction with the parasitic capacitance of the comparator can distort the signal, is that the problem? I reduced it to 2k2 and it improved, but is not by far as perfect as the MAX038 output. Any sugestions? Best regards

30th July 2003, 07:55 AM	#168
phase_accurate diyAudio Member Join Date: May 2002 Location: Switzerland	Hi Sergio I was just talking about a possible cause. Be aware that the dv/dt of a triangular may not look very high but it is still a demanding signal. To avoid unnecessary distortion an opamp must be faster by at least an order of magnitude. The gain-bandwidth product has to be quite high as well: Just try to imagine how much NFB is left at the fundamental frequency of the triangular, let alone it's harmonics ! And don't forget that the lower and upper edge of the triangular represent a very high slew-rate signal. I forgot to look at your circuit again before starting to post but I will have another look to see, why you use this opamp at all. Regards Charles

30th July 2003, 08:02 AM	#169
phase_accurate diyAudio Member Join Date: May 2002 Location: Switzerland	Hi Sergio I assume it is there only to increase the amplitude of the triangular by a factor of two. While a large triangular amplitude is a good thing, regarding susceptibility of the modulator to spikes etc, it's linearity is important as well. So I would suggest to try it with and without this opamp stage and compare the results (the feedback has to be dimensioned accordingly). Regards Charles

30th July 2003, 08:39 AM	#170
koldby diyAudio Member Join Date: Jun 2003 Location: Ruds Vedby	Very nice ssanmor! I specially like the very straight foreward driver circuit. It looks like you have made the schematics in Prolte? Can you upload (or e-mail) the .sch files??? Koldby

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)