A Triangular waveform compleetly defeets the point of a class D amp. As with this sort of amp the output devices are supposed to be in their on or off state, where they dissipate littel power (due to either no curent flowing, or only a very small voltage drop when fully turned on). If you but anything other than a sqare wave through them, you will be not fully on or off very much at all, so there will be a lot of power dissipated in the transistors and the amps efficienct will fall quite a lot.
You misunderstand, bigparsnip 
The triangle/sawtooth wave is what you use in a class-d amp as a reference to produce the PWM wave. amp_man_1 has no intention of trying to force a triangle wave through the output!
And for amp_man_1, triangle is best as soren has mentioned.
I designed a class-d amp on paper a couple of years ago and am now thinking of implementing it It was spookily similar to the one you posted recently. However, the NE5532 you used is too slow IMO. I am currently looking into very fast op-amps to try.
edit: Triangle is also best because you can make it symetrical about 0V very easily. With a sawtooth you have to align exactly to the mid-point (not easy) to have 'zero' output from the PWM, i.e. 50:50 mark:space ratio.
The triangle/sawtooth wave is what you use in a class-d amp as a reference to produce the PWM wave. amp_man_1 has no intention of trying to force a triangle wave through the output!
And for amp_man_1, triangle is best as soren has mentioned.
I designed a class-d amp on paper a couple of years ago and am now thinking of implementing it
It was spookily similar to the one you posted recently. However, the NE5532 you used is too slow IMO. I am currently looking into very fast op-amps to try.edit: Triangle is also best because you can make it symetrical about 0V very easily. With a sawtooth you have to align exactly to the mid-point (not easy) to have 'zero' output from the PWM, i.e. 50:50 mark:space ratio.
Op-Amps
Here's a hint. Yes Op-Amps are too slow. The really good Class D amplifiers use the LM319 precision comparitor. Two of them with the audio signal applied to each in cophase but the same triangle applied to each. You get cophase PWM signals and if you apply those in bridge mode to a bridge output stage, a lot of the grunge disappears.
Here's a hint. Yes Op-Amps are too slow. The really good Class D amplifiers use the LM319 precision comparitor. Two of them with the audio signal applied to each in cophase but the same triangle applied to each. You get cophase PWM signals and if you apply those in bridge mode to a bridge output stage, a lot of the grunge disappears.
1. You will still need OP amps for the generation of the triangular waveform and for the NFB.
2. Use of the LM319 per se does not mean that you will end up with a high-quality amp.
3. Compared to some modern op-amps (and some other comparators of course) also the LM319 is still quite sluggish !
Regards
Charles
P.S. One advantage of the LM319 is that it is dirt cheap !
2. Use of the LM319 per se does not mean that you will end up with a high-quality amp.
3. Compared to some modern op-amps (and some other comparators of course) also the LM319 is still quite sluggish !
Regards
Charles
P.S. One advantage of the LM319 is that it is dirt cheap !
Op-Amps
I'm only talking about the stage where the audio is converted to PWM. The LM319 seems to be the device of choice with class-d designers. It is certainly better than using an op-amp. Besides, the speed is still quite fast enough. What is more important is having a low and stable offset voltage and a match between the comparitors for this spec.
The ICEPower modules use the 33078 and 33079 op-amps for the places where op-amps are needed and the LM339 for the shut down circuits.
I'm only talking about the stage where the audio is converted to PWM. The LM319 seems to be the device of choice with class-d designers. It is certainly better than using an op-amp. Besides, the speed is still quite fast enough. What is more important is having a low and stable offset voltage and a match between the comparitors for this spec.
The ICEPower modules use the 33078 and 33079 op-amps for the places where op-amps are needed and the LM339 for the shut down circuits.
Commentable Thoughts
Extreme Warm ThanX to SIR, CHARLES
For his Phase_Accurated Comment!
BUT i recently viewed some high-end amps schematics in which they are using Sawtooth @ 300KHz instead of Triangular and they also claim that using Sawtooth improves linearity in Pulse Switching.
Whats ur Opinion,SIR.
Regards
Ampman
Extreme Warm ThanX to SIR, CHARLES
For his Phase_Accurated Comment!
BUT i recently viewed some high-end amps schematics in which they are using Sawtooth @ 300KHz instead of Triangular and they also claim that using Sawtooth improves linearity in Pulse Switching.
Whats ur Opinion,SIR.
Regards
Ampman
hi.
>ICE1000A amplifier module, 1000W RMS and they sound great>
its a subwoofer amplifier , and it goes to 1 khz
i guess your definition of great sound is different from most of us, but nice marketing again
rgds karsten madsen - www.cadaudio.dk
>ICE1000A amplifier module, 1000W RMS and they sound great>
its a subwoofer amplifier , and it goes to 1 khz
i guess your definition of great sound is different from most of us, but nice marketing again
rgds karsten madsen - www.cadaudio.dk
O.K.
Those convenient SMPS ICs (which I have never used, so I can't comment on them) use an integrated sawtooth generator.
If you want to generate the waveform by YOURSELF then it is much easier to generate a decent triangular than a sawtooth, as others already have mentioned before.
If the waveform and the comparator were ideal, both the sawtooth and the triangular would give a perfectly linear modulation though.
I assume that the claims about sawtooth being better are just marketing BS (as are many claims amongst switching amps).
There is one property of a sawtooth that is disturbing me (though I don't know how much it affects quality): The rectangular is not only modulated in pulse-width but also in phase.
Regards
Charles
Those convenient SMPS ICs (which I have never used, so I can't comment on them) use an integrated sawtooth generator.
If you want to generate the waveform by YOURSELF then it is much easier to generate a decent triangular than a sawtooth, as others already have mentioned before.
If the waveform and the comparator were ideal, both the sawtooth and the triangular would give a perfectly linear modulation though.
I assume that the claims about sawtooth being better are just marketing BS (as are many claims amongst switching amps).
There is one property of a sawtooth that is disturbing me (though I don't know how much it affects quality): The rectangular is not only modulated in pulse-width but also in phase.
Regards
Charles
ICEPower
http://www.medicom.bang-olufsen.com/sw1062.asp
AES publications about ICEPower.dmfraser said:
http://www.medicom.bang-olufsen.com/sw1062.asp
hi.
amp_man i wasnt answering you.
>I find the ICEPower are the best class D amplifiers available in terms of cost and sound quality.>
what other types did you hear and compare to?
others dont seem to share your high thoughts about their sound quality either but im glad you can at least confirm that they are cheap (we had that discussion here some time ago in a discussion with av reality who claimed the opposite
rgds karsten madsen - www.cadaudio.dk
amp_man i wasnt answering you.
>I find the ICEPower are the best class D amplifiers available in terms of cost and sound quality.>
what other types did you hear and compare to?
others dont seem to share your high thoughts about their sound quality either but im glad you can at least confirm that they are cheap (we had that discussion here some time ago in a discussion with av reality who claimed the opposite
rgds karsten madsen - www.cadaudio.dk
Isn't Jeff Roland Design using the ICEpower modules now too? Roland is well known for his perfectionism. Hard to believe he would have chosen them if he weren't satisfied with them.
I heard from many people that attended the recent CES that heard them that they were quite excellent. I have also been told that these new Roland amps are almost impossible to get at the moment as the demand has greatly outstripped supply. Clearly more than a few people like them.
I'm curious Mr. Madsen, you seem to be quite critical of many current class d amps on the market. Are there any you do like?
I heard from many people that attended the recent CES that heard them that they were quite excellent. I have also been told that these new Roland amps are almost impossible to get at the moment as the demand has greatly outstripped supply. Clearly more than a few people like them.
I'm curious Mr. Madsen, you seem to be quite critical of many current class d amps on the market. Are there any you do like?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.