Filter - less D class amplifier

Status
Not open for further replies.
Jump to Latest
#21
vacuphile said:
A pulse contains all frequencies, you are confused between the switching frequency, and the spectral content of the switch on and off pulses, which because of the fast rise times contains high frequency hash. These frequencies are not related to the switching frequency.
Click to expand...
An isolated pulse contains all frequencies from DC to infinity. A periodic pulse does not.

Two different ways to think about this:
1. Any periodic waveform consists of just fundamental and harmonics (says Mr. Fourier).
2. A set of pulses will each generate their individual components but at different phases, some of these will add and some will (partly) cancel, so put together an infinite sequence of regular pulses and most of the components will cancel leaving just the harmonics.
The nice thing about mathematics is that you can't escape it!

Mikelandjelo said:
What is the catch?
Click to expand...
Copious RF interference, and no audio advantage, so why do it? Filterless Class D is only used by two groups:
- commercial very low power mass market designs where saving a few pence and some space is deemed worthwhile, and the necessary spread-spectrum logic can be squeezed into the audio chip
- unthinking DIYers
 
#22
Julf said:
Filterless class D is accomplished by noise modulation, but that doesn't actually reduce the noise, it just spreads the spectrum over a wider range. If you have managed to come up with some great new scheme that gets around that, then good for you! Looking forward to reading the AES or IEEE paper about it!
Click to expand...

Spread spectrum is one way but not a solution.

DF96 said:
Copious RF interference, and no audio advantage, so why do it? Filterless Class D is only used by two groups:
- commercial very low power mass market designs where saving a few pence and some space is deemed worthwhile, and the necessary spread-spectrum logic can be squeezed into the audio chip
- unthinking DIYers
Click to expand...

So its still dont do it.... 🙂 .... Can we start with exact clear discussion of how it can be achieved to work?We all here have expirience and a tones of working hours in circuit designing....I never woke up with thoughts - it is impossible dont even try to achieve it .... Is there any constructive thought how filterless output can be achieved (in theory) to work?
 
#24
Mikelandjelo, which parameter(s) do you think you will improve by leaving out the filter? Whats your tecnical design goal and what will it achieve in terms of audible improvements?

I applaud your ambition and spirit!

//
 
#25
Mikelandjelo said:
Is there any constructive thought how filterless output can be achieved (in theory) to work?
Click to expand...

Well, the reason is probably that filterless class D simply does not exist. All you do by leaving out the output filter is having the speaker drivers themselves doing it instead. And since they are highly reactive and variable with temperature (induced or ambient) and many other parameters they make very poor and unreliable filters.

You mentioned earlier you wanted to slow the rise time down but that gives up efficiency. In fact, you could probably compare such an amp with an FM receiver with a class AB output.

I can think of a way that you might achieve high(er) output power with just a ferrite bead filter though and that is interleaved multi-phase modulation but to my knowledge attempts at that has not been successful yet.
 
Last edited:
#28
DF96 said:
An isolated pulse contains all frequencies from DC to infinity. A periodic pulse does not.
Click to expand...

Of course a periodic pulse does contain the same frequencies as the same pulse in isolation, but I am not going to waste any more time on this. Except for saying that it is the fast switching pulses which create all the rf hash, and not the switching frequency per se.
 
Last edited:
#29
TNT said:
@TNT ?
Click to expand...


TNT thank you too.

For those who think that its not worth of trying i can say that u r very wrong.I thought that we will have here constructive dsicussion but as far as i can see we cant.I am not DIYer.I am living from designing and making new solutions for companies.I have 18 years of professional expirience and i expected some other type of discussion.Thank you all for efforts to help (me/us all) to scratch the surface.I am gratfull for any comment posted here.Wish u all the best.
 
#30
Mikelandjelo said:
I thought that we will have here constructive dsicussion but as far as i can see we cant.
Click to expand...

Well, barging in and saying "I am going to make the best amplifier in the world, now can you tell me how to do it?" is maybe not the best way to encourage constructive discussion?

I am not DIYer.
Click to expand...
Then diyaudio is maybe not the best forum for you.

I am living from designing and making new solutions for companies.I have 18 years of professional expirience and i expected some other type of discussion.
Click to expand...
In that case I would have expected you to be familiar with spread spectrum techniques, and the Tripath implementation of class D.
 
#32
Julf ..... I am sure that u r good guy.I dont fell any positive energy from your writing so i will not include my self a lot in it.

About discussion i thought that we will discuss possible working scenarios (globaly-not in details so i can make something directly from there).

I am not DIYer BUT this is right place to discuss and share thoughts,ideas and design concepts about audio.If u think that professionals should not be here then please say that loud to all of them here.I am sure that i am not the only one.I asked you nicely to keep your thoughts in your head for a longer time before u decide to write them down....otherwise u will have war with many people here.We all r here couse on first place (for me absolutely) share same passion about audio tachnology....Did u think about that?If i said that i will make good-best class d amplifier that was not a call for a fight.Thats just what i will really do...its just a matter of time when i will be there.If that hits u somewhere then u should ask your self why that hits you so hard to try to make an enemy from me.Anyway u will not make an enemy from me whatever u decide to write.

U said that i should know all about Tripath.Why?Did u ask your self why i dont?Maybe i know a lot and think that thats not enough for something that i have in my mind to do?

I am frendly and positive and that will not change no matter what.If u personaly want to share some thoughts (global and wide ones) about this topic i will dedicate all time that i will need to your post 100% and will be glad to hear.I at least sheared some of mine but no one tried to make them deeper.

Peace.
 
Last edited:
#34
Mikelandjelo, I am sure you are a good guy too, and I am sure you can write proper English if you try.

Mikelandjelo said:
I dont fell any positive energy from your writing so i will not include my self a lot in it.
Click to expand...

I apologize for the fact that my vibrations are so anti-sympathetic.

If u think that professionals should not be here then please say that loud to all of them here.I am sure that i am not the only one.
Click to expand...
Many of us are professionals of some form or other. No problem with that, as long as we discuss and share openly.

But do you see how you barging in and saying "I am going to make the best amplifier in the world, now can you tell me how to do it?" is maybe not the best way to encourage constructive discussion?

I asked you nicely to keep your thoughts in your head for a longer time before u decide to write them down....otherwise u will have war with many people here.
Click to expand...
Thank you for your friendly advice.

If i said that i will make good-best class d amplifier that was not a call for a fight.Thats just what i will really do...
Click to expand...
Fine - it is good to have ambitious goals.

U said that i should know all about Tripath.
Click to expand...
No, I didn't. So let me give you a piece of advice in return - how about actually reading what I write, and thinking about it, before you respond?

And no, I am not just picking nits - if you don't think there is any difference between "know all about" and "be familiar with", then we have a communication issue.

I am frendly and positive and that will not change no matter what.If u personaly want to share some thoughts (global and wide ones) about this topic i will dedicate all time that i will need to your post 100% and will be glad to hear.I at least sheared some of mine but no one tried to make them deeper.
Click to expand...
Several people did, but you seem to have ignored their input.

See responses from Saturnus, PetruV, DF96 and others...

So just like Saturnus, I have to ask why you feel the need to go for a filterless design?
 
#35
We r going out of topic definately.I dont want to share reasons why i want to go that way.English is not my mother tongue language which is obvious and it can happen that i will miss something in communication.If that happens i apologize for that.This will be my last post in this topic.If i wrote something offensive or wrong i apologize again.Wish u all the best and inovative designing ideas in the future.
 
#36
Mikelandjelo said:
We r going out of topic definately.
Click to expand...

I agree.

I dont want to share reasons why i want to go that way.
Click to expand...
Then it becomes hard to discuss it.

English is not my mother tongue language
Click to expand...
And just to be clear - it is not my mother tongue either. In fact, it is my third language, and I now live in a country where they speak my fifth language, so that doesn't help either.
 
#37
vacuphile said:
Except for saying that it is the fast switching pulses which create all the rf hash, and not the switching frequency per se.
Click to expand...
Yes, the fast pulses create the RF hash and the switching frequency determines exactly what frequencies they appear. This fact can be used to determine the switching frequency by looking at the spectrum of the RF hash - can be used to find the source of interference.

Mikelandjelo said:
I have 18 years of professional expirience and i expected some other type of discussion.
Click to expand...
I'm not sure what sort of discussion you were hoping for. Encourage you to do something which is pointless, by using some technique (other than spread spectrum) which doesn't work?

Is there any constructive thought how filterless output can be achieved (in theory) to work?
Click to expand...
We have told you: spread spectrum. It doesn't remove the RF hash (for that you need a filter), but it does spread it around. Having smoother pulses simply moves the problem to lower RF frequencies and negates the whole point of Class D (efficiency).

You still haven't told us why you want to do filterless Class D. It has no audio advantage, and considerable RF disadvantage so why do it?
 
#38
Sir, please, please, do not leave ... 🙁
We were only joking, ho ho! Here is the start of what you asked for ...

class d filtless 3110 1.jpg

class d filtless 3110 2.jpg

class d filtless 3110 3.jpg

from

http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=tpa3110d2&fileType=pdf

Ha, ha ha. We were just having a bad day here at diyaudio, :nownow: we're not usually like this, no sir! 😱
 
Last edited:
#39
Last edited:
Status
Not open for further replies.
Top Bottom