phase_accurate said:
Bricolo said:
In general, these are accurate statements. But the open loop performance of a comparator followed by a switching power stage is pretty awful by itself
. But thanks to feedback and noise shaping we can get good fidelity in the audio range anyway.That's why I was wondering how much difference the filter coil makes in the end, when feedback is taken from the filter output...
Hi,
Makes a huge difference in a UCD type of design..
but feedback wise ...having the filter inductor included in the loop is key to the high performance as noise generated within it/picked up by it is now compensated for.....that was the holy grail we were all wanting a little while back right?
Non linearity wise of the core....hmmm well that's included in the feedback loop as well, and so more easily minimised over the EMI radiating all over the rest of the circuit as it seems an air core would be. It should never saturate anyway right?
Given the major advantage by having it inside the feedback loop, simplicity of said loop, when compared to noise shaping schemes, lower output impedance..part count..load independance...just far too many huge advantages to not want to include it, I think anyway. Before we had that class d was "good for sub woofer applications".
Regards,
Chris
Makes a huge difference in a UCD type of design..
but feedback wise ...having the filter inductor included in the loop is key to the high performance as noise generated within it/picked up by it is now compensated for.....that was the holy grail we were all wanting a little while back right?
Non linearity wise of the core....hmmm well that's included in the feedback loop as well, and so more easily minimised over the EMI radiating all over the rest of the circuit as it seems an air core would be. It should never saturate anyway right?
Given the major advantage by having it inside the feedback loop, simplicity of said loop, when compared to noise shaping schemes, lower output impedance..part count..load independance...just far too many huge advantages to not want to include it, I think anyway. Before we had that class d was "good for sub woofer applications".
Regards,
Chris
Quite precisely. That's why I think the most important thing is getting the comp/PS/filter done right. The control loop can be saved for last. You could even consider not placing the modulator on the same PCB with the power stage so you have a Universal Test Mule to try anything you like.classd4sure said:
Bricolo said:
In fact, here is one
http://www.diyaudio.com/forums/attachment.php?s=&postid=418321
I'll take the blame for it. Don't trust those values, but it should give you the idea of it.
UCD is simply a hysteresis free phase modulator with a dual purpose single feedback loop which is taken after the filter coil. Circuit shown is right off the patent and is excellent if done right.
Much of it can be replaced by a single differential comparator.
Chris
classd4sure said:
I even have difficulties to guess the part symbols, so don't worry about the values, I won't trust them since they are too small
Does this qualify as UcD?
But In fact I wanted to see something more simple, with block diagrams (integrator, comparator...) no discrete for the momment
A nice starting to understand how Class-D Amplifiers work is to read this tutorial from IRF;
http://www.irf.com/product-info/audio/classdtutorial.pdf
Regards,
Jan-Peter
www.hypex.nl
http://www.irf.com/product-info/audio/classdtutorial.pdf
Regards,
Jan-Peter
www.hypex.nl
phase_accurate said:
It's even using the original circuit from the UcD patent !
Regards
Charles
Nice
But take care, I remember (with some PASS patents) that patent text and schematics aren't really made to help people (and DIYers
) to understand the principe. Most of the time it's quite the opposite. And the patent's schematics aren't "optimal integration of the theory" not "ready to work" circuits.But since we have the inventor here, maybe things will be easier
Hi,
Originally this was a starting point thread (I thought), but it seems people need at least 100 watts with great quality..that's not starting point anymore.
Had you been happy with say 20watts, you could be listening to it play really bad in about ten minutes flat, probably for under 10$, learn all about the how's and the why's....and move on to bigger and better things. That's what a reference is after all.
The way this has turned now, It doesnt' seem anyone has any concern over that, and want a good "this is how it's done right" design.
Whatever the end result winds up being I've no doubt will be documented sufficiently.
Things like "what is class d" probably shoulnd't be covered in more than a sentence or two, see google.
Try looking for "phase shift oscillator", same sort of idea. I can link you to the patent it's got the block diagrams and that but....I'm too burnt to dig it up right now, will get it for you later.
Personally, I'm still all for keeping this a class d beginners ideal circuit (cheap, quick, kind of works, learn a heck of alot!)
Once you decide you like it or not, you can decide to start spending more time and money on it....but someone wanted 100w
Regards,
Chris Crashed
Bricolo said:Please, let's not start with a "experienced users only" discussion
this is a starting point thread, maybe we should first explain and define class D, the parts needed, the topology, the theory. I'm new in class D and I begin to get lost with your technical discussion
Originally this was a starting point thread (I thought), but it seems people need at least 100 watts with great quality..that's not starting point anymore.
Had you been happy with say 20watts, you could be listening to it play really bad in about ten minutes flat, probably for under 10$, learn all about the how's and the why's....and move on to bigger and better things. That's what a reference is after all.
The way this has turned now, It doesnt' seem anyone has any concern over that, and want a good "this is how it's done right" design.
Whatever the end result winds up being I've no doubt will be documented sufficiently.
Things like "what is class d" probably shoulnd't be covered in more than a sentence or two, see google.
Try looking for "phase shift oscillator", same sort of idea. I can link you to the patent it's got the block diagrams and that but....I'm too burnt to dig it up right now, will get it for you later.
Personally, I'm still all for keeping this a class d beginners ideal circuit (cheap, quick, kind of works, learn a heck of alot!)
Once you decide you like it or not, you can decide to start spending more time and money on it....but someone wanted 100w
Regards,
Chris Crashed
classd4sure said:Hi,
Originally this was a starting point thread (I thought), but it seems people need at least 100 watts with great quality..that's not starting point anymore.
Had you been happy with say 20watts, you could be listening to it play really bad in about ten minutes flat, probably for under 10$, learn all about the how's and the why's....and move on to bigger and better things. That's what a reference is after all.
The way this has turned now, It doesnt' seem anyone has any concern over that, and want a good "this is how it's done right" design.
Whatever the end result winds up being I've no doubt will be documented sufficiently.
Things like "what is class d" probably shoulnd't be covered in more than a sentence or two, see google.
Try looking for "phase shift oscillator", same sort of idea. I can link you to the patent it's got the block diagrams and that but....I'm too burnt to dig it up right now, will get it for you later.
Personally, I'm still all for keeping this a class d beginners ideal circuit (cheap, quick, kind of works, learn a heck of alot!)
Once you decide you like it or not, you can decide to start spending more time and money on it....but someone wanted 100w
Regards,
Chris Crashed
Here's the patent
http://l2.espacenet.com/espacenet/viewer?PN=WO03090343&CY=ch&LG=en&DB=EPD
What I want, is to understand how what I build works. That's all... (maybe that's not as easy as it seems)
As for the power, can't we simply agree on a design, and adapt the PS rails (and the output mosfets) for more or less power?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.