Zero crossing PWM, I remember having read a paper on something like that and I think it is the following. Pictures are included in the second link. Is the Class O modulation based on something similar ?
https://ieeexplore.ieee.org/document/4341577
https://www.semanticscholar.org/pap...-Liu/e9d588edfc3786ded5e0c53038b16d075fe78b3f
https://ieeexplore.ieee.org/document/4341577
https://www.semanticscholar.org/pap...-Liu/e9d588edfc3786ded5e0c53038b16d075fe78b3f
I am not able to read some of these articles but it sound like ZPWM is what they talking about. if so that is what is used in class O, it is like Class B and Class D but it is stand alone class of amplification, as it is not using standard PWM like Class D, as for Class B it uses identities positive and negative going pules that comes from the ZPWM.
Advanced Audio Amplifier
Analog Processor Core
Pulse Function
New Class O Amplifier
+ others
Here some information on my research and development work since 2017, I tried to keep it in a basic form.
Advanced Audio Amplifier
Analog Processor Core
Pulse Function
New Class O Amplifier
+ others
Here some information on my research and development work since 2017, I tried to keep it in a basic form.
The OP maybe interested in knowing the methodology behind your techniques, but would you really disclose them ? Unfortunately, I have difficulties following your articles, mainly due to inconsistencies in terminology and the absence of diagrams. For example, what do you imply when you say "identities" ?
Yes, it is not easy get this information across, some people like a lot of details as other want an simple over view of the subject therefore I kept it short and to the point. If would like more information, maybe best to send me a message via LinkedIn? and I can provide you with more detail.
The last time I tried, I could upload pdf files, and I think that was after the forum migration. File types that are not supported can always be zipped and uploaded, but pdf files can just be uploaded directly.
I didn't mean to ask you for any classified / protected information on your techniques, but would just like to know what "identities" in your above post means.
A little more on Class-0 (class zero), not sure if that's the same as what is being said here, seems to be a linear amplifier.
https://ieeexplore.ieee.org/document/7062915
Last edited:
Thanks for that, here is PDF file of my articles, *note there are nothing more than basic over view of what I been working on. I still working my way through how I would design an audio amplifier, that is why I am here to get some useful feedback.
Grantwt:
It's difficult to get any feedback from a forum like this, until you disclose your methods openly. Besides, if your method is indeed "cutting edge", then it may be best to submit the same as a research paper to a reputed, peer-reviewed international journal in Electrical Engineering, such as those run by the IEEE. That way, the feedback/comments you would be receiving (after your paper has been peer-reviewed) would also be ultimately professional, as reviewers are usually chosen for their expertise in the relevant field of study (switching power conversion in this case).
All the best.
It's difficult to get any feedback from a forum like this, until you disclose your methods openly. Besides, if your method is indeed "cutting edge", then it may be best to submit the same as a research paper to a reputed, peer-reviewed international journal in Electrical Engineering, such as those run by the IEEE. That way, the feedback/comments you would be receiving (after your paper has been peer-reviewed) would also be ultimately professional, as reviewers are usually chosen for their expertise in the relevant field of study (switching power conversion in this case).
All the best.
Good point, the way I work is to develop working prototypes, make available for testing, that is how I make my discoveries. As for a research papers it not something I get involved in, if you would like to do one, I can provide with all information on how this topic of switching amplifier design and why it functions in the way it does. As there is a lot to this area, for example there are two very important aspect to this, they are the pulse function and the processing function blocks. Another thing is the way classes of amplification is done, seems poorly thought out as it seems everything is a form of Class D which I can show that this is the wrong approach, as class should be partly based on their pulse function, as there is 4 + 1. They are SPWM, ZPWM, PPM, and PLM, the plus one is PDM as it has function, but it not directly related to the other four. As for PCM it has no pulse function as it just has value representing an amplitude component.
Anyway getting onto more interesting things, I have now workout how layout of this this new switching amplifier design, and I am now making a start on doing some of VHDL code for the analog processing core, from all the research I have done on this topic, DAC integrated switching amplifier is 100% doable, looks Iike I will be working on this project for the next few years.
Anyway getting onto more interesting things, I have now workout how layout of this this new switching amplifier design, and I am now making a start on doing some of VHDL code for the analog processing core, from all the research I have done on this topic, DAC integrated switching amplifier is 100% doable, looks Iike I will be working on this project for the next few years.
Well, it's not enough that you believe the convention to be wrong, as anything in science needs to be properly backed up with enough proof. Note that even people like Albert Einstein have written papers, attended conferences and successfully answered all kinds of questions from reviewers.
So, don't you want you ideas to be reviewed by any leading experts in the field? I ask that because that's what people primarily write papers for. The credit and copyright are all only secondary. Further, all research submitted to reputed journals must include experimental results.
From the document you have posted I don't feel like you want to disclose anything at all. However, if I'm wrong and you indeed intend to, please do so here publicly, on the forum, so that its members (like the OP in this thread) could benefit from your findings.
People primarily write peer-reviewed papers because they work at university research groups that don't get money unless they publish peer-reviewed papers. Companies sometimes get involved with it because they either have a deal with a university research group or desperately need the publicity.
Einstein published papers because he didn't want to keep working as a patent clerk. Papers on theoretical physics usually don't include experimental results for obvious reasons.
Einstein published papers because he didn't want to keep working as a patent clerk. Papers on theoretical physics usually don't include experimental results for obvious reasons.
It is all public, except what I am current working on over the last six months, If you would like more information what parts would like to know more about?
Circuits I will make public once I test the prototype, as for block diagrams I can draw something up showing how I plan to build this amplifier, but still very much in the research stage at moment. As current prototypes they are all radio frequency modulator / amplifier designs using Class Q, that is designed to work with QAM (quadrature amplitude modulation), that would have very interest here. This is my first time looking at using this switching technology for an audio amplifier design.
If would you would like I could start a new topic, where I can post updates on parts of this project as I work through putting it all together, as will testing each part of circuit, once that is done then I make public that part of circuit. As posting circuits at this stage there would issues with them as may not work, so this is long term project so will time for work it all in spare time. Inventing new technologies very time consuming, and I expect I will end with up with a lot of failures a long the way.
Circuits I will make public once I test the prototype, as for block diagrams I can draw something up showing how I plan to build this amplifier, but still very much in the research stage at moment. As current prototypes they are all radio frequency modulator / amplifier designs using Class Q, that is designed to work with QAM (quadrature amplitude modulation), that would have very interest here. This is my first time looking at using this switching technology for an audio amplifier design.
If would you would like I could start a new topic, where I can post updates on parts of this project as I work through putting it all together, as will testing each part of circuit, once that is done then I make public that part of circuit. As posting circuits at this stage there would issues with them as may not work, so this is long term project so will time for work it all in spare time. Inventing new technologies very time consuming, and I expect I will end with up with a lot of failures a long the way.
What I will do if I get R & D finding, I will write peer-reviewed papers, but been outside of any university research groups, this is almost impossible to get any R & D funding, it is just the way this system works.
It is not always money. Sometimes, it's about a concept and researchers do submit certain articles just to get feedback on how and where they stand in comparison to existing ones. Now, if the feedback is encouraging (to the authors in any way), they would proceed with the deeper aspects of their research, or else leave it (or put it aside) so that no more time/resources are wasted. Also, note that all reputed engineering journals / letters / transactions need experimental results to avoid desk rejection, of course with the only exception of survey papers.
Actually, Einstein took the menial clerk job so that he would have enough spare time to develop his theories. And he wrote papers because that was the only platform that allowed his groundbreaking concepts to have the serious attention of the scientific society (e.g. writing a column in a magazine would not have sufficed).
https://oxsci.org/einstein-at-the-patent-office/
Now, in this case, Grant says that conventional amplification is the "wrong approach", but whether that's correct (or not) is something that we'll be knowing about only later. However, my opinion is that such "big" ideas (that challenge the norm) need to be raised before a larger (and more science-speaking) audience, like at the next AES convention, instead of being limited to internet forums. Of course, such ideas would see severe discouragement at first, but as long as the "inventor" can explain himself convincingly, there should be no problems regarding the eventual acceptance of his/her ideas.
All parts, to be fair, but how about starting with the term "identities" for now, as already asked for, in post #26 ?
Okay, "identities" is the relationship between positive and negative parts of the sampled waveform, this can be used with a Sine / Cosine function this could be phase or amplitude information. Where amplitude +1 or -1 as the outer circle always has a value of 1, as this uses logic levels, for phase +π or -π, as this works with both real and imaginary values. So therefore the type of pulse used will have a victor component, for audio applications this is not important as there is only an amplitude part that is used. When it comes radio frequency applications, such as a signal balanced modulator, or with identities information used with dual vector processing, where it is possible to build a double balanced modulator out of logic gates. As with all switching amplifiers they are not just an amplifier, but they are a form modulator / amplifier. So when it comes to Class O as it is used for audio, the identities have little use, the outcome would much the same using two locked SPWM pulses, one that is positive going and the other that is negative going configuration, as in dual Class D. So therefore Class O can be used as analog processing block, along with others as building blocks to create a more complicated wave forms, this is where analog processing core comes into play. For this to work there are two forms of modulation one is the pulse function that will be some form of pulse modulation, and radio frequency modulation that is present after the low pass file stage. If you would like will do some diagrams show how this process works?
Last edited:
Thanks for the above explanation, Grant.
I sincerely hope the OP manages to understand something of what's written above, and I don't think I want to trouble Grantwt by asking any more questions.
I sincerely hope the OP manages to understand something of what's written above, and I don't think I want to trouble Grantwt by asking any more questions.
No problems, I have started a new topic called, the DAC switching Amplifier, I have also uploaded files of what I have done so far.
With the huge advancement made in computing power since the first DACs came alone, wouldn't it be possible to take a totally different approach; let's say you feed an analogue signal to a ADC, and use a computer to work out an algorithm to convert the digital signal to precisely match the original analogue (for example you could perhaps use "beats" in the high frequencies to detect frequencies higher than the sample frequency), then you output the signal as a PWM or similar, drive speakers with it, then you anouther computer to find the discrepancy with the output of the speakers with the original analogue signal, and modify the algorithm to account for the distortion in the power amplifier stage (for instance the effect of the low pass filter on the output of the amp.