Lars Clausen said:It is, but that doesn't mean i am nailed to the building
So i have quit, and now doing only lab work, development, experiments etc. Which i enjoy a lot !
Hi Lars,
So you've hinted at your recent work a few times now extremely vaguely other than to say it solves the inherent problems known with class d today, will be the amp of tomorrow.
What in your opinion would those inherent problems with class d today be? I can only assume we're not talking about post filter feedback or other foolishness.
It is my belief the inherent problems which kept it from being a viable high end option have been left in the past already, or are you attempting to hint at a whole new level, like with lossless switching at 5MHz or something ? Maybe you're playing around with those new Mueta samples?
Just curious
Regards,
Chris
Hi Chris: No it's the good old well known problems, like dead time, switching time, filter distorsion etc. I just had time now to find some new and very effective solutions, and the result has simply lifted the sound quality through the roof!
But in the same moment i must say that we could not benefit from using post filter NFB. It simply didn't contribute with anything useful in this design. In fact we couldn't even preserve the open and precise sound stage when the NFB was moved to after the output filter, with a simple switch. It simply destroyed the dynamics!
This effect is not especially known from UcD but also in many many other post filter NFB designs. Of course i must admit post filter NFB dampens some of the midrange harshness found in earlier Class D designs. (Including my own). But post filter NFB still (IMO) cost too much in terms of dynamics, openness, soundstage etc.
And in the new designs, midrange harshness is completely a thing of the past! When you listen to it you are not thinking Class D, you are thinking Class A, Single End, Tubes etc. But with ultra high precision, and rock solid dynamics in the bass.
From other threads here on the forum i think it would be fair to say that the sound quality of ZAPpulse 2.3SE vs. UcD400 AD roughly balances, maybe with a small win for UcD because of little more sweetness in the midrange. On the other hand bass dynamics and treble resolution is slightly better on the 2.3SE.
But i have seen several test teams testing ZAPpulse 2.3SE vs. 700XE and they all sayd that 700XE is 10 times better in terms of micro detail, midrange warmth, resolution, precision etc. Also when compared with some very expensive Class A amplifiers!!! So i think it's a big step forward for Class D.
But in the same moment i must say that we could not benefit from using post filter NFB. It simply didn't contribute with anything useful in this design. In fact we couldn't even preserve the open and precise sound stage when the NFB was moved to after the output filter, with a simple switch. It simply destroyed the dynamics!
This effect is not especially known from UcD but also in many many other post filter NFB designs. Of course i must admit post filter NFB dampens some of the midrange harshness found in earlier Class D designs. (Including my own). But post filter NFB still (IMO) cost too much in terms of dynamics, openness, soundstage etc.
And in the new designs, midrange harshness is completely a thing of the past! When you listen to it you are not thinking Class D, you are thinking Class A, Single End, Tubes etc. But with ultra high precision, and rock solid dynamics in the bass.
From other threads here on the forum i think it would be fair to say that the sound quality of ZAPpulse 2.3SE vs. UcD400 AD roughly balances, maybe with a small win for UcD because of little more sweetness in the midrange. On the other hand bass dynamics and treble resolution is slightly better on the 2.3SE.
But i have seen several test teams testing ZAPpulse 2.3SE vs. 700XE and they all sayd that 700XE is 10 times better in terms of micro detail, midrange warmth, resolution, precision etc. Also when compared with some very expensive Class A amplifiers!!! So i think it's a big step forward for Class D.
Hi Lars,
Thank you for that reply.
So it's safe to assume then that it is the 700XE you speak of when you mention the amp of tomorrow?
Or are you furthering that research for something new? (Always I hope).
The loss of dynamics /openess or whatever you see with post filter feedback could be a limitation of the one design.
How do you compensate for the load effectively without taking feedback at the output? Does not seem possible.
You mention a slight edge to UCD in that it has a sweeter midrange, let's assume it not ideal and susceptible to the all the usual passives/active input stage/PSU.. same as any other amp.
Can you imagine the topology as a pure, non commercially implemented amp? You can run it differentially straight off the comparator, demonstrating extreme control over the load from 0 to 20kHz + with the kind of vibrant detail that slaps you across the face across the same frequency range, and holds the same detail and character for a 400W 3 way 4ohm as it does for on a 3W 8ohm or a half watt 8 ohm. Mind blowing really.
I found core material played a roll here, likely the Q of it and nothing seems to beat an air core, just as a baseline of clearity.
I can't help thinking if you tried that kind of experiment you'd reconsider post filter feedback?
Regards,
Chris
Thank you for that reply.
So it's safe to assume then that it is the 700XE you speak of when you mention the amp of tomorrow?
Or are you furthering that research for something new? (Always I hope).
The loss of dynamics /openess or whatever you see with post filter feedback could be a limitation of the one design.
How do you compensate for the load effectively without taking feedback at the output? Does not seem possible.
You mention a slight edge to UCD in that it has a sweeter midrange, let's assume it not ideal and susceptible to the all the usual passives/active input stage/PSU.. same as any other amp.
Can you imagine the topology as a pure, non commercially implemented amp? You can run it differentially straight off the comparator, demonstrating extreme control over the load from 0 to 20kHz + with the kind of vibrant detail that slaps you across the face across the same frequency range, and holds the same detail and character for a 400W 3 way 4ohm as it does for on a 3W 8ohm or a half watt 8 ohm. Mind blowing really.
I found core material played a roll here, likely the Q of it and nothing seems to beat an air core, just as a baseline of clearity.
I can't help thinking if you tried that kind of experiment you'd reconsider post filter feedback?
Regards,
Chris
Hi Chris: I have done a lot of experiments with post-filter NFB, but no matter what i get the same negative results every time: lack of dynamics and the top sounds like the membrane is a plastic cup. (In comparison) I could never get a natural sound from a post filter NFB amplifier - my own or other's.
So i will likely never go down that dead end road.
You may have let someone convince you that load dependency is a big problem that can only be fixed by the post filter NFB. But just consider this: you have a 12-15 uH choke in series with the signal in the amplifier. Later when you are in the loudspeaker you have coils in the order of 470 uH and up in series with bass and midrange, plus the inductance of the drivers themselves. They typical are in order of 70-100 uH for a dome tweeter. So maybe you can see from this that the 12-15 uH inductance is really a very small problem. The driver coil falls out at 5-8 times lower frequency compared the amplifier coil. So does it matter?
You can compensate it out with use of feedback, but like i said you end up paying on other accounts.
In short if the feedback loop can see back EMF from the loudspeaker (like in post filter NFB) you always lose bass dynamics. This works in any kind of amplifier also Class D.
The long and short of this discussion is, that if you measure THD and frequency resonse of both types of amplifiers, with a typical loudspeaker as a load, you get THD in order of 1-2% (also on post filter NFB) and a frequency resonse that is the same.
About air coils, this is another wrong assumption. We started out in L C Audio using air coils, because we made the same wrong assumption in the beginning (2000). But after a lot of experimentation, and listening tests we found that a good ferrite coil will always be better, and if designed correctly, will have no mentionable distortion.
There are some people who have studied this particular area all their life, and i have of course got some good tips, from some of the best experts in the world.
So i will likely never go down that dead end road.
You may have let someone convince you that load dependency is a big problem that can only be fixed by the post filter NFB. But just consider this: you have a 12-15 uH choke in series with the signal in the amplifier. Later when you are in the loudspeaker you have coils in the order of 470 uH and up in series with bass and midrange, plus the inductance of the drivers themselves. They typical are in order of 70-100 uH for a dome tweeter. So maybe you can see from this that the 12-15 uH inductance is really a very small problem. The driver coil falls out at 5-8 times lower frequency compared the amplifier coil. So does it matter?
You can compensate it out with use of feedback, but like i said you end up paying on other accounts.
In short if the feedback loop can see back EMF from the loudspeaker (like in post filter NFB) you always lose bass dynamics. This works in any kind of amplifier also Class D.
The long and short of this discussion is, that if you measure THD and frequency resonse of both types of amplifiers, with a typical loudspeaker as a load, you get THD in order of 1-2% (also on post filter NFB) and a frequency resonse that is the same.
About air coils, this is another wrong assumption. We started out in L C Audio using air coils, because we made the same wrong assumption in the beginning (2000). But after a lot of experimentation, and listening tests we found that a good ferrite coil will always be better, and if designed correctly, will have no mentionable distortion.
There are some people who have studied this particular area all their life, and i have of course got some good tips, from some of the best experts in the world.
Lars Clausen said:Hi Chris: No it's the good old well known problems, like dead time, switching time, filter distorsion etc. I just had time now to find some new and very effective solutions, and the result has simply lifted the sound quality through the roof!
But in the same moment i must say that we could not benefit from using post filter NFB. It simply didn't contribute with anything useful in this design. In fact we couldn't even preserve the open and precise sound stage when the NFB was moved to after the output filter, with a simple switch. It simply destroyed the dynamics!
This effect is not especially known from UcD but also in many many other post filter NFB designs. Of course i must admit post filter NFB dampens some of the midrange harshness found in earlier Class D designs. (Including my own). But post filter NFB still (IMO) cost too much in terms of dynamics, openness, soundstage etc.
And in the new designs, midrange harshness is completely a thing of the past! When you listen to it you are not thinking Class D, you are thinking Class A, Single End, Tubes etc. But with ultra high precision, and rock solid dynamics in the bass.
From other threads here on the forum i think it would be fair to say that the sound quality of ZAPpulse 2.3SE vs. UcD400 AD roughly balances, maybe with a small win for UcD because of little more sweetness in the midrange. On the other hand bass dynamics and treble resolution is slightly better on the 2.3SE.
But i have seen several test teams testing ZAPpulse 2.3SE vs. 700XE and they all sayd that 700XE is 10 times better in terms of micro detail, midrange warmth, resolution, precision etc. Also when compared with some very expensive Class A amplifiers!!! So i think it's a big step forward for Class D.
Hi Lars: It sounds like the 700XE has completely resolved the midrange leanness/edginess that the 2.3SE can have. Is that correct?
It also sounds like the 700XE has dramatically better detail and focus. Is that correct?
When will the 700XE be available?
OK Jan Peter, we can bash each other with words all night, which could be kind of fun .. but i don't think people want to waste their saturday evening listening to us.
The Class D technology is still advancing, and on many fronts improvements are being made. And i do admit that the topology i use may expose the early childhood flaws, while your topology has a tendency to disguise the same flaws. I am of course talking about deadtime distortion.
But in the same way, when this and some other hurdles are solved, i think you will find that your topology still disguises something, while my topology reveals everything. Now i am talking about music.
And i am sorry to say, any way you look at it post filter NFB has some inherent limitations that you can never solve. And it is my prediction that post filter NFB will not be used for quality demanding audio in 5-6 years, when solutions for deadtime distortion, switching errors etc. are widely known in the industry. Simply because i have heard the difference.
But again lets talk about this when the market has actually listened to the 700 XE (and other amplifiers with the new core) and compared to your best UcD. Then i don't think you look at your self and think you are the king anymore
(Sorry i just couldn't resist a small hit on your royal highness)
Jan-Peter i hope you have a nice saturday evening, and instead of sitting on the computer, why not take a couple af Heineken, and call Bruno to go to a pub and spot some nice dutch women. I know they are very nice and good looking
Cheers from Lars
The Class D technology is still advancing, and on many fronts improvements are being made. And i do admit that the topology i use may expose the early childhood flaws, while your topology has a tendency to disguise the same flaws. I am of course talking about deadtime distortion.
But in the same way, when this and some other hurdles are solved, i think you will find that your topology still disguises something, while my topology reveals everything. Now i am talking about music.
And i am sorry to say, any way you look at it post filter NFB has some inherent limitations that you can never solve. And it is my prediction that post filter NFB will not be used for quality demanding audio in 5-6 years, when solutions for deadtime distortion, switching errors etc. are widely known in the industry. Simply because i have heard the difference.
But again lets talk about this when the market has actually listened to the 700 XE (and other amplifiers with the new core) and compared to your best UcD. Then i don't think you look at your self and think you are the king anymore
(Sorry i just couldn't resist a small hit on your royal highness)
Jan-Peter i hope you have a nice saturday evening, and instead of sitting on the computer, why not take a couple af Heineken, and call Bruno to go to a pub and spot some nice dutch women. I know they are very nice and good looking
Cheers from Lars
Lars, if you don't work for lcaudio anymore, so maybe "700xe" developed not by you personally? BTW, "Lowest THD in the industry!" etc, what about -3db(350W@4ohm) THD? If claimed only 1,10 or 100w level, then i'll (sure that Bruno can better else) win the "700xe" too easy (.00075%@15w/4ohm .0014%@25w THD/Freq=Flat), with the TDA8939 in the any variations - UcD or hystersis based modulator.
Attachments
Hello Ivan.
It's true i don't work in L C Audio anymore. But i did finish the dev work on the 700XE before i left, so it's all mine
About 'lowest THD in the industry' i did comparison of all the commercially available amplifiers i could find on the market, and compared with them. On this field the 700XE does have the lowest THD in the industry.
Of course i cant speak for all the imaginary fantasy products on the forum Or can you tell me where i can buy that amplifier you are referring to in your post?
It's true i don't work in L C Audio anymore. But i did finish the dev work on the 700XE before i left, so it's all mine
About 'lowest THD in the industry' i did comparison of all the commercially available amplifiers i could find on the market, and compared with them. On this field the 700XE does have the lowest THD in the industry.
Of course i cant speak for all the imaginary fantasy products on the forum
Or can you tell me where i can buy that amplifier you are referring to in your post?
Lars Clausen said:Jan-Peter i hope you have a nice saturday evening, and instead of sitting on the computer, why not take a couple af Heineken, and call Bruno to go to a pub and spot some nice dutch women. I know they are very nice and good looking
Cheers from Lars
Nice Dutch women????? Where in heavens name Lars?
Lars Clausen said:
In the same place where we could be buy all others imaginary fantasy products (eg "700XE")
i guess in the future? )) Actually, PRO must be able to fast detect degree of the claiming realism, and they will not see something wrong here. I'm still just a DIYer, but i expect, that "700XE" will give much more THD at 350w/4ohm vs .015%@100w, and THD/FREQ will not flat. So what, .5%? Again? Attachments
Lars Clausen said:
Yes exactly! Both questions...
I think it will be available in about 5 weeks. But still i don't work in LC any more, so i can't be sure.
Hi Lars: So, 700XE has dramatically better detail and focus, then the 2.3SE, and has completely resolved the midrange leanness/edginess that the 2.3SE can have. That's sounds very exciting.
Lars, the 2.3SE can be a little cold or lacking in high frequency harmonic extension/detail/micro detail. How is the 700XE in this area (ie. AIR?)?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.