Sound Quality Vs. Measurements

Status
Not open for further replies.
My worry about class D is that it is a new toy . A toy which might put people off of trying to learn . My friend who has PA equipment says the sound of class D is fine and the weight advantages critical . We are talking 3000 W sometimes . As he says 40 or 25 Kg is a big deal . He has a slight preference for AB excluding the weight . Some of the class D story is a myth . 78.5% verses 95% is not such a big deal . Class H is not as complex as D and for real music makes it a narrow margin of difference . I noticed Philips state crest factors of 10 in the application notes . Obviously this expects the average TV to be shacking to pieces long before the heat-sink becomes an issue . In old fashioned class B in vinyl days a crest factor of 6 was advised and 3 for PA . In my work crest factors of 0.5 are wise ( reactive loads ) . Thus what Philips recommend and I can use are miles apart . The problem I have is I can not shift the heat from such a small area albeit less than if class AB. My brother had a simple rule , always use TO3 . now the big plastic packages compete . Class D , fools gold ?
 
Hi Nigel,

I would just like to point out that class AB only has 78.5% efficiency when driven to max output. In normal situations the efficiency is much lower, usually around 30 - 40%. Class D does much better in this regard. But as always there are choices to be made, you can tune class D amps for maximum performance or maximum efficiency, (idling time, gate drive). The Hypex nCore white paper is an interesting read about state of the art class D amps: http://www.hypex.nl/docs/papers/ncore wp.pdf
 
Mark you are absolutely right . It is the industrial applications that need stressing as in my book if something says it is 100 watts I expect it to offer 5 years continuously at 100 watts ( 18 months is an undisclosed industry standard in things friends do , contracts exchanged on that basis , modules replaced each year when critical ) . When is a watt not a watt ? When it is VA or music . Like red-lines on engines to me it matters most , this is the same . I seek very high reliability in what I build . As simply put as I can the die area of any device is where my quest begins .
 
I think class D is like GM food . Logical but ..... ? In normal plant research GM ideas have been used . Not GM but much faster research times using ideas common to both GM and standard . My objection to GM and class D is it transfers power of choice away from me . On the farm where I worked as a boy we saved seeds from broad beans , now it is illegal unless a club and not for profit .
 
Well it is simple . The broad beans is how life was for thousands of years . Sensible farmers kept the best to grow on , silly ones grew the less good ones . From this we learned selection . Brother Mendel went on to show how this would be made into a science ( peas ) . Now some rightly or wrongly move to outlaw what was common practice for years . One has to ask why ? The parallel is if we loose the skills to make the internal parts of things we built how long before we become just consumers ? Metaphors , the world was full of them once . Time will come when no one remembers them or how to use them ?

Is saving seeds Illegal? - Yahoo! Answers
 
dvv,

LOL, I used the terminology that applies to woofers (long throw) without realizing it.
you are right about the engines. I'm familiar with motorcycle engines which are different in enough ways and I extrapolated from there, my mistake. the current tendency with motorbikes is to use oversquare engines though. I think it has a lot to do with the higher rpm's reached (my bike has a 12000 rpm redline and it's not even a new engine design, it's basically a pre-90's oil-cooiled engine with some upgrades) and the larger forces the conrod sees at stroke ends, with today's motorcycles rpm's of 16000 are not uncommon and I'm referring to common production models. may also have something to do with geometry which is more restrictive with bikes?

anyway, back to D amps.
well yes, at least some manufacturers seem to design by measurement. but a 50k$ Mark Levinson? am I missing something? a 50k$ amp does not address exactly to the same public that buys a $50 DVD player based on the number of formats it can read or number of different inputs or outputs. or the 20k TAD or even the "cheap" Devialet? these are respected manufacturers with a long history, what would make them employ a paradigm shift?

(this is all assuming that these amps are indeed somehow inferior to their linear counterparts which is not a given but speculation based on reviews)
 
Mr PP and others . Is there a class D amp which can be built " dead bug " and is > 90 % efficient and < 1% THD . It would be great to have something to work on . I always suspected a NE555 and some FET's should be able to get us somewhere . Hypex is too much like a cake mix for me ( look mummy I made a cake ) .

Google tells me this site is a risk ( Trojan horse or the like , policeman sign says " get me out of here " ) . hope me being here is sensible ? Other than that we are a threat to world order and we are being censored , great !
 
diyAudio Member RIP
Joined 2005
Mr PP and others . Is there a class D amp which can be built " dead bug " and is > 90 % efficient and < 1% THD . It would be great to have something to work on . I always suspected a NE555 and some FET's should be able to get us somewhere . Hypex is too much like a cake mix for me ( look mummy I made a cake ) .

Google tells me this site is a risk ( Trojan horse or the like , policeman sign says " get me out of here " ) . hope me being here is sensible ? Other than that we are a threat to world order and we are being censored , great !
Chrome was warning me for a while about diyaudio, so I avoided it for a bit. Not sure if it was a real threat or not.
 
diyAudio Member RIP
Joined 2005
Hypex is too much like a cake mix for me ( look mummy I made a cake ) .
The genesis of Hypex, whose products use, among other later developments, Putzeys' patent that belongs to Philips, is that Bruno started off when he was with Philips to make the lowest-cost class D amp, period. At the time he did it, Philips (having not yet spun off NXP) made certain parts from which the earliest UcD designs were constructed. As a result there are more discretes than might be seen today as cost-effective, although they are still pretty cheap. But it means that board area is larger than more integrated designs. Note that you can use exactly the original reference design if you use NXP semiconductors, no licensing fees required.

The surprise, though, was how well things worked. By requiring only a small delay inserted in the loop, essentially no hysteresis, and counting on the output L-C for most of the phase shift, one makes a self-oscillating amplifier with the output filter inherently enclosed in the feedback loop. Many benefits thereby accrue. :cool:
 
Nigel,
not sure what you're getting at.
maybe I should emphasize that IMO efficiency is not the biggest advantage of class D. my electrical bill is less than GBP 35 during the winter when I turn the electrical heater on now and then. and in all honesty, likely my bike with its low mileage does a decent job at getting few species closer to extinction so I'd be a hypocrite to pretend that I'm green. ok, lecture me, it won't help LOL
it's about size and weight that translates to price. and yes, I don't have an infinite income therefore I need to take that into consideration.
about Hypex, THD and efficiency, I'm even more confused.
the standard UCD180 datasheet lists efficiency at 92%, NCORE at 93%, full power. but as we all know, at idle the output devices still switch so...
I have done distortion measurements with my implementation. at ~100W into resistive 4.7ohms, a 18k+19k intermodulation tests gave a highest product at -80dB. pretty decent. Stereophile reviewed the Channel Islands amp and their measurements pretty much confirm it (actually -70dB at 50W but there must be something wrong in their methods as I have personally tested it).
and NCORE? as per Hypex datasheet it's pretty much unequaled.
so, you don't like Hypex? cake mix? I don't get it. no I'm not a fan boy. I own one but I have no problems admitting it has its issues. in no way perfect. quite the contrary I'd like to find out what's wrong.
 
Last edited:
Looking over the sites of some digital audio amp manufacturers, I noticed that when one added all the extras, which are supposed to make them work and sound better, in the end, they turn out to be not at all cheap. While for the price they will produce nominally more power, I still feel that for the money, I can do better in analog sound-wise, even at the cost of the power output

Of course, every new technology has its gestation period, the time during which its milk teeth need to fall out and be replaced by proper teeth, and I have no doubt that the future belongs to them.

I also feel their time is yet to come. In the meanwhile, I'll stick to analog.
 
Well run it at 1 ohm and let me know ..i can do that with the ab amps here :)

Nigel,
not sure what you're getting at.
maybe I should emphasize that IMO efficiency is not the biggest advantage of class D. my electrical bill is less than GBP 35 during the winter when I turn the electrical heater on now and then. and in all honesty, likely my bike with its low mileage does a decent job at getting few species closer to extinction so I'd be a hypocrite to pretend that I'm green. ok, lecture me, it won't help LOL
it's about size and weight that translates to price. and yes, I don't have an infinite income therefore I need to take that into consideration.
about Hypex, THD and efficiency, I'm even more confused.
the standard UCD180 datasheet lists efficiency at 92%, NCORE at 93%, full power. but as we all know, at idle the output devices still switch so...
I have done distortion measurements with my implementation. at ~100W into resistive 4.7ohms, a 18k+19k intermodulation tests gave a highest product at -80dB. pretty decent. Stereophile reviewed the Channel Islands amp and their measurements pretty much confirm it (actually -70dB at 50W but there must be something wrong in their methods as I have personally tested it).
and NCORE? as per Hypex datasheet it's pretty much unequaled.
so, you don't like Hypex? cake mix? I don't get it. no I'm not a fan boy. I own one but I have no problems admitting it has its issues. in no way perfect. quite the contrary I'd like to find out what's wrong.
 
Status
Not open for further replies.