I don't see any PWM there, so it is strictly speaking not an amplifier. It is just a push-pull arrangement of 2 transistor switches, in order to made a high power high efficiency transistor based RF output that has to be modulated by other means.
I had a project in university back in 1978 where I used such an arrangement in power supply (I heard a rumor about Japanese power amp that used switched mode PS, so was eager to implement mine). I had to implement a time gap between switching of transistors in order to avoid huge current peaks when both transistors were open. High voltage transistors were sooo slooow then... 😉
A difference between comparator threshold and switching time effects on the output of a Class D amplifier vs class B xover distortion is that the latter is (mostly) displayed in a range of relatively small amplitude levels commonly occupied by program material which tends to reduce the relative efficacy of amelioration options for the latter as compared to the former. Standard crossover distortion besides showing up as elevated distortion levels at low levels also exhibits itself as instantaneous changes in damping, frequency response, phase response, stability, PS rejection and possibly gain. I doubt that any of these anomalous behaviors that could result from comparator or switch performance limitations will have such a specific and characteristic amplitude sensitivity in most Class D amplifier implementations. These particular characteristics are what makes Class B crossover distortion crossover distortion, IMO.
Last edited:
It does not matter. What matters, in class D amps cross-over distortions indeed present and can't be avoided, where in class A amps they are absent.
However, the majority of implementations of class B and class AB amps have worse impact of them on sound reproductions than have class D amps.
However, the majority of implementations of class B and class AB amps have worse impact of them on sound reproductions than have class D amps.
Well, this makes me wonder - if a comparator in a Class D amplifier has a consistent amount of hysteresis throughout its functional range, wouldn't that show up more as a loss of general detail and possibly a DC offset more than what one would recognize as conventional 'crossover distortion'? The local mechanism may be similar but the effect at the speaker terminals might be significantly different.
Also, what about A/D and D/A comparators? Couldn't a 'crossover distortion' be associated with the switch for each conversion bit depending on the slope of the output signal?
Then there is FM which depends on basically a zero crossing detector scheme. Couldn't distortions associated with this also be defined as 'crossover distortions'?
I'm just curious as to when 'crossover distortion' may appear as something different, IYO.
Also, what about A/D and D/A comparators? Couldn't a 'crossover distortion' be associated with the switch for each conversion bit depending on the slope of the output signal?
Then there is FM which depends on basically a zero crossing detector scheme. Couldn't distortions associated with this also be defined as 'crossover distortions'?
I'm just curious as to when 'crossover distortion' may appear as something different, IYO.
Last edited:
I don't care, actually, about such theoretical details. What I know, bass guitar players adore class D amps. They sound clean, loud, and are light to carry between gigs. Today I received a package from Parts Express, let's try how it will perform driving my woofers and subwoofer. 😉
I need now stiff 24V PS that would allow 400W power consumption.
I need now stiff 24V PS that would allow 400W power consumption.
Attachments
Part Express makes/sells good quality stuff. I have a Crown K1 and a Crown K2, and they are indeed pretty good in the bass, and have no glaring defects up higher, but my DC coupled OTL beats them across the whole range.😀
Last edited:
Actually you pointed out how important low noise on the power supply rails is for a switching amp.
In 74 a buddy built a PWM amp just using a 555. It was quite noisy.
The 555 was introduced in 1970 and pin 5 was used to provide PWM!
In 1970'th (I don't remember exactly the year) I used a discrete monostable multi-vibrator with voltage-controlled time constant, integrating R-C network after it for feedback, and an opamp as an error corrector. The carrier was stabilized by quartz resonator. The thingy was used as a record amplifier (actually, a class D PW-modulated HF bias source) for a tape recorder that I built cannibalizing vacuum tube studio tape recorder. 😀
I strongly believed then that the future belongs to opamps, and vacuum tubes are already obsoleted. 😀
I strongly believed then that the future belongs to opamps, and vacuum tubes are already obsoleted. 😀
Last edited:
I remember playing with the record bias oscillators for my Ampex 351 tube recorders which I had retrofitted with roller guides, Saki glass-ferrite heads and AG440 servo motor drives. I was able to push the record oscillator frequency as high as 600Khz (where it was starting to get difficult to get enough record bias for the Maxell open reel tape I liked to use) and did notice SQ and noise floor improvements until at least 400Khz. The standard bias frequency I believe was about 100Khz.
I should mention that I used no erase bias with the above scheme. What I would do is bulk erase the tape and record directly onto it afterwards. This got rid of the erase head tape hiss which was several dB higher than the record head tape hiss, but it was a cumbersome technique if I wanted to re-record over only a portion of the tape. If I had got beyond the experimental stage here, I probably would've added a second erase bias oscillator that was switchable.
Later, I actually built a tube bias amplitude modulator that provided several dB HF headroom extension at 3 3/4 and 7 1/2 ips by reducing the oscillator level when there was significant energy in the top octaves of the recorded sound based on the idea that the HF boosted record signal itself began acting as a self bias.
I should mention that I used no erase bias with the above scheme. What I would do is bulk erase the tape and record directly onto it afterwards. This got rid of the erase head tape hiss which was several dB higher than the record head tape hiss, but it was a cumbersome technique if I wanted to re-record over only a portion of the tape. If I had got beyond the experimental stage here, I probably would've added a second erase bias oscillator that was switchable.
Later, I actually built a tube bias amplitude modulator that provided several dB HF headroom extension at 3 3/4 and 7 1/2 ips by reducing the oscillator level when there was significant energy in the top octaves of the recorded sound based on the idea that the HF boosted record signal itself began acting as a self bias.
Last edited:
I don't think that this is clear at all. If you read 'The Audio Critic' back issues; link:- The Audio Critic, you will see that the change in audio reviewing all dates from the decision by a single reviewer, whose name I cannot immediately recall, to abandon measurement based reviewing in favour of 'subjective' reviewing. I can't be bothered to trawl through the magazines to unearth the individual's name, but the story is correct in essence.
It requires less stretch of the imagination on my part to envision that the situation we now find ourselves in is more a result of a credulous audience than any perceptible differences in performance which are nevertheless undetectable by instrumented test.
Obviously no audience likes to find itself described as credulous (having or showing too great a readiness to believe things), it's basically a way of calling people stupid without using the word stupid, but if you divide the population statisticallly, 50% are more stupid than the average, and 50% are less so, which makes for a considerable number of stupid people, by comparison.
Historically we enjoy the morality tale, 'Jack and the Beanstalk.' This is NOT told to children to encourage them to give the totality of the family resources over to a stranger in exchange for a handful of magic beans, but there are always those who fail to get the message. Add to this a communications medium the capacity of which to amplify rumour is sometimes not appreciated, and as I said before, even a comparatively level-headed individual can be overwhelmed by the seeming universality of agreement which can prevail at times.
w
Don't take any wooden nickels.
Now I understand that the more I accept practical truth, and the more of theoretical illusions I discard, the better are results. That's why I use vacuum tubes.
I remember when self-styled 'objectivists' were trumpeting CD's 'perfect sound forever' and casting doubt on the intelligence of all who felt differently, whether their ears or familiarity with the limitations of Redbook standard audio were the cause.
While these 'true believers' were braying away their ignorant certainty that 1930's -vintage continuous waveform measuring techniques applied at the maximum possible encodable level told everybody all they needed to know about the SQ of CD audio, the professional sound industry quietly abandoned the 12 to 16 bit 32, 44.1 and 48 KS/s standards for their insufficient SQ and robustness for 24 bit and 32 bit 96 and 192 KS/s standards, which showed that they at least had the ears and self respect not to allow themselves to be cowed by the meter readers' dogmatism.
The bad news here is that the high quality audio in the consumer market is more or less relegated to HD DVD's and Blu Ray discs (with the exception of the rare DVDA and SACD) while CDs are stuck at the technical level of the 1980's ignorami.
Meanwhile, the nihilistic meter readers succeeded in dragging most everybody else down to lossy encoding such as MP3 standards that sound like trash on a stick for the most part.
Lesson learned, I hope.
While these 'true believers' were braying away their ignorant certainty that 1930's -vintage continuous waveform measuring techniques applied at the maximum possible encodable level told everybody all they needed to know about the SQ of CD audio, the professional sound industry quietly abandoned the 12 to 16 bit 32, 44.1 and 48 KS/s standards for their insufficient SQ and robustness for 24 bit and 32 bit 96 and 192 KS/s standards, which showed that they at least had the ears and self respect not to allow themselves to be cowed by the meter readers' dogmatism.
The bad news here is that the high quality audio in the consumer market is more or less relegated to HD DVD's and Blu Ray discs (with the exception of the rare DVDA and SACD) while CDs are stuck at the technical level of the 1980's ignorami.
Meanwhile, the nihilistic meter readers succeeded in dragging most everybody else down to lossy encoding such as MP3 standards that sound like trash on a stick for the most part.
Lesson learned, I hope.
Last edited:
I used a trigger that divided by 2 bias signal. 120 KHz was used for bias, 60 KHz for erase.
TEAC used it in their cassette recorders.
Now we see at play one of the driving forces of the 'New Age' revolution, anti-science, or anti-intellectualism. By all means, let us revert to barbarism, it's a 'known fact' that we were all happier as Noble Savages.
w
w
No. Read Wakibaki. He broadcasts beliefs of the majority, including those who make decisions.
The rule of the game is simple: swap measurements, so people would judge products of your company VS products of competitors by the parameter you choose, no matter is it relative to the end result, or not. In 1930'th THD was right and good, because distortions grew up with power. Everyone accepts it already without understanding of it's meaning, so you may introduce enormous distortions on low power; it does not matter because on high power you beat'em all.
Last edited:
You mean the retrograde 1930's meter readers are still attempting to spread their propaganda? In a way, their continued persistence in the face of all facts is a paean to mass advertising over reality.
Even Philips was touting CD as a lower-fi alternative to LP until the Madmen of Sony got ahold of the format, and look at all the damage that resulted in. A generation of recording lost to high SQ, for a start.
Hey, meter readers - Sony made their bucks pushing 'Perfect Sound Forever' a generation ago and have moved on to SACD and Blu Ray. You should move on too, or at least stop rotating on your stupid schticks.
Even Philips was touting CD as a lower-fi alternative to LP until the Madmen of Sony got ahold of the format, and look at all the damage that resulted in. A generation of recording lost to high SQ, for a start.
Hey, meter readers - Sony made their bucks pushing 'Perfect Sound Forever' a generation ago and have moved on to SACD and Blu Ray. You should move on too, or at least stop rotating on your stupid schticks.
Last edited:
No, I mean they found the way to satisfy 1930'th requirements well introducing distortions that in 1930'th did not exist. "Multiple choice" way of thinking: instead of understanding of basics searching for "right answers".
MP3 haters !!
I agree , most MP3's should be banned. Downloading (and redownloading) a whole discography at X bitrate just to find that it was encoded in the wrong psychoacoustic algorithm or you find one at a higher bitrate is frustrating. With over 10TB of free space , I almost have my whole collection in flac and only with certain music and passages can one realize the difference between 320kbs Mp3's and lossless formats. "trash on a stick" (corndogs ??) yes , your typical purchased lame codec VBR mp3 is worthy of deletion 🙁 I have a whole library's classical collection encoded 320kbs with the original fraunhofer codec in true stereo and I have quite the time differentiating it from my redundant flac version. On the best amp/speakers/ (and soundcard), along with high volumes the flac's superiority becomes noticable. But .... I can't call the MP3 version "trash".
OS
I agree , most MP3's should be banned. Downloading (and redownloading) a whole discography at X bitrate just to find that it was encoded in the wrong psychoacoustic algorithm or you find one at a higher bitrate is frustrating. With over 10TB of free space , I almost have my whole collection in flac and only with certain music and passages can one realize the difference between 320kbs Mp3's and lossless formats. "trash on a stick" (corndogs ??) yes , your typical purchased lame codec VBR mp3 is worthy of deletion 🙁 I have a whole library's classical collection encoded 320kbs with the original fraunhofer codec in true stereo and I have quite the time differentiating it from my redundant flac version. On the best amp/speakers/ (and soundcard), along with high volumes the flac's superiority becomes noticable. But .... I can't call the MP3 version "trash".
OS
Well, I agree that 320 Kb/s can sound.....as good or better than some CDs. But 128Kb/s or lower - Puhleeze! And that's about where most MP3's seem to be encoded. And now we have HighDeafnation Radio at a whopping 64Kb/s a channel at best. Excuse me while I blow chunks.
Certainly I will listen to MP3's - I just don't feel it necessary to tell myself it's in the same audible category as 24 bit 96KS/s lossless or high quality analog.
Certainly I will listen to MP3's - I just don't feel it necessary to tell myself it's in the same audible category as 24 bit 96KS/s lossless or high quality analog.
Last edited:
- Status
- Not open for further replies.