Those of you who believe measurements aren't the whole story, do you have a hypothesis why that is?

[escksu]

I probably could have worded that better- my point was simply that a design can be a proper voltage source and the fact that it may not be able to double power from its 8 Ohm full power level to 4 Ohms does not affect that. Heck, there are tube amps that act as voltage sources too, but instead of doubling their power from 8 Ohms going down, whatever their 4 Ohm power is they cut that in half as the load impedance is doubled. Expensive way to do it IMO.

Oh ok, i get what you meant. Yes, you do not have to double the power and it will still work fine. Eg. 100w 8ohms but 150w 4ohms... 150w is still alot of power and its not like you will hit this limit unless you are playing at ear bursting volume.

 

[solderdude]

To me this comment seems really suspect! Otherwise why is there an ongoing effort to keep it down? This comment must be out of context.

Can I ask if you are Ralph Karsten ?

The remark wasn't out of context. But yes there is an ongoing effort to look for technical possible borders. The manufacturer creating one can boast with lowest distortion and get numbers with many 0's after the first decimal.

It's all moot though below 0.01% because when we listen at 100dB peaks in music we can't hear anything harmonically related 80dB below it.
The louder the signal gets the wider the masking is. At 80dB peaks (casual listening) we really can't hear 0dB SPL either.

Even the best OTL amplifier in the world is below 0.5% at 500W and as this is the best amplifier and 0.5% seems to be enough I think you would have to agree that < 0.01%, which is easy to make in SS, should be low enough not to impart any change in sound.

Soundstage measured its little brother to reach just below 0.1% in optimal conditions and having a (for OTL low) 10Ω output R.
The amps look nice and will (100% certainty) impart a clearly different sonic signature which is highly speaker dependent.
It's an effect box basically with nice looks. There will be quite a lot of people drawn to it.

I can also see why getting an audible lowest possible distortion with these designs is an enormous effort and difficult to reach.

 

[ahofer]

You can certainly make that argument (as the departed Mr March did in extremis and contra Magico) but if the amp/speaker combination works for you (or me) it's moot.

“Joined Dec 6, 2021”? How is it you have come to selectively absorb this old site history?

 

[egellings]

Why on earth so much power?

 

[atmasphere]

Can I ask if you are Ralph Karsten ?

The remark wasn't out of context. But yes there is an ongoing effort to look for technical possible borders. The manufacturer creating one can boast with lowest distortion and get numbers with many 0's after the first decimal.

It's all moot though below 0.01% because when we listen at 100dB peaks in music we can't hear anything harmonically related 80dB below it.
The louder the signal gets the wider the masking is. At 80dB peaks (casual listening) we really can't hear 0dB SPL either.

Even the best OTL amplifier in the world is below 0.5% at 500W and as this is the best amplifier and 0.5% seems to be enough I think you would have to agree that < 0.01%, which is easy to make in SS, should be low enough not to impart any change in sound.

Soundstage measured its little brother to reach just below 0.1% in optimal conditions and having a (for OTL low) 10Ω output R.
The amps look nice and will (100% certainty) impart a clearly different sonic signature which is highly speaker dependent.
It's an effect box basically with nice looks. There will be quite a lot of people drawn to it.

I can also see why getting an audible lowest possible distortion with these designs is an enormous effort and difficult to reach.

That's me.

We have been working on a class D project for about the last 5 years. It was about 6 years ago that I heard class D amps that convinced me the writing was on the wall and if an audio manufacturer didn't have a class D amp of their own in due time they would be left behind. My goal has always been to make amps that are neutral rather than amps that imbue a nice coloration. You are quite correct that OTLs have limited application since output impedance is always an issue except with the most powerful examples (which apparently we make). Some people do use them to get colorations as they do SETs but I don't count myself amongst them.

The amp we designed employs about 37dB of feedback at all frequencies in the audio band. Its distortion is considerably lower than our OTLs, but interestingly if both amps are on a speaker benign to the OTL, they are difficult to tell apart. This makes sense because most of the distortion of our OTLs is lower ordered; the amps are fully differential from input to output so the 3rd harmonic tends to be the primary distortion component, and the ear assigns the same tonality to the 3rd as it does the 2nd (which is to say it imparts a bit of warmth). There is enough of it that it is masking the higher orders. So the big difference you hear between the two tends to be in the bass because the class D is acting as a voltage source (its output impedance is a few milliOhms) while the OTL isn't.

I feel that you need to get well below 0.01% THD unless the distortion signature allows for the higher orders to be masked. If not an amp with that much THD can still sound bright; that is how sensitive the ear is to the higher orders (as I explained earlier). Somewhere around 0.001% I feel is really getting somewhere. You need to be -100db or more IMO.

IME the sensitivity the ear has to the higher orders is grossly downplayed by the industry and its incantations of 'negligible distortion' are often false. Like many things in life, its more complex than that. To me the missing link is understanding how the ear/brain system perceives sound; what's important and what is not (and so applying engineering to take advantage of that). A lot of that has only been found out in the last 30 years or so; IMO not enough of that understanding is applied when we do the measurements; distortion vs frequency is a good step in the right direction due to that pesky Gain Bandwidth Product thing.

 

[solderdude]

It was about 6 years ago that I heard class D amps that convinced me the writing was on the wall and if an audio manufacturer didn't have a class D amp of their own in due time they would be left behind.

I suppose your clientele would rather be interested in something with tubes than something that switches.

Completely own design modulators and output stages or just some module built into a case ?
Some tubes in front of it ?

You may want 100dB or more when you hit 110 to 120dB peaks. Any distortion then would be guaranteed to be below any audible thresholds.
On that note... this kind of performance will not be possible with tubes unless you use a lot of amplification and a lot of feedback.
A simple opamp can easily reach that performance.

 

[bigguyca]

Its pretty much the same thing. If your voltage remains the same and your resistance is 1/2, your current doubles. High current is what is hard to achieve. You need bigger transformers, more power transistors..... Whats even worse is that BJT current output drops as their temperature increase. Your cost goes up, multiple transistors in parallel has it own design challenges too. You need output resistors.

This is why not that many amps out there (esp. budget ones) are able to drive lower impedance speakers.

The current gain of BJT's increases, NOT decreases with temperature. BJT based amplifiers have compensation circuits that reduce the bias of the power transistors as their temperature rises. This increase in gain with increasing temperature is fundamental to the design of BJT power transistor based power amplifiers and is one of the design challenges with these amplifiers.

Without compensation circuits the output transistors will be subject to thermal runaway as temperature increases cause in increase in current, which in turn further increases temperature, which increase current until bad things happen. You can read about how all this works in any book on power amplifiers that covers amplifiers with BJT output transistors.

 

[atmasphere]

I suppose your clientele would rather be interested in something with tubes than something that switches.

Completely own design modulators and output stages or just some module built into a case ?
Some tubes in front of it ?

You may want 100dB or more when you hit 110 to 120dB peaks. Any distortion then would be guaranteed to be below any audible thresholds.
On that note... this kind of performance will not be possible with tubes unless you use a lot of amplification and a lot of feedback.
A simple opamp can easily reach that performance.

Its our own circuit and no tubes. The problem with using tubes is you can't really put that sort of feedback on them. Futterman did 60dB in his OTLs but due to gain bandwidth issues by the time you get to 10KHz the feedback is quite a bit lower. Despite that there were nevertheless stability issues as well. At least with a self oscillating class D amp stability is really a non-issue

 

[solderdude]

Without compensation circuits the output transistors will be subject to thermal runaway as temperature increases cause in increase in current, which in turn further increases temperature, which increase current until bad things happen. You can read about how all this works in any book on power amplifiers that covers amplifiers with BJT output transistors.

Thermal runaway in AB output stages is not because of increased current gain at higher temperatures but caused by a decreasing Vbe.
The trick is to bias the power stage with a 'zener' voltage that is thermally coupled and also decreases in the same amount as the Vbe of both output devices.

 

[solderdude]

Its our own circuit and no tubes. The problem with using tubes is you can't really put that sort of feedback on them. Futterman did 60dB in his OTLs but due to gain bandwidth issues by the time you get to 10KHz the feedback is quite a bit lower. Despite that there were nevertheless stability issues as well. At least with a self oscillating class D amp stability is really a non-issue

What's more fun (or challenging) designing tube amps or switching amps (especially the modulator) ?

 

[Momotaro]

“Joined Dec 6, 2021”? How is it you have come to selectively absorb this old site history?

December 7. Are you asking a question or begging the question?

 

[LTig]

Thermal runaway in AB output stages is not because of increased current gain at higher temperatures but caused by a decreasing Vbe.

Isn't this essentially the same as stating that current gain increases when Vbe does not change but temperature increases?

 

[solderdude]

Not really. When the Vbe would not change as drastically as it does and current gain only increases somewhat there would just be a somewhat higher current which would be compensated by the emitter resistors.
As the zener circuit is low impedance and a voltage source the relatively much higher change in Vbe is what makes the transistor conduct much more, heat up more, etc.
Emitter resistors are meant to counter that and this (as well as thermal coupling of the Zener) ensures this doesn't happen.

The early NAD amps did not have emitter resistors and solely relied on thermally coupled zener circuits.
Those amps blew up quite often.

 

[ahofer]

December 7. Are you asking a question or begging the question?

Not sure how this would be circular reasoning. It seems like you just joined but are referring to Alan March, who was escorted out many months ago, and was hardly best known here for this particular line of thinking. I’m just wondering if you have been lurking, or actually went over the archives carefully this month.

 

[Momotaro]

Not sure how this would be circular reasoning. It seems like you just joined but are referring to Alan March, who was escorted out many months ago, and was hardly best known here for this particular line of thinking. I’m just wondering if you have been lurking, or actually went over the archives carefully this month.

I didn't know about the "escorted out" part, just the recent absence and some commentary. I've followed ASR reviews in general as they come out, for a few years. I'm interested in Purifi amps for example, and March makes them locally so an obvious choice to buy. His comments on the Magicos were unusual/memorable, as they were vehement, but also because I have some 4-way passive speakers (although not Magico) that may or may not be a good match on that basis. Probably an amp thread or one discussing those speakers, class D and/or soundstage perceptions (I've read back through a few of those). Many of the explanations therein appear speculative (power supplies, phase, negative feedback levels and topologies, distortions "good" and "bad" etc) but I did have a disappointing amp purchase a while back that I'd rather not repeat.

Beautiful cat by the way, yours?

 

[DanielT]

What's more fun (or challenging) designing tube amps or switching amps (especially the modulator) ?

My guess, if relatively low budget is set, given the cost of outpot transformers: Tubes.

Now I can not construct amplifiers but one thing is completely obvious tubes are more fun (regardless of sound quality).

In the dark, you do not get this feeling with any Topping or Aiyima class D, see attached picture.

Then I like to sometimes get that distorted tube sound.

 

[ahofer]

I didn't know about the "escorted out" part, just the recent absence and some commentary. I've followed ASR reviews in general as they come out, for a few years. I'm interested in Purifi amps for example, and March makes them locally so an obvious choice to buy. His comments on the Magicos were unusual/memorable, as they were vehement, but also because I have some 4-way passive speakers (although not Magico) that may or may not be a good match on that basis. Probably an amp thread or one discussing those speakers, class D and/or soundstage perceptions (I've read back through a few of those). Many of the explanations therein appear speculative (power supplies, phase, negative feedback levels and topologies, distortions "good" and "bad" etc) but I did have a disappointing amp purchase a while back that I'd rather not repeat.

Beautiful cat by the way, yours?

The cat is ours, he is our third Cornish Rex (“Apollo”). Complete goofball, so the tongue-out photo is hardly just a weak moment.

I have a March Audio amp (p252) purchased in early 2019 and going strong. I had a good experience with Alan, but I guess he was pretty inflexible and outspoken on the forum. I might well buy again from him. I’d love to try his speakers, but shipping back to Australia from NYC seems daunting.

 

[escksu]

The current gain of BJT's increases, NOT decreases with temperature. BJT based amplifiers have compensation circuits that reduce the bias of the power transistors as their temperature rises. This increase in gain with increasing temperature is fundamental to the design of BJT power transistor based power amplifiers and is one of the design challenges with these amplifiers.

Without compensation circuits the output transistors will be subject to thermal runaway as temperature increases cause in increase in current, which in turn further increases temperature, which increase current until bad things happen. You can read about how all this works in any book on power amplifiers that covers amplifiers with BJT output transistors.

Sorry, its power, not current. My mistake. I meant to say that max power output of the transistor drops with temp.

The current gain of BJT's increases, NOT decreases with temperature. BJT based amplifiers have compensation circuits that reduce the bias of the power transistors as their temperature rises. This increase in gain with increasing temperature is fundamental to the design of BJT power transistor based power amplifiers and is one of the design challenges with these amplifiers.

Without compensation circuits the output transistors will be subject to thermal runaway as temperature increases cause in increase in current, which in turn further increases temperature, which increase current until bad things happen. You can read about how all this works in any book on power amplifiers that covers amplifiers with BJT output transistors.

Sorry, i was confused between current and power dissipation. I meant to say power dissipation, not current. These transistor can get rather hot (50-70c).

 

[Momotaro]

The cat is ours, he is our third Cornish Rex (“Apollo”). Complete goofball, so the tongue-out photo is hardly just a weak moment.

I have a March Audio amp (p252) purchased in early 2019 and going strong. I had a good experience with Alan, but I guess he was pretty inflexible and outspoken on the forum. I might well buy again from him. I’d love to try his speakers, but shipping back to Australia from NYC seems daunting.

I forgot he was also making a speaker. Explains a bit more why he has strong-ish opinions on speaker design. They do look good (I'd also like to hear them). As do his amps. I'm glad yours is going well.

 

[atmasphere]

What's more fun (or challenging) designing tube amps or switching amps (especially the modulator) ?

A tube amp for me is more fun (mostly because I can tamper with one without causing channel 4 of the TV to be blocked out ); class D amps are far more challenging IMO although the modulator is the easy part. We're working with GaNFETs and they are the difficult bit as they are so fast- you can have noise at crazy frequencies at which they are perfectly capable of switching. So you can run into heat problems due to this kind of noise, which might be 50MHz. Obviously this affects cerification issues; every time we make even a minor change we have to do an enourmous amount of testing to see to it that there isn't some sort of noise problem introduced with the change.