Mastersound PF100 PSET Amplifier

[Republicoftexas69]

How are the Class D GaN amps vs the Class A . I find most class D detailed sounding but with odd Timbre


Regards

Still trying to figure out what "odd" is. So I will say that I do prefer my AH X200 but this is a class A PP tube. So I am not sure I can really answer your question.

Cheers.

 

[Audiohertz2]

Still trying to figure out what "odd" is. So I will say that I do prefer my AH X200 but this is a class A PP tube. So I am not sure I can really answer your question.

Cheers.

Ok , lets try poor instead of odd ..!

 

[Republicoftexas69]

Ok , lets try poor instead of odd ..!

I actually find the GaN amps do sound very tube like and very detailed/articulate, still a tad sterile but not like Ice or Purfi amplifiers. Using a tube pre-amp will take a bit of the odd sound out,

 

[bazelio]

What is an unacceptable distortion level? There is not a simple answer to this question as it depends heavily on the amp topology and also the SPL at which one is listening. So, I don’t see how you can make the statements you make regarding usable power.
For example, if an amp has a high order distortion without significant low order masking (orders above 5th probably aren’t masked at all) at low power, one could argue from a psychoacoustic perspective it has 0% usable power.
My amp generates 35 SET watts at 1% THD…most of which is 2nd and 3rd harmonic…it is rated at 42 watts. I would argue that is a lot more than 25% of the rated power as usable. Part of the usable power thing is semantics. Most 8 watt 300B amps are really 4-6 watt amps and should be regarded accordingly.
If an amp is predominantly low order distortion up to close to clipping, what % THD is considered unacceptable? 1%? 5%?
Many amps with low measured distortion still don’t sound good…then what? Geddes demonstrated that there is NO correlation between THD or IMD and sound quality. The pattern matters much more.

Stereophile used to (and maybe still does) consider clipping to be 1% THD, which some might say was never "fair" to SET amps because without exceeding 1% you can't get anywhere near the max power output spec of any SET tubes without feedbavk. Max power output is 5% THD in many or most DHT datasheets for example. I've seen some pentodes specify max power at 10% THD. Now Ralph keeps saying most SET amps don't use feedback. I don't have any idea when that survey was taken, but I do know several that do. What is the topology of your SET amp that generates 35W at 1% THD. Which tubes does it use?

 

[bazelio]

Seriously, trying to get that 'SET magic' from an amp of this size, and actually needing the power is IMO a Bad Idea. What you might not be keeping in mind is that any SET sans feedback (which is almost all of them) really only has about 20-25% (the latter on a good day) usable power.

20-25% seems generous to me, honestly. I've been playing with various DHTs recently and have found that to be closer to 10-15% usable power. My method has simply been to flatten out the load line until THD comes down to 1%. This usually means the load reaches 5-10x the nominal datasheet value, at which point, there's hardly any output power left on tap. 25% would actually be phenomenal.

 

[morricab]

Stereophile used to (and maybe still does) consider clipping to be 1% THD, which some might say was never "fair" to SET amps because without exceeding 1% you can't get anywhere near the max power output spec of any SET tubes without feedbavk. Max power output is 5% THD in many or most DHT datasheets for example. I've seen some pentodes specify max power at 10% THD. Now Ralph keeps saying most SET amps don't use feedback. I don't have any idea when that survey was taken, but I do know several that do. What is the topology of your SET amp that generates 35W at 1% THD. Which tubes does it use?

Ec86 input, 6N6P driver and GM70 output.

 

[morricab]

20-25% seems generous to me, honestly. I've been playing with various DHTs recently and have found that to be closer to 10-15% usable power. My method has simply been to flatten out the load line until THD comes down to 1%. This usually means the load reaches 5-10x the nominal datasheet value, at which point, there's hardly any output power left on tap. 25% would actually be phenomenal.

Your assumption is that above 1% is too much distortion of the type SETs produce…this despite the fact the studies have shown that 2% of 2nd harmonic is inaudible on pure sine waves let alone a complex music signal. Also, let’s not be fooled by what manufacturers claimed power is…they all are generous with the ratings…speaker designers are the worst at this padding sensitivity numbers. They do this because they think it will help sales I guess.
Keep in mind also that most of the time these amps are paired with pretty sensitive speakers and power well below 1 watt is needed for most listening. For short loud peaks the amp will produce the needed power with acceptable distortion.
This brings up the topic of dynamic vs. static power. Peter Van Willenswaard found that tube amps (SET in particular) could generate huge peak power for short bursts…way more than SS amps. Even though it looked quite distorted on a scope, the listeners did not hear an issue…psychoacoustic factors again play a huge role in this.

This 10 or 25% usable power argument is based on false premises of where the power becomes unusable. I would argue that a good SET will be able generate > 100% of its rated power as usable for short durations, as demonstrated by PVW, thus leading to the common perception that they sound powerful for their rating.

 

[morricab]

20-25% seems generous to me, honestly. I've been playing with various DHTs recently and have found that to be closer to 10-15% usable power. My method has simply been to flatten out the load line until THD comes down to 1%. This usually means the load reaches 5-10x the nominal datasheet value, at which point, there's hardly any output power left on tap. 25% would actually be phenomenal.

Tubes Do Something Special

Most people who now listen to tube amplifiers began with a transistor amp, and know from experience that a tube amp of a given measured power output sounds louder than its nominally identical transistorized equivalent. The unofficial consensus is that you need two to four times the transistor...

www.stereophile.com

 

[bazelio]

Ec86 input, 6N6P driver and GM70 output.

Topology? A single ended GM70 is rated for 30W @ 5% THD in Class A1.

 

[morricab]

Topology? A single ended GM70 is rated for 30W @ 5% THD in Class A1.

No this is not correct. A GM70 will make 43 watts with 1200V on the anode. I think my amp is 3% THD at 42 watts and 1% at around 35 watts

 

[bazelio]

Your assumption is that above 1% is too much distortion of the type SETs produce…this despite the fact the studies have shown that 2% of 2nd harmonic is inaudible on pure sine waves let alone a complex music signal. Also, let’s not be fooled by what manufacturers claimed power is…they all are generous with the ratings…speaker designers are the worst at this padding sensitivity numbers. They do this because they think it will help sales I guess.
Keep in mind also that most of the time these amps are paired with pretty sensitive speakers and power well below 1 watt is needed for most listening. For short loud peaks the amp will produce the needed power with acceptable distortion.
This brings up the topic of dynamic vs. static power. Peter Van Willenswaard found that tube amps (SET in particular) could generate huge peak power for short bursts…way more than SS amps. Even though it looked quite distorted on a scope, the listeners did not hear an issue…psychoacoustic factors again play a huge role in this.

This 10 or 25% usable power argument is based on false premises of where the power becomes unusable. I would argue that a good SET will be able generate > 100% of its rated power as usable for short durations, as demonstrated by PVW, thus leading to the common perception that they sound powerful for their rating.

Oh Lord, listening to sine waves? No thank you. Music is what matters and is where most people would be likely to hear it. Please publish these studies. The methodology is important. It's difficult to "hear distortion" without a baseline. Anyone might listen to 2% 2nd harmonic in isolation and be unsure if they can hear it. But I contend that they're hard of hearing if they can't hear a difference between 1% and 2%. Now, as to which sounds subjectively better will vary person to person. Stereophile considers clipping to br 1% THD as I stated. This is because 1% is clearly visible. And from a purely objective standpoint, that's a reasonable stake in the ground. You're correct that most amps aren't pushed to maximum power continuously which is why Ralph explained the common symptom of distortion as "exciting dynamics". The heavy load steps of large transients become very distorted.

 

[bazelio]

No this is not correct. A GM70 will make 43 watts with 1200V on the anode. I think my amp is 3% THD at 42 watts and 1% at around 35 watts

Woops, I was thinking of the 845. I just looked it up. My bad. I can't actually find the power output spec for GM70 in any datasheet. But, it does look like a preferable tube. Do you know the topology of your amp? Class A1 single ended, zero feedback? Or what?

 

[morricab]

Oh Lord, listening to sine waves? No thank you. Music is what matters and is where most people would be likely to hear it. Please publish these studies. The methodology is important. It's difficult to "hear distortion" without a baseline. Anyone might listen to 2% 2nd harmonic in isolation and be unsure if they can hear it. But I contend that they're hard of hearing if they can't hear a difference between 1% and 2%. Now, as to which sounds subjectively better will vary person to person. Stereophile considers clipping to br 1% THD as I stated. This is because 1% is clearly visible. And from a purely objective standpoint, that's a reasonable stake in the ground. You're correct that most amps aren't pushed to maximum power continuously which is why Ralph explained the common symptom of distortion as "exciting dynamics". The heavy load steps of large transients become very distorted.

Please explain how this so called “exciting dynamics “ can be created and why specifically from SET amps. It is well known that tubes tend to compress before they actually clip but how does that make dynamics more exciting rather than less exciting? Subjectively 1% means nothing.

 

[morricab]

Woops, I was thinking of the 845. I just looked it up. My bad. I can't actually find the power output spec for GM70 in any datasheet. But, it does look like a preferable tube. Do you know the topology of your amp? Class A1 single ended, zero feedback? Or what?

Class A1 no global feedback... Perhaps some local but not sure.
Info about GM70

GM70 tube amplifier.

www.jacmusic.com

 

[bazelio]

Please explain how this so called “exciting dynamics “ can be created and why specifically from SET amps. It is well known that tubes tend to compress before they actually clip but how does that make dynamics more exciting rather than less exciting? Subjectively 1% means nothing.

Ralph already explained it. The bottom line is that you want speakers as sensitive as possible so you have as much dynamic headroom as possible in order to (try to) stay out of the highest distortion region during large transient peaks. I'll bet that a great many systems are incapable of avoiding 5% distortion peaks because so many people build their systems at the margin. And what does or does not "mean nothing" subjectively is up to the subject lol. There are many audiophiles who find lower distortion to sound more pleasing than higher distortion.

 

[bazelio]

Class A1 no global feedback... Perhaps some local but not sure.
Info about GM70

GM70 tube amplifier.

www.jacmusic.com

Chances are high based just on the numbers you've posted that your amp uses feedback.

 

[morricab]

Ralph already explained it. The bottom line is that you want speakers as sensitive as possible so you have as much dynamic headroom as possible in order to (try to) stay out of the highest distortion region during large transient peaks. I'll bet that a great many systems are incapable of avoiding 5% distortion peaks because so many people build their systems at the margin. And what does or does not "mean nothing" subjectively is up to the subject lol. There are many audiophiles who find lower distortion to sound more pleasing than higher distortion.

Read the papers from Earl Geddes. There is no correlation between THD or IMD and subjective sound quality...at least over the wide range they evaluated. This is what I mean by "means nothing". It's just a number... not a threshold for sonic quality. It's also why your statement about audiophile preference is questionable. If nearly all audiophiles really preferred ultra low distortion electronics, not only would SETs NOT have made a comeback, Ralph would not have been able to sell OTLs for over 30 years. Ralph's "explanation" is simplistic and almost certainly wrong. SS amps often have much higher distortion of high order harmonics with increasing frequency...this doesn't make them sound more dynamic... Usually they just sound annoying.

 

[morricab]

If Stereophile wanted to give a useful guide to amps that probably sound better they would use their technical data and then plug it into a psychoacoustic metric like the one developed by Geddes or by Cheever. The better that metric the more likely it is to sound good…there would still be room for preferences at the top but it would help weed out ones that do poorly psychoacoustically.

I actually did this once using the much simpler metric from DEL Shorter and data from Stereophile and SoundStage. Based on that metric the Wyetech Tooaz 211 SET had the best outcome of all the amps data I processed. That amp was highly praised in the early 2000s. Lame ML2 did well too.

 

[microstrip]

Read the papers from Earl Geddes. There is no correlation between THD or IMD and subjective sound quality...at least over the wide range they evaluated. This is what I mean by "means nothing". It's just a number... not a threshold for sonic quality. It's also why your statement about audiophile preference is questionable. If nearly all audiophiles really preferred ultra low distortion electronics, not only would SETs NOT have made a comeback, Ralph would not have been able to sell OTLs for over 30 years. Ralph's "explanation" is simplistic and almost certainly wrong. SS amps often have much higher distortion of high order harmonics with increasing frequency...this doesn't make them sound more dynamic... Usually they just sound annoying.

Earl Geddes work in the early 2000's on distortions of electronics was just the start of a promising work that was never tested or proceeded. Very interesting read, but unfortunately nothing more than that.

Although there are no established firm rules of correlation between THD or IMD and subjective sound quality, we can't pretend that an amplifier having more than 10% of distortion at a few watts is accurate when used with medium efficiency speakers. IMO consumers should be delivered measurements of what they get - in this aspect Lamm is exemplary. Fortunately magazines such as Stereophile, Audio Critic, HiFi New's or the german Stereo still publish them in their reviews.

 

[microstrip]

If Stereophile wanted to give a useful guide to amps that probably sound better they would use their technical data and then plug it into a psychoacoustic metric like the one developed by Geddes or by Cheever. The better that metric the more likely it is to sound good…there would still be room for preferences at the top but it would help weed out ones that do poorly psychoacoustically.

I actually did this once using the much simpler metric from DEL Shorter and data from Stereophile and SoundStage. Based on that metric the Wyetech Tooaz 211 SET had the best outcome of all the amps data I processed. That amp was highly praised in the early 2000s. Lame ML2 did well too.

Such metric still does not exist and I doubt that it can be established for stereo purposes. And no way we should put in the same bag the careful work of Earl Gueddes and the Cheever amateurish MSc thesis.