If you consider a gain of -1 is gain, then yes.
Phase Splitter - Electronics-Lab.
I shall also stop.
Cheers,
Patrick
Looks like Papa agrees with me too
"Q2 acts as a gain device and also a
signal splitter, driving both Q3 and Q4 simultaneously, but out of phase with each other. " Nelson Pass
http://www.firstwatt.com/pdf/art_plh.pdf
Similar
I did say similar and not the same. To clarify, as I gather many are getting a little hot under the collar:
a. Both amps have the same input arrangement
b. Both amps have the same feedback arrangement
c. Both amps power the input stage from the feedback point.
d. Both amps use a single polarity power supply.
e. Both amps cap couple the output.
f. Both amps use a minimum of devices to achieve their sound.
While the F8 output stage is SE and the JLH is PP, if you look at the JLH's output stage, unless the output devices are perfectly matched at increased drive the output stage will go SE. In this situation the bootstrap arrangement still controls the current of the upper device. While the F8 uses an opt coupler to control the current there is similarity or should I say "vague" similarity in having a current controlled upper device.
The point of the email was to say that maybe there is something in these similarities in delivering good sound, despite the phase splitter and the difference in the devices (bipolars vs JFETs). After all compared to your 99% typical Lin style amplifier both amps are very similar.
Oh and both amps were designed by of two of audio's greats JL Hood and Nelson Pass. I think that they were on to something here and it is worth examining the similarities.
Peace to my brothers in audio.
I did say similar and not the same. To clarify, as I gather many are getting a little hot under the collar:
a. Both amps have the same input arrangement
b. Both amps have the same feedback arrangement
c. Both amps power the input stage from the feedback point.
d. Both amps use a single polarity power supply.
e. Both amps cap couple the output.
f. Both amps use a minimum of devices to achieve their sound.
While the F8 output stage is SE and the JLH is PP, if you look at the JLH's output stage, unless the output devices are perfectly matched at increased drive the output stage will go SE. In this situation the bootstrap arrangement still controls the current of the upper device. While the F8 uses an opt coupler to control the current there is similarity or should I say "vague" similarity in having a current controlled upper device.
The point of the email was to say that maybe there is something in these similarities in delivering good sound, despite the phase splitter and the difference in the devices (bipolars vs JFETs). After all compared to your 99% typical Lin style amplifier both amps are very similar.
Oh and both amps were designed by of two of audio's greats JL Hood and Nelson Pass. I think that they were on to something here and it is worth examining the similarities.
Peace to my brothers in audio.
Unless the output stage of the F8 is biased at around 2.7A, it is likely to also have some level of push pull.
In fact the Firstwatt F8 website spec page states it is Push Pull.
So both amps are a little similar in that regard, they both can be adjusted for various amounts of push pull operation. Both of them also cannot operate into Class B
I guess there is also the possibility of some sliding bias effect depending on how Papa setup the output stage. I did something like this with lateral fets with zero degen in a mu follower arrangement, there was some sliding of the CCS due to the way I set it up.
Most of the similarities are trivial, the one major difference is significant.
Kind of like saying dogs and cats are similar, in some ways yes, in some ways no, some of the differences are significant.
Both amp designs are elegant and to be admired at least from the point of view of musical enjoyment.
I think I will leave it at that.
Either way, it's always exciting to see something new being released.
I thank Papa for sharing so much of his knowledge and experience with us.
It's accurate to say he taught me everything I know (well maybe not ohms law).
In fact the Firstwatt F8 website spec page states it is Push Pull.
So both amps are a little similar in that regard, they both can be adjusted for various amounts of push pull operation. Both of them also cannot operate into Class B
I guess there is also the possibility of some sliding bias effect depending on how Papa setup the output stage. I did something like this with lateral fets with zero degen in a mu follower arrangement, there was some sliding of the CCS due to the way I set it up.
Most of the similarities are trivial, the one major difference is significant.
Kind of like saying dogs and cats are similar, in some ways yes, in some ways no, some of the differences are significant.
Both amp designs are elegant and to be admired at least from the point of view of musical enjoyment.
I think I will leave it at that.
Either way, it's always exciting to see something new being released.
I thank Papa for sharing so much of his knowledge and experience with us.
It's accurate to say he taught me everything I know (well maybe not ohms law).
I had a check today out of curiosity.
If you wish to keep the same bias and dynamic current ratio between the MOSFETs, then some changes in resistor values are necessary.
The reward is 2x lower distortion.
Does not necessarily mean better sound, only more NFB, as expected.
It also clips earlier (at 4V input instead of 4.3V), also as expected.
And in real life, it will drift more with temperature due to the positive tempco of the IRFP.
But you can solve that with a DC servo and plenty of heat sink.
Patrick
.
Hello 2 picoDumbs and Ejam,
The highlighted sentence above is due to the use a mu-follower by Mr. Pass;which he taught in US 5,170,522. Patent is attached.
There is a new article about "Leaving Class A".
Where does an amp like M2X output stage fit into this article?
It idles at about 1.4A but has a special bias arrangement so it raises bias when more power is needed?
Does it fit into the 2.56A bias example regarding distortion and contents of harmonics?
Would you call M2X a "Pure Class A" or a "Dynamic Class A" amp?
Where does an amp like M2X output stage fit into this article?
It idles at about 1.4A but has a special bias arrangement so it raises bias when more power is needed?
Does it fit into the 2.56A bias example regarding distortion and contents of harmonics?
Would you call M2X a "Pure Class A" or a "Dynamic Class A" amp?
With "Dynamic" I was thinking of one the Pass patents to control bias and as I have understood the patents it causes that an amp can deliver more Class A watts than the "idle bias" indicates?
M2 can deliver about twice the idle bias current into the load without any of the output devices cuts off.....or is that not true?
You need more than 1.4A into 8 ohm to get 25 Class A watts or 40W in 4 ohm?
Or how do we explain those Class A watts with only 1.4A bias?
M2 can deliver about twice the idle bias current into the load without any of the output devices cuts off.....or is that not true?
You need more than 1.4A into 8 ohm to get 25 Class A watts or 40W in 4 ohm?
Or how do we explain those Class A watts with only 1.4A bias?
Can it do that without any of the output devices cuts off?
If I test in a load it is difficult to know if both output devices are conducting current which is a requirement for Class A?
What about 40 Class A watts in 4 ohm.....you need more than twice the bias current?
M2 is specified as 40 Class A watts into 4 ohm?
If I test in a load it is difficult to know if both output devices are conducting current which is a requirement for Class A?
What about 40 Class A watts in 4 ohm.....you need more than twice the bias current?
M2 is specified as 40 Class A watts into 4 ohm?
I may have one question more.
When Class A watts are specified it is instant max. power where output devices are still conducting?
I can't expect like say 40W RMS class A into 4 ohm if I use a sine generator for testing?
If I measure voltage over source resistors during a test I can see when voltage drops to zero if I want to know when an output device cuts off?
The 2x bias current for a PP output stage is a "rule of thump"?
I guess it can be a bit different which kind of output devices are used (square law vs exp. law devices)?
When Class A watts are specified it is instant max. power where output devices are still conducting?
I can't expect like say 40W RMS class A into 4 ohm if I use a sine generator for testing?
If I measure voltage over source resistors during a test I can see when voltage drops to zero if I want to know when an output device cuts off?
The 2x bias current for a PP output stage is a "rule of thump"?
I guess it can be a bit different which kind of output devices are used (square law vs exp. law devices)?
read about M2 and about XA25, find the difference ( ignore different parts in output) and you'll get your answer
tip - source resistors
of smaller amps circulating around, only Babelfish M25 is able to (almost) double power @4R, comparing to 8R, deliberately made following XA25 footsteps
no FW amp is made to have sliding bias - no need for that , while staying in chosen form factor - these tricks were reserved for higher powered amps on market, and to still have them made in reasonably sized package, which resorting to full Class A would certainly made impossible
some makers were frank, some were utter liars , speaking of sliding bias in their gadgets ...... and some of them coined their own names and rulez, even then selling more ordinary things under novelty badge
tip - source resistors
of smaller amps circulating around, only Babelfish M25 is able to (almost) double power @4R, comparing to 8R, deliberately made following XA25 footsteps
no FW amp is made to have sliding bias - no need for that , while staying in chosen form factor - these tricks were reserved for higher powered amps on market, and to still have them made in reasonably sized package, which resorting to full Class A would certainly made impossible
some makers were frank, some were utter liars , speaking of sliding bias in their gadgets ...... and some of them coined their own names and rulez, even then selling more ordinary things under novelty badge
You might enjoy the square law article newly posted at FIRST WATT
which discusses Class A envelope in the context of degeneration of
push pull stages.
Undegenerated outputs can remain in conduction for more than twice
the idle current, making the definition of Class A a bit more complicated.
I will read the article in more details.
I had difficulties finding articles which explain M2 and XA25 designs in details.
I remember I read somewhere that the LM385 voltage ref. in M2 was to be able to deliver more power in low impedances. They prevents to much "degeneration"?
What I am a bit confused about reading the M2 manual is that distortion is more or less pure 3rd harmonic while my own M2X measurements and also Zen Mod's M25 are slightly 2nd harmonic dominant. Is that because original M2 uses matched output devices (even that they are source followers)?
Also Whammy which seems to have same kind of PP output stage topology is 2nd harmonic dominant. I assume that this comes from a slight "unbalance" in PP output stage?
I had difficulties finding articles which explain M2 and XA25 designs in details.
I remember I read somewhere that the LM385 voltage ref. in M2 was to be able to deliver more power in low impedances. They prevents to much "degeneration"?
What I am a bit confused about reading the M2 manual is that distortion is more or less pure 3rd harmonic while my own M2X measurements and also Zen Mod's M25 are slightly 2nd harmonic dominant. Is that because original M2 uses matched output devices (even that they are source followers)?
Also Whammy which seems to have same kind of PP output stage topology is 2nd harmonic dominant. I assume that this comes from a slight "unbalance" in PP output stage?
DC Coupled Headphone Version
I feel like some more simulations today.
How about a headphone version, DC coupled ?
In such application, the 2SK1058 behaves very similar to the Semisouth.
The transconductance of the lateral is ~1/6 of the Semisouth.
But Bias for headphone is also about 1/6, and the load impedance about 6x.
Vgs are both ~1V at bias.
More importantly, both have negative tempco at bias.
So no fear of thermal run-away without source degenration.
How does it sim ?
Actually VERY similar in performance.
You should pick ones with a high-ish Vgs (say >1V at 250mA) if you want low distortion.
They can take a lot of power, so 24V 300mA is no issue at all.
And experience suggests that this will have control / authority in abundance.
Patrick
.
I feel like some more simulations today.
How about a headphone version, DC coupled ?
In such application, the 2SK1058 behaves very similar to the Semisouth.
The transconductance of the lateral is ~1/6 of the Semisouth.
But Bias for headphone is also about 1/6, and the load impedance about 6x.
Vgs are both ~1V at bias.
More importantly, both have negative tempco at bias.
So no fear of thermal run-away without source degenration.
How does it sim ?
Actually VERY similar in performance.
You should pick ones with a high-ish Vgs (say >1V at 250mA) if you want low distortion.
They can take a lot of power, so 24V 300mA is no issue at all.
And experience suggests that this will have control / authority in abundance.
Patrick
.