Capacitance multiplier: which Darlington to choose?

[thor.zmt]

The customers who rewarded advertisers who hyped performance specifications by buying their SKUs deserve the blame/credit for how the industry evolved.

Customers cannot know better, usually. They lack the knowledge. And they like to buy, these days on "scientific evidence".

So, the system becomes self reinforcing.

Step 1 - Marketeer one decides disingenuously to promote some specific spec as figure of merit, to sell more macguffins to the unsuspecting pee-pull.

Step 2 - it works, more items are sold.

Step 3 - now everyone wants to make macguffins even better in this figure.

Step 4 - test gear makers fall over themselves to provide new gear or add ONS / upgrades to existing gear to quantify this figure of merit.

Step 5 - with standardized tests and engineering effort focused, more and more macguffins get better and better in this metric, until practical limits are reached. Nobody asses sound quality impairment effects. Gear sounds worse but measures better. Saturation is reached.

Step 6 - GoTo Step 1

This is what happens when advertising of figures of merit is done without forcing an evidence based approach that requires the Marketeers to prove that improving metric x reliably improves sound quality without any penalties and sound quality impairment in other areas.

It is easy to promote "low distortion" specifically as low THD as a good thing, because nobody wants a more distorted hifi system.

But the customer do not know and without huge highly specialisef research cannot know that THD has no proven or reliable link with reduced sound quality impairment.

The blame rests squarely with first the huckster in step one and secondly with an industry that went along with the BS instead of calling it out and educating customers as to the nature AND offering proven reliable alternatives.

But it is always easier to copy than to be original.

 Thor

 

[thor.zmt]

It's all about learning and improving on the best. That's asian philosophy. If you know the best, only then your competent to improve on it. If one doesn't know what others have achieved and how they managed to do so, then improving is much more difficult.

I tend to go with Charles Mordechai:

"Practice without Theory is blind, Theory without practice is sterile."

The Asian approach is repetition send copying without understanding is practice without theory.

Once you understand how, in theory, a transistor works, including behaviour under unintended condition (avalance/first breakdown, secondary breakdown etc.) you can look at a circuit and understand why for example a low voltage transistor might work in a high voltage supply etc. and you are able to extend the application out of its original context.

That will not tell you the results in sound quality, but the electrical operation. And you do not need to.ask "can I use part X", but you will understand what using part X implies.

And you would for example understand that correctly implemented protection circuits do not alter the operation of the circuit under normal conditions or, where they do, just how much.

For example, a TIP122 could be used reliably in a high voltage circuit if you connect a sufficiently high power zenner diode across the transistor to limit the voltage across the transistor to values that do not cause breakdown.

This trick also works for LM317 regulating a 300V PSU.

Thor

 

[Walter66]

For example, a TIP122 could be used reliably in a high voltage circuit if you connect a sufficiently high power zenner diode across the transistor to limit the voltage across the transistor to values that do not cause breakdown.

This trick also works for LM317 regulating a 300V PSU.

Thor

Dear Thor,
not playing in champions league in this game, more of a simple amateur who has to do small steps to learn.
Has this trick you mentioned be applied in this Cortese PSU with the LM 317 included, too? It regulates a 300V output. Don't really understand how it does work with high voltage.

 

[metalockpick]

I don't know what your problem is with that. Just after WW2 negative feedback was barely 10 years old. Lots of companies had explored it in detail as part of their war work and the literature is littered with articles about its application to sound reproduction in order to lower distortion. All Leak did was apply that experience. He didn't make a thing about low distortion it was already a thing.


Funnily enough so did every other tube power amp using negative feedback where the low frequency open loop bandwidth was limited. Not many manufacturers were brave enough to quote it though


Not it isn't. It has a global feedback loop and local feedback too.

Applying that degree of negative feedback to a tube amplifier in the mid 1940s was no mean feat given the components they had available. And once again I agree Leak was an excellent marketer just like all the other hi-fi brigade. Don't forget vinyl records with their extended frequency response only became commonplace after the war.

Leak Point One Amplifiers

THD was already an established figure of merit when Leak released his point one range. He didn't invent it. There was a huge sea change in fidelity as a whole post WW2 and Leak played a part in it but by no means did he have the degree of influence you seem to imagine.

Cheers

Ian

My problem is using excessive or inappropriate feedback in order to game the "test equipment" of the day. Your argument here is akin to defending bloodletting and the balancing of humors as effective medicine and somehow still relevant. Feedback after a transformer can be used to affect gain and frequency response, but it can not fix time domain issues. All those count towards "distortion"; the time domain issues are far more egregious than the frequency domain issues.
Current FFT test equipment does a good job (when used properly) and can correlate to what can be heard. That's not the case with measurements prior to FFT

 

[thor.zmt]

Has this trick you mentioned be applied in this Cortese PSU with the LM 317 included, too? It regulates a 300V output.

View attachment 109740

Not really.

There an MJE340 is used as a simple "regulator" that regulates the voltage across the LM317 to.~ 10V.

It's a very generic circuit documented for ages. I do not consider this a very good sounding solution. It's ok if followed by RC filtering or in a PP output stage.

BTW, does the schematic miss the capacitor after the bridge rectifier?

Thor

 

[JohnRoberts]

Customers cannot know better, usually. They lack the knowledge. And they like to buy, these days on "scientific evidence".

So, the system becomes self reinforcing.

Step 1 - Marketeer one decides disingenuously to promote some specific spec as figure of merit, to sell more macguffins to the unsuspecting pee-pull.

Step 2 - it works, more items are sold.

Step 3 - now everyone wants to make macguffins even better in this figure.

Step 4 - test gear makers fall over themselves to provide new gear or add ONS / upgrades to existing gear to quantify this figure of merit.

Step 5 - with standardized tests and engineering effort focused, more and more macguffins get better and better in this metric, until practical limits are reached. Nobody asses sound quality impairment effects. Gear sounds worse but measures better. Saturation is reached.

Step 6 - GoTo Step 1

This is what happens when advertising of figures of merit is done without forcing an evidence based approach that requires the Marketeers to prove that improving metric x reliably improves sound quality without any penalties and sound quality impairment in other areas.

It is easy to promote "low distortion" specifically as low THD as a good thing, because nobody wants a more distorted hifi system.

But the customer do not know and without huge highly specialisef research cannot know that THD has no proven or reliable link with reduced sound quality impairment.

The blame rests squarely with first the huckster in step one and secondly with an industry that went along with the BS instead of calling it out and educating customers as to the nature AND offering proven reliable alternatives.

But it is always easier to copy than to be original.

 Thor

It is human nature to look elsewhere to assign blame... Back in the 80s I had to deal with one competitor in the mixer business who spent more on advertising one mixer SKU than my boss did for over a thousand different SKUs. Of course it worked in the short term and I lost mixer market share.

I get angry when I see too many (any?) ads....

JR

 

[ruffrecords]

My problem is using excessive or inappropriate feedback in order to game the "test equipment" of the day.

I have a problem with that too.

Your argument here is akin to defending bloodletting and the balancing of humors as effective medicine and somehow still relevant.

What argument? Nobody is arguing the merits or otherwise of NFB. We were arguing about historical fact and its consequences.

Cheers

Ian

 

[Walter66]

Not really.

There an MJE340 is used as a simple "regulator" that regulates the voltage across the LM317 to.~ 10V.

It's a very generic circuit documented for ages. I do not consider this a very good sounding solution. It's ok if followed by RC filtering or in a PP output stage.

BTW, does the schematic miss the capacitor after the bridge rectifier?

Thor

Well, I'm not sure about this. The circuit comes from an obscure website, they may have changed parts and did experiment to make it work. In general, nobody knows what IC types were used, as all designations have been deleted. No original schemos available.
Cortese is a very natural and excellent sounding amp which had some great reviews. A friend of mine purchased two of them, one with 300B, the other with F2A output. McIntosh is dead for him since he discovered them.

 

[Walter66]

I have a problem with that too.

What argument? Nobody is arguing the merits or otherwise of NFB. We were arguing about historical fact and its consequences.

Cheers

Ian

In germany, we had a change in the general audio discussion in the mid 1980's. Until this time, the three major mags published measured data to all tested gear from their own labs, not only replicating the measurements of the manufacturers. They gave hierarchical ratings to it, strongly bounded to the excellence of those measurements. The best gear had the best measurements and the best sound.

That was all changed when mags were published without any measurements. The so called author and editor focused subjective listening test was the only criteria for the hierarchical classification of tested gear, because the experts knew many decades ago, that there was no connection with measurement of THD and good or bad sound.

This is a well known common opinion in audio forums today, its the majority which puts subjective listening tests more important than measurements.
We had to learn that 0.000001% THD not make for the most musical or best sounding amp in general.

I'm not a hater of NFB and use it all the time in my designs. If applied correctly and with the best dosage, it could help to achieve better sonic and measurement qualities, although I don't measure gear but just listening to it.

 

[thor.zmt]

I get angry when I see too many (any?) ads....

Don't get angry, get even. The problem is however, business people and marketing people know less about the subject matter they work with than the customers and they assume by definitions that engineers cannot know about business or marketing, because if they were any good they's be business or marketing MBA - right?

So in the end we get my little self reinforcing homeostatic system, which even forces engineers that know better to get with the program.

As said, I think there is a better way, but that is a difficult path.

Thor

 

[thor.zmt]

Well, I'm not sure about this. The circuit comes from an obscure website, they may have changed parts and did experiment to make it work. In general, nobody knows what IC types were used, as all designations have been deleted. No original schemos available.

I see. I used to get the photo's from the glossy german mag's and reverse engineer Shindo's stuff from that. Gave it up after a while as there was nothing more much interesting.

Cortese is a very natural and excellent sounding amp which had some great reviews. A friend of mine purchased two of them, one with 300B, the other with F2A output. McIntosh is dead for him since he discovered them.

Yes, absolutely.

In the early 00's I rebuild the Shanling SP-80 Amplifiers (and their CD-Players etc) to more "interesting" circuitry for the UK Importer. These were really custom commissions.



All the solid state stuff got niXed.

Output Stage switchable Triode/UL/Pentode, usually KT-100 fitted, but I made a few examples with 300B outputs as well.

Driver stage differential 6N12 (think of it as an Octal 5687WB from China - Nangking Tube Factory, usually excellent tubes, real sleepers) with CCS to Bias Rail, input ground referenced, loads actually with positive feedback from output anodes (literally resistor from opposite polarity Output Anode).

Input Stage stayed 6SL7, but was made differential on CCS from Bias supply, XLR input, local feedback from degeneration.

Shunt feedback from output anodes to 6SL7 anodes, DC coupled. This in effect operates the input stage as Voltage controlled current source and the output stage as current to voltage converter. The 6SL7 anodes "look" in effect into a near short, so miller effect is defeated.

Small trimmable capacitors made from teflon coax (simply push/pull the core in/out to adjust) to trim square wave response.

Silver Mica and Audio Note Silver Foil coupling capacitors.

IRF830 Capacitor Multipler for output stage and second one for frontend.

These were technically quite excellent and sounded great.

I believe the Cortese has some parallels with this.

I designed this however more based on the HK Citation 2 and the V69, I was of course aware of what Shindo did around that time (is that really two decades ago?).

We asked Shanling to put my design into production as special edition, but they were not interested.

Commonly these were sold with an open baffle speaker of my own design, ideally fitted with Supravox 8" Fieldcoil full range drivers, plus REL Subwoofers and Visaton TL-16H tweeters....



This shows the "cheap" variant with the Supravox Bicone permanent magnet speaker and a Linn Sub).

For several years (until I left England) I use a pair of the bastardised Shanling Amp's with 300B's and the Field coil Supravox open baffles as my main system (with additional subs and supertweeters of course.

Thor

 

[thor.zmt]

I'm not a hater of NFB and use it all the time in my designs.

NFB is a tool. All design techniques are tools to achieve a certain goal set.

And the goal set determines how we verify that goals have been achieved.

I don't measure gear but just listening to it.

I do both. Again, back to Charles Mordechai: "Practice without Theory is blind, Theory without practice is sterile."

Listening tells me if I got "good sound". Measuring tells me if it works as designed. Both are necessary.

Thor

 

[Walter66]

NFB is a tool. All design techniques are tools to achieve a certain goal set.

And the goal set determines how we verify that goals have been achieved.



I do both. Again, back to Charles Mordechai: "Practice without Theory is blind, Theory without practice is sterile."

Listening tells me if I got "good sound". Measuring tells me if it works as designed. Both are necessary.

Thor

"Gray, dear friend, is all theory,
And green the golden tree of life." Johann Wolfgang von Goethe

Sakuma San engineered his famous amps by listening to the sound of his circuits.
With most electronics, you hear the effort and not the art.

In the end, no technical databook will tell if a choosen workpoint is perfect for music reproduction. I've just again experienced the truth of this statement. In the end, when it's a good electric amp design, it can tell WHY the musicians play together, whats their intention on the notes being played. This is much more than sound, at least to me.

 

[thor.zmt]

Sakuma San engineered his famous amps by listening to the sound of his circuits.

And he made multiple systems, all mono, with special amplifiers and speakers for specific music. I visited his restaurant when I apprenticed for a few weeks with Kondo San.

I more prefer a Tokyo Jazz Kissa System, small room, huge vintage JBL speakers, vinyl, turntables and a DJ Booth and usually well restored vintage Tube Gear in backlit glass vitrines, Vintage Scotch Whiskey and lady Ella singing her heart out.

In the end, no technical databook will tell if a choosen workpoint is perfect for music reproduction.

No, for that we listen. And adjust. And repeat.

But it is possible to have unnecessary and substantial problems in a device or system, which show up under some conditions and are better resolved and easily identified by objective tests.

In the end, when it's a good electric amp design, it can tell WHY the musicians play together, whats their intention on the notes being played. This is much more than sound, at least to me.

Who quipped "I don't want to hear what the musicians do on stage, I want to hear why they are there!"?

For AMR I coined "closest approach to the original emotion".

But we still need something that also works as expected, trust me my friend.

Thor

 

[abbey road d enfer]

The difference between the two circuits according to AP2 was exactly that, 20dB difference of H2 at @ below 18bit dynamic limits.

Now, the difference between the two circuits was heard clearly in blind testing.

But that would not make me conclude that the difference between -110dB and -130dB H2 is audible with music
We know it is not.

So often I suspect people hear differences, between devices, but what they hear is not the difference in the advertised figure of merit.

They may well have heard a difference in IMD. Not always, but most of the times IMD and THD increase hand in hand.
IMD, not being harmonically correlated with signal, is much more noticeable.

 

[abbey road d enfer]

IRF730 or IRF830. These are Mosfets.

In this case I recommend:

T1 = IRF830
R1 = 6M8
R2 = 1k (this is a gatestopper in series with with the gate)
R3 = 100R (this is a protection resistor in series with source/emitter before Vout)
ZD1 = 1N5245B (or similar - Gate protection diode from Vout [anode] to Gate [cathode])
C1 = 100nF/450V Film (e.g Wima MKP2)
C2 = 100uF/450V Film or Electrolytic (this capacitor is attached between Vout & GND)
View attachment 109510

Out of curiosity I made a simulation of this circuit and I have quite different results, with HF regulation tending to zero. The MOSFET seems to behave as a negative impedance at about 200kHz.
Also I made a comparison with a similar circuit using a Darlington.
The Darlington behaves as expected.
The output impedance of the circuit is entirely dominated by the 100r series resistor and the 100uF shunt capacitor. No benefit from the active circuit in this area.

 

[thor.zmt]

They may well have heard a difference in IMD. Not always, but most of the times IMD and THD increase hand in hand.

In this circuit IMD & HD are indeed rising in proportion, however IMD is lower than HD. More to the point, the lower HD circuit used aggressive HD cancellation in the output stage and increased loop feedback.

IMD, not being harmonically correlated with signal, is much more noticeable.

Are we suggesting that the IMD cause by nonlinearities that are H2 dominant SMD have < -110dB HD are audible at normal listening levels?

I consider this a bit unlikely.

If we can agree that HD and IMD at -110dB H2 dominant and -130dB H3 dominant should be inaudible, then it cannot be the difference in HD & IMD that are audible.

Instead, the difference in HD/IMD are indicative of different and audible differences between the two largely identical circuits, differences not captured directly using traditional tests.

Thor

 

[abbey road d enfer]

Are we suggesting that the IMD cause by nonlinearities that are H2 dominant SMD have < -110dB HD are audible at normal listening levels?

I consider this a bit unlikely.

Then, what is it that they heard? Something that's more an elephant in the room than THD/IMD/whateverD, but is unnoticed by measurement?

If we can agree that HD and IMD at -110dB H2 dominant and -130dB H3 dominant should be inaudible, then it cannot be the difference in HD & IMD that are audible.

THD and IMD are very crude tests that don't really account for the interactions between multiple complex signals. I'm inclined to think that the actual distortion effects people hear are far higher than these figures. Also audition has a unique capability of pinpointing and discriminating that no traditional test equipment has.

Instead, the difference in HD/IMD are indicative of different and audible differences between the two largely identical circuits, differences not captured directly using traditional tests.

Since the times where traditional test have been criticized/rejected, no one came with a more significant and commonly accepted test.
The Rohde & Schwarz system PEAQ does not seem to be adequate.
The Mosquito app seems to be right on the money if it delivers what it promises. Let's see.
I have a strong belief that if something is audible it's measurable.

 

[JohnRoberts]

I have a strong belief that if something is audible it's measurable.

In my experience I could measure distortions that I could not hear, but I never heard a distortion that I couldn't measure. That said I had to roll some of my own semi-custom test equipment to measure two-tone IMD before it was widely available.

I am pretty sure that people hear IMD. They just don't know what it is.

JR

 

[thor.zmt]

Then, what is it that they heard?

Good question. I am asking myself that as well.

I have a strong belief that if something is audible it's measurable.

I agree. I also have a fundamental understanding that not everything that is measured is audible AND that not everything that is audible is currently covered by standard measurements.

Thor