Capacitance multiplier: which Darlington to choose?

[Walter66]

Thank you, Ian!

 

[abbey road d enfer]

And regardless of chosen parts, all capacitance multipliers have what I consider as a fatal design fault. Any overcurrent fault on the output, the pass transistor usually fails with a collector to emitter short circuit.

That's why I mentioned the need for a resistor between the cap and the base.

 

[abbey road d enfer]

Mr Shindo is a pragmatist. In the claret he used only RC filtering and it sounds great too.

Can you expand on how (apart from the effects on noise performance), a regulator can influence "sound".
Unless the stages are so poorly designed the rail impedance and (lack of) PSRR degrade performance.

 

[Walter66]

Dont want to start a war about subjective Sound impressions, my EL156 series Regulator has a huge Impact on the linestage Sound. The 50W pentode Sounds rich and full, But seem to tame Dynamics and Impulses of music. Thats why maybe an unregulated Output Stage can have Sonic Advantages. Fairchild 670 did it this way, Western Electric 130B didnt. I suggest to try it in practise, Gray is all theory.

 

[Matador]

You can use a much lower voltage BJT to sample the current across the 100 ohm series resistor in the MOSFET design above, to steal the gate drive and lower dissipation in the pass element in the case of shorted output. If you really want to get fancy, you can use a high-voltage regulator on the gate and get a regulated supply with just a few more parts.

And agreed: you can play with the values of the RC filter on the gate in order to get different ramp-up times, which are a huge advantage in a tube design.

 

[ccaudle]

Unless the stages are so poorly designed the rail impedance and (lack of) PSRR degrade performance.

Shindo is known for producing amps using single ended triode circuits (SE-300 power triode). Think 1930's design.
I would expect the circuits have very little PSRR.

 

[thor.zmt]

What about IRF830 when applied in this circuit? It has an internal protection Diode already implemented. So why Not Do it this way, please? Resistor value can be changed.

My circuit is like this, just elaborated with measures to make the circuit more robust.

A Mosfet has a much higher transconductance than a BJT, even a triple emitter foloower / triple darlington. This means when used as follower it will easily make an good Colpitts oscillator. In my experience careful implementation (mechanically) allows the gate stopper (R2) to be omitted, but the circuit is much more reliably replicated with R2.

You could omit R3 & Z1, they just make the circuit much more robust in practice.

All this is easily wired on a small vero board attached to the legs of the TO-220 Resistor.

My suggestion is to keep these items. You have a much greater chance of everything working fine.

Thor

 

[Walter66]

Thanks For your explanation, Thor.

 

[thor.zmt]

Can you expand on how (apart from the effects on noise performance), a regulator can influence "sound".
Unless the stages are so poorly designed the rail impedance and (lack of) PSRR degrade performance.

Many circuits, especially older single rail Tube and Solid state circuits have very limited PSRR.

Even modern circuits often rely on looped feedback for PSRR and can be improved by regulating supply rails.

It must be remembered that currents flow in loops (Kirchoff) and powersupplies usually complete the current loops.

Any non-zero impedance will produce signal correlated error voltages in the power supply rails, which in turn enters the circuit through the power rails and the limited PSRR of the circuit.

Material distortion in, for example, electrolytic capacitors is well documented.

Thus the potential of Sonic degradation always exists, even if only as re-entrant distortion. And no, the argument "if the circuit has poor PSRR it's badly designed" doesn't hold.

In many applications with high gain the NE5534 (for example) has poor PSRR.

Thor

 

[abbey road d enfer]

It must be remembered that currents flow in loops (Kirchoff) and powersupplies usually complete the current loops.

Exactly. That's why I try to avoid as much as possible to rely on the PSU's stiffness. Actually I deliberately introduce power distro resistors, so the loop is closed via the decoupling caps rather than by the regulators. Indeed the decoupling caps must be dimensioned for the task. The typical 100nF ceramics are not adequate for this task, they are mainly for stability. The bulk of the current is handled by large electrolytics.
As a result, the interaction between stages is reduced accordingly.
Electrolytic technology improvements have allowed that.

 

[thor.zmt]

Im seeking For an elegant Method of further filtering the PSU For a phonostage and got the Message, IRF830 seem to be a very powerful Filter.

Yes, it is.

Maybe you have an opinion wether I should Filter the whole Tube preamp PSU with Tube regulated voltage by using Tube series Regulator (WE 130B did it this way) or leave the Output section For passive filtering (Fairchild 670 Style PSU Filter). Fully Tube regulated PSU seem to Limit Impulse responds, means Sound May become smoothed Out in Terms of Dynamics. I figured this Out by using an old Tube regulated EL156 PSU vs. unregulated CLCRC filtered PSU. Complete different Sound. But I want to Go Pro active Tube regulated PSU Like WE or Fairchild did.I Like those Classic designs. Because of that importance of the Filter Method to the Sound Im seeking advise.

I have build many designs over time, all tube, tube regulated and rectified (I also build a Claret Copy among many others).

Something that I'd consider the cumulation of my Journey in Tubes is documented here (300B SE Amp, Transformer Volume Control, LCR EQ Phono...

300B-SE-Amp von Thorsten Loesch

Wieder mal ein neues stueck Voodoo vom Hohenpriester des kultes des Audio und Roehren Voodoos

All Powersupplies are LC/RC filtered without regulators of active filtering of any kind anywhere in sight. Strict avoidance of polarised/electrochemical capacitors. Minimal number of stages using high transconductance / high perveance tubes, no looped or local feedback.

Nowadays I listen with a restored Marantz PM-75 and all sources are digital. I live by the beach and moved country or city five times in two decades... I need a simple system these days.

Thor

 

[thor.zmt]

Exactly. That's why I try to avoid as much as possible to rely on the PSU's stiffness. Actually I deliberately introduce power distro resistors, so the loop is closed via the decoupling caps rather than by the regulators. Indeed the decoupling caps must be dimensioned for the task. The typical 100nF ceramics are not adequate for this task, they are mainly for stability. The bulk of the current is handled by large electrolytics.
As a result, the interaction between stages is reduced accordingly.
Electrolytic technology improvements have allowed that.

There are many ways to achieve a given final result. My preference is often for shunt/parallel regulation or "very fast" series regulators (faster than the supplied circuit) for each individual functional block or stage.

Ultimately an Engineer sets requirements, evaluates solutions against them and then iterates until the requirements are met. Some define requirements strictly in abstract objective performance, for which there is no demonstrated or proven reliable links with sound quality, others focus strictly on subjective sound and still others synthesise to differing degrees.

The proof of the engineering is if the machine maketh a sound, pleasing to the ear.

Thor

 

[Tubetec]

Cheers for those insights Thor ,

 

[abbey road d enfer]

There are many ways to achieve a given final result.

Indeed. I'm not an imaginative guy, and I don't know what is the receipe for good sound; however I know a few receipes for achieving objective performance results. That's what I've done for about 40 years and never regret. Whenever I made mistakes, it was due to my own stupidity, not my modus operandi.
The problem with subjective evaluation is there are as many test results than there are pairs of ears.
And how does one translate a perceived defect into a corrective action, I don't know...

 

[thor.zmt]

Indeed. I'm not an imaginative guy, and I don't know what is the receipe for good sound; however I know a few receipes for achieving objective performance results. That's what I've done for about 40 years and never regret. Whenever I made mistakes, it was due to my own stupidity, not my modus operandi.

Different strokes for different blokes.

The problem with subjective evaluation is there are as many test results than there are pairs of ears.
And how does one translate a perceived defect into a corrective action, I don't know...

Therein lies the rub. Mind you, compared to 40 Years ago we have a much better understanding of how human hearing works and it is possible to adjust objective goals in line with this.

I find that an excessive objective performance focus makes one blind (and ABX deaf) to less than glaring but instead subtle fidelity impairments not covered by traditional measurements (at one point TIM and Jitter were such fidelity impairments) while an excessive subjective performance focus makes one blind (and deaf) to glaring objective fidelity impairments that ultimately can become subjective problems with different music than used in subjective testing (e.g. acoustic Jazz/Classical vs. EDM).

Worse, any excessive emphasis often leads to a quite intolerant and belligerent mindset in the face of challenges to ones worldview. I find that keeping an open mind and rather to seek conflict between Thesis and Antithesis trying to learn from all sides and engage in synthesis gets me further towards somewhere where gear measures respectable and sound great.

Thor

 

[Walter66]

There are many ways to achieve a given final result. My preference is often for shunt/parallel regulation or "very fast" series regulators (faster than the supplied circuit) for each individual functional block or stage.

Ultimately an Engineer sets requirements, evaluates solutions against them and then iterates until the requirements are met. Some define requirements strictly in abstract objective performance, for which there is no demonstrated or proven reliable links with sound quality, others focus strictly on subjective sound and still others synthesise to differing degrees.

The proof of the engineering is if the machine maketh a sound, pleasing to the ear.

Thor

What can be auditioned with newer Shindo Gear (which is IC regulated PSU in Combination with Tube rectification) is a very Low noise floor. It cant be Judged from the amps noise Level wether its switched to phono or Line Input. Thats remarkable Low noise, I dont have achieved such a black Background with passive filtering yet. We're talking of Tube amps in Combination with 100dB horn speakers, which I employ and Shindo offers, too. Out of that black totally silent Background, Impulses seem to be generated with much Dynamics and ease. Its still a mystery to me, how He achieved this. He must have tried some technical solutions and have found one which Sounds good and provide extreme good PSRR at the Same time. Not an easy Task, many have failed to achieve. Im aware that Shindo is still a Sound create Producer instead of a Studio Sound Producer and thats where I want to move in that direction, but His actual PSU capabilities, achieved with relative simple IC circuit solutions, are quite astonishing.

 

[JohnRoberts]

Different strokes for different blokes.



Therein lies the rub. Mind you, compared to 40 Years ago we have a much better understanding of how human hearing works and it is possible to adjust objective goals in line with this.

If it wasn't subjective they would already be objective metrics.

Indeed I have paid attention to psychoacoustics, and ergonomics (human factors engineering) in the context of product design for decades. Numerous papers were published in the JAES last century.

I find that an excessive objective performance focus makes one blind (and ABX deaf) to less than glaring but instead subtle fidelity impairments not covered by traditional measurements (at one point TIM and Jitter were such fidelity impairments) while an excessive subjective performance focus makes one blind (and deaf) to glaring objective fidelity impairments that ultimately can become subjective problems with different music than used in subjective testing (e.g. acoustic Jazz/Classical vs. EDM).

That is a leap in mind reading even for here. Abbey is an experienced circuit designer with good judgement from my experience reading his comments.

Worse, any excessive emphasis often leads to a quite intolerant and belligerent mindset in the face of challenges to ones worldview. I find that keeping an open mind and rather to seek conflict between Thesis and Antithesis trying to learn from all sides and engage in synthesis gets me further towards somewhere where gear measures respectable and sound great.

Thor

Yes, that sounds worse... are you still talking about Abbey?

JR

PS; Back decades ago I found psychoacoustics most applicable in effects design, and for side chain manipulations inside dynamic processors. Path technology has improved so much since back then that many of these tricks to conceal noise floors or limited bandwidth have become moot.

 

[Walter66]

Shindo is known for producing amps using single ended triode circuits (SE-300 power triode). Think 1930's design.
I would expect the circuits have very little PSRR.

Single ended and Push pull Tube Power and preamps. The Design of class A SET or SEP is very old indeed, but still offers lowest distortion Sound. Modern Shindo hasnt much else in Common with those old dinosaurs.

 

[thor.zmt]

Thats remarkable Low noise, I dont have achieved such a black Background with passive filtering yet. We're talking of Tube amps in Combination with 100dB horn speakers

Back then I used Tannoy Monitor Red's, which I measured at 97dB/2.83V. My purely passive, purely "no electrochemical capacitors" designs were absolutely quiet on those speakers standing next to them.

Mind you, LCLCLCLC with 20H/120uF per cell is quite a filter. There is Feck All noise after that.

The 300B SE CLCL with resonant first choke kills +B noise by > 106dB but the output stage has a PSRR enhancing circuit that adds another 26dB noise reduction.

Any noise was 50Hz coupled via magnetic or capacitive means. It's not that hard to achieve. But it will be much bigger, heavier and costly than an IC regulator.

Thor

 

[thor.zmt]

If it wasn't subjective they would already be objective metrics.

Yet they all started out subjectively, or as marketing gimmick (eg. THD&N as quality metric, that is on Harold J. Leak who just changed the way he qualified rather generic Mullard circuits as "point one" series claiming his Amplifiers had .1% THD while others claimed 3% or 10% THD for the exactly identical circuits).

That is a leap in mind reading even for here. Abbey is an experienced circuit designer with good judgement from my experience reading his comments.

I spoke very generally, it applies as much to hardcore subjectivists as to objectivists and applying also outside audio. I did NOT target any specific individual, but more like...

PS; Back decades ago I found psychoacoustics most applicable in effects design, and for side chain manipulations inside dynamic processors. Path technology has improved so much since back then that many of these tricks to conceal noise floors or limited bandwidth have become moot.

If we are honest, every microphone, amplifier and speaker is to some degree an "effect".

What is the point of making a 200W/8R Amplifier with 0.00001% THD&N at half power when it drives a speaker with 0.1% THD&N at 1 Watt input?

Thor