IOaudio Inventor of the parrallel 408a tube configuration in U47 Clone ?

[o3misha]

Any sound examples of good and bad sound of your U67? Measurements?

 

[marcus4audio]

Any sound examples of good and bad sound of your U67? Measurements?

I cant make any soundsamples in next few days but also I'm not sure how to make it acurate because of performance/position... Anyway, I never liked the pad function on U67 (also on 87). Besides higher noise (what is clear why it's worse) it lacks some details. I hate to express like this but it sounds "slow" with pad "on". Is it connected somehow with input capacitance-resistance=time constant? Sorry if I ask dumb question :/
If the question is out of topic I'm really sorry. But someone here mentioned the lack of transients

 

[o3misha]

This is the problem that there are no measurements - only "feelings". And I , for example, often use pad on u67. On the drum overheads. And don't understand how I would have managed without it. And I don't see any changes in the frequency spectrum. The increased noise is absolutely natural, because the signal amplitude is reduced to a subsequent amplifier, not after.
Ps: I have often noticed that some forum members are trying to link their sympathy-antipathy to theory, forgetting about practical research, supported by measurements. This leads to useless discussions, instead of which only one statement: "I don't like it". Or - "I like it".  Not enough bottom with a 408a? And I don't have enough low frequencies with 90% VF 14s, which I tried in my original U47s. I have now bought Maxs  kit, which is not yet assembled. However, the transformer I already tried. And it is much better sounds than the most expensive transformer AMI, for me(!!!)

 

[poctop]

Any sound examples of good and bad sound of your U67? Measurements?

I cant make any soundsamples in next few days but also I'm not sure how to make it acurate because of performance/position... Anyway, I never liked the pad function on U67 (also on 87). Besides higher noise (what is clear why it's worse) it lacks some details. I hate to express like this but it sounds "slow" with pad "on". Is it connected somehow with input capacitance-resistance=time constant? Sorry if I ask dumb question :/
If the question is out of topic I'm really sorry. But someone here mentioned the lack of transients

http://repforums.prosoundweb.com/index.php?topic=1137.0
this is  a good read about pads and the why and the how and the why again.
Best,
Dan,

 

[granger.frederic]

This is the problem that there are no measurements - only "feelings".

I have now bought Maxs  kit, which is not yet assembled. However, the transformer I already tried. And it is much better sounds than the most expensive transformer AMI, for me(!!!)

it will surprise you, but i think you're right Misha,

books about microphones are very rare and not enough accurate usually...
Measurements in a mic are very complicated because of the ultra high impedance of the capsule (among other things)
we must use our ears and our knowledge in the same time.
i add that we often use the tube in a non linear part of the curves

that's why we can't trust Spice simulations wich produce a theorical approximation of the choosen design, and there is a total lack of pentode modelisations, even less in triode mode ... 408a in triode mode : don't even think about it ...

miller effect wikipedia definition :

In electronics, the Miller effect accounts for the increase in the equivalent input capacitance of an inverting voltage amplifier due to amplification of the effect of capacitance between the input and output terminals. The virtually increased input capacitance due to the Miller effect is given by

C_{M}=C (1+A_v)\,
where -A_v is the gain of the amplifier and C is the feedback capacitance.

Although the term Miller effect normally refers to capacitance, any impedance connected between the input and another node exhibiting gain can modify the amplifier input impedance via this effect. These properties of the Miller effect are generalized in the Miller theorem. The Miller capacitance due to parasitic capacitance between the output and input of active devices like transistors and vacuum tubes is a major factor limiting their gain at high frequencies. Miller capacitance was identified in 1920 in triode vacuum tubes by John Milton Miller.

your ears are telling you that one transformer is better than another, i respect that.

i really can't understand why it's so hard to share an opinion in this forum....

 

[o3misha]

There is no problem with "opinions". The problem in substitution of concepts. It's one thing to say- "I don't like"based on feelings. Another thing to say "it's not professional" based on feelings and not on knowledge about the subject. I am convinced that the Max's product is professional and competitive. It's a great job. I still think that EF 14 - least successful lamp from all that you have listed as a lamp for the microphone.

 

[gyraf]

could we move on...?

 

[granger.frederic]

yes, we could but it's very interesting...

call it just as you want: opinion, concept, feelings, knowledge...
that's exactly what we're doing all the time : giving our opinion, and this is the principle of every topics in every forums :

just 2 posts before: "And it is much better sounds than the most expensive transformer AMI, for me(!!!)"
just 1 post before : " It's a great job. I still think that EF 14 - least successful lamp from all that you have listed as a lamp for the microphone."

and i'm ok with that!
even if i think the opposite

but it's very sad to see how things are getting hot , when someone thinks different or go against economical interests

 

[MicDaddy]

C_{M}=C (1+A_v)\,
where -A_v is the gain of the amplifier and C is the feedback capacitance.

major factor limiting their gain at high frequencies.

High frequencies in the audio band or high frequencies kHz, MHz?

 

[Matador]

The Miller capacitance due to parasitic capacitance between the output and input of active devices like transistors and vacuum tubes is a major factor limiting their gain at high frequencies.

Again, nobody is claiming the Miller effect doesn't exist in a tube microphone (and this has been stated several times by posters other than me).  What I am saying is that with a capacitive source, it doesn't have the effect that you are claiming it does.

The origin of the difference is in how output sources are modeled:  generally, you can think of any voltage source as a capacitor and real resistance in series:  most "typical" tube circuits follow this model:  like a triode output stage could be modeled as an ideal voltage source, with a resistance in series (equal to the plate resistance in parallel with the dynamic plate resistance of the tube) to the coupling cap.  Once the frequencies become high enough that the coupling cap appears like a short (relative to the other impedances in the circuit), the Miller effect swings into full force.  It's why you would become frustrated if you pump 100MHz into the grid of a triode, and see absolutely nothing on the plate, yet everything is fine below 1MHz.  At some point parasitic inductances start to come into play, and other factors start to dominate frequency response.

In any case, none of these effects are playing in the audio ranges.  We're recording acoustic guitars and singers, not microwave ovens.  ;D

A capsule has no real series resistance component (this isn't absolutely true, there is always leakage current at DC, however I would hazard to guess it is no worse than a typical film cap which leaks a few pA), but even the insulation resistance of the capsule is in parallel with the capsules capacitance, not in series with it.  I would guess the "real" component of impedance is actually happening in the mechanical domain, and doesn't come in to play in the electrical domain, since we are cause charge to change by changing the C (e.g. Q = CV), which is an altogether different circuit action than we would see in a typical tube circuit.

 

[micaddict]

You know, after some editing, this thread could be(come) a valuable document, actually. 

 

[poctop]

You know, after some editing, this thread could be(come) a valuable document, actually. 

Maybe after all of this someone will be elected 
D

 

[Matador]

You know, after some editing, this thread could be(come) a valuable document, actually. 

Maybe after all of this someone will be elected 
D

We really should write an eBook - Analysis of Microphone Circuits, collectively authored by the GroupDIY community...

The 10 people that would buy it would be very satisfied I'm sure.  ;D

What we would need to know is the ESR of a capsule, then we could figure out where in the frequency domain the Miller effect became significant.

Let's pretend that the capsule shows ESR figures typical of a polyester film cap (I would guess in reality that the ESR of a capsule is actually much less, simply due to the distance between the membranes...but I have no measurements to back that up).  For example, the datasheet for a typical 1uF Vishay polypropylene film cap is 54 mOhms (that's 0.054 Ohms).  At 100uF, it is 2.5mOhms (or 0.0025ohms).

If we say the relationship between capacitance and ESR is linear (this is almost definitively not true, I'm just spitballing here), then a 100pF cap extrapolated from the above numbers would have an ESR of 540 Ohms.  If we greatly exaggerate the Miller capacitance, and say the tube looks like a 1000pF capacitor (typical high gain triodes look more like 200-300pF), then the -3dB point of the frequency response of 540ohms into 1000pF (also including 80pF capsule capacitance and a 100MEG grid resistor) happens at 4.3MHz.

In reality, the ESR and Miller capacitance are likely smaller (ESR would likely be immeasurably small), meaning the dip starts to happen at even (much) higher frequency.

I really think people who are looking to Miller effect to explain frequency response in a typical capsule + tube circuit are really barking up the wrong tree, and are best served spending their energy elsewhere.

I wonder if anyone has tried measuring the ESR of a condenser capsule?  Most online reports of people trying to measure small value ceramic disc caps show no results, because you need enormous bandwidth and the scope lead resistance (usually sub 0.1 ohms) always dominates the measurement.

 

[0dbfs]

I'd buy the book!

"GDIY Mic's: Myth-Busted !"

Capacitive pads and miller effect: don't different caps have different sounds due to many detailed elements and construction materials as well as implementation in the ckt?

Cheers!
-jb

 

[granger.frederic]

Even if the miller effect is more effective outside 16khz (audible range) ,
It doesn't mean that we can't hear it!
Especially with the transients....
An overshoot on a 20 khz square wave is clearly audible ...

I don't why the miller effect should be less significant with a mic capsule
The miller effect is inside the tube : it a parasitic capacitance  feedback beetween the output and the input

Put a jfet 2sk170 and then a j305 , all other parts equal, and hear it´s clearly audible that the 2sk170 is more soft on transients due to its higher capacitances

 

[ricardo]

http://repforums.prosoundweb.com/index.php?topic=1137.0
this is  a good read about pads and the why and the how and the why again.

There's a couple of inaccuracies in this.

The most obvious is his assertion that a 'Miller pad' like in U87 increases noise.  It doesn't cos it is done by feedback.

There's also a load of BS about the evils of feedback and how it mucks up phase.  In fact, in this case, the extra feedback makes the phase response MUCH flatter.

Guru Scott Wurcer has a couple of excellent articles in the dead tree magazine, Linear Audio, which explain this clearly and are mandatory reading for wannabe mike electronics (pseudo?) gurus.

But Klaus's main recommendation, that the pad is used as a last resort, is valid.

The U67 pad though, with capacitance across the capsule, does increase distortion.  There's an old article by Bob Schulein on the Shure SM81, and IIRC also by Sennheisser, in AES that explain the evils of stray capacitance.

Matador, your explanation of 'Miller' in condensor mikes is accurate & excellent. 

But for capsule ESR, its more like a Polystyrene or NPO/COG ceramic  ..  ie almost impossible to measure directly with simple methods.  I did it with 'noise measurements' at Calrec.

The main ESR is due to the air load including acoustic resistance.

But as you say, it is waaa....ay beyond any audio transient frequencies.

Put a jfet sk170 and then a j305 , all other parts equal, and hear it´s clearly audible that the j305 is more soft on transients due to its bigger capacitances

Err.rrh!  sk170 has a LOT more capacitance than j305 

 

[granger.frederic]

Too fast
Fixed

Do the test ...!
But only if you are less then 45 years old
After you will waste your time ....

Capacitive feedback makes the frequency response more linear it is true but it produce in the same time some phase rotations and then it changes the sound ( good or bad : matter of taste)
It´s the most well known factor in the audio world

Sometime it' s usefull to tweak with theory and simulators
Sometime it's more effective to listen...( and more funny)

 

[Matador]

I'd buy the book!

"GDIY Mic's: Myth-Busted !"

Capacitive pads and miller effect: don't different caps have different sounds due to many detailed elements and construction materials as well as implementation in the ckt?

Cheers!
-jb

That is a whole other thread my friend.  ;D

 

[ricardo]

Capacitive feedback makes the frequency response more linear it is true but it produce in the same time some phase rotations and then it changes the sound ( good or bad : matter of taste)
It´s the most well known factor in the audio world

Wanna show us some evidence of this truly amazing result .. especially about

"frequency response more linear it is true but it produce in the same time some phase rotations" ?

'Real life' would be good but even SPICE world evidence would be interesting.

Can you tell us which mike you did your sk170 & j305 experiments with?  Got a schematic for it?

 

[granger.frederic]

a few time ago , i was fixing a vintage hifi integrated tube amp : a Hitone H300
a very high end french product of the 60's with two marvelous Chretien output transformers
anyway, i focused on a quite huge overshoot on 20khz square waves...
i analyzed that i was probably the 12pf (only) on the Volume Pot
when i removed it, the overshoot disappeared and the sex appeal of this amp at the same time..
so i've quickly putted back the capacitor at its original place
in theory the overshoot (or parasitic over-oscillation) frequency is "wa....ay beyond" audible range

same story with C4 in a M49c, put a too big capacitor and bye bye the mojo...

just to say that theory and measurements are sometimes misleading