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

[Matador]

This really should be the quote that threadkill's this thread:

if you just checked, it was surely a bad measure...
however this won't change my point of view

"In God we trust, all others bring data."

 

[ricardo]

Max shows that 2 x 408a in parallel works well.

Anyone tried the Saturn Sound 407a (supposedly with both triode halves in parallel)?  Anyone confirm Oliver's assertion that this is the Saturn Sound replacement?
______________________

Meanwhile, in the Bat Cave ...

Because the capsule has capacitive output Z, any capacitance on the input of the valve, whether real or dynamic, doesn't affect frequency response [1] but LEVEL.

Try a LTspice simulation, using a FET (or a tube if you have a good SPICE model) and varying Cgs & Cgd

[1] There are effects but well outside the audio range in a competent design.

 

[granger.frederic]

Dear Matador

this is a misunderstanding , i wanted to say that ioaudio probably made a bad measure

the only thing that's killed is your credibility, you can't stand different opinions , and it's now clear that you're here to only make some business

Dear Ricardo, the miller effect is in cause , google it....!

Dear kcatthedog, feel free to read GS if you don't like my posts or tell us something interesting...

Sorry but it's enough for me , i came here to share my humble knowledge for free, but now i'm disgusted ...

I give up !
Ciao !

 

[kcatthedog]

Yoo hoo Frederic ,

Don't go away angry, its too bad we weren't all enjoying a few bottle or more of good wine and talking this all through cus the internet is a bad communication forum.

It just seemed like you were splitting hairs.

When  knowledgable, credible people like Max and Matador get shot down for what seem like valid and well documented positions the counter argument seems more personally motivated: like the thread topic theme to start with.

If you like the design, buy it and enjoy a great mike, if you don't: don't: full stop ! 

 

[ricardo]

Dear Ricardo, the miller effect is in cause , google it....!

Perhaps you should also research Miller effect on capacitive sources. 

 

[shabtek]

407a does not work; sounds like broken overdrive fuzzbox with dirty contacts missing battery in bypass mode.

don't bother

 

[ricardo]

407a does not work; sounds like broken overdrive fuzzbox with dirty contacts missing battery in bypass mode.

Anyone confirm that the Saturn Sound replacement ISN'T 2x407a in parallel?

ie was Oliver wrong?

 

[granger.frederic]

Dear Ricardo, no need to look far : groupdiy.com/index.php?topic=12529.0
cheers

i'm not spitting hairs, i have another opinion, that's all.
should i be blamed for that ?

"When  knowledgable, credible people like Max and Matador get shot down for what seem like valid and well documented positions the counter argument seems more personally motivated: like the thread topic theme to start with. "

here we are:

1: "credible" am i less credible than others ?
2: "well documented" : where are the well documented datas about the max and matador kits ?
3 : what you find "seem like valid" is not always for me
4:  motivation: i'm here (maybe was) for free, i don't want to make some PROFIT

to conclude : Two 408A design shows a lot of non described issues :

1: microphonic noises

2: lower than capsule  input impedance thus the low end suffers (http://www.neumann.com/forums/view.php?bn=neumann_archive&key=993770179&v=f) and thanks Oliver again...

3: doubled miller effect that softens a bit the transients , but here there are surely other causes

4: the design don't react like a true VF14 : the power gain is increased and the current also (not the voltage gain)
this produces a different  BV8 reaction

5: if the tubes are not perfectly matched (that's complicated) there is some non linearities

6: i don't give a S.... who has invented it

However it works, and for some reasons some people like it.
i can hear that.
so please hear me.

 

[ricardo]

Dear Ricardo, no need to look far : groupdiy.com/index.php?topic=12529.0

You may like to study dale116dot7's posts in that thread.

 

[granger.frederic]

are you trying to say that the Miller effect is null in a tube microphone ?
you can't be serious...

 

[kcatthedog]

Morning Frederic,,

Glad you are hangin in as you do obviously have a lot of knowledge to share and I do respect that.

You said:

"However it works, and for some reasons some people like  it i can hear that.
so please hear me."

I do hear and understand your perspective.  I think you have clearly identified the nub of the issue in your comment:

"However it works, and for some reasons some people like  it ",  exactly if the individual is happy/satsified wtih the mike : knowing it uses the  2 tube design that is their perogative.

People have been frank on the build threads about needing to test tubes and trying different tubes; that is true of almost all tube gear.

I am for continuing to understand and to  constructively discuss the typology of this build and its implications, but for me the bottom line is not criticism on paper but the sound of the mike and its usabililty, both of which according to users is very very good and very useful.

Forgive me, if my comments seemed dismissive, which was not my intention.

 

[ricardo]

are you trying to say that the Miller effect is null in a tube microphone ?

The Miller effect is alive and well in microphones.

But the effect isn't on frequency response but on level.  Neumann (& others) use it to adjust level.

Why don't you try and understand dale116dot7's posts.  You might also like to check out Scott Wurcer's articles on LN mike electronics in Linear Audio.

Better still, if you can measure microphone frequency response .. measure one of yours.  Then stick a 100p NPO/CGO ceramic on the grid of the tube to ground (or a 5p from Anode to grid for Miller) and measure it again.

 

[Matador]

Granger, one of the things that makes this forum stand out for me (and maybe others), is that it's a place were we can discuss facts, not truth.  To quote Abbey's signature:  it doesn't matter who's right, it matters what's right.

If you make an assertion (and especially an assertion about other people's assertions), you should be prepared to debate the facts, and to back up your informed opinions of them.

For example, you have made many assertions in this thread:

[quote author=granger.frederic]
two tubes in parallel are adding their noise , inter electrodes capacitance ( higher miller effect : different transient response) and other issues if they aren't perfectly paired
[/quote]

What is the change in noise?  What is the mechanism and basis for your noise assessment?  What is the weighting factor you are using?  What exact tubes were you looking at at the time?  What is the circuit action that links "different transient response" due to the paralleling of tubes in this application (a pentode tube)?

[quote author=granger.frederic]
thus the noise floor doesn't really comes down with paralleled tubes
[/quote]

All of the same questions as above.

[quote author=granger.frederic]
In fact , Ioaudio's intentions are,  in fine , to sell his work on this site
[/quote]

Do you really feel qualified and knowledgable enough to assert, as fact, what IOAudio's intentions are?  I certainly don't.

[quote author=granger.frederic]
if you just checked, it was surely a bad measure
[/quote]

Why is it bad?  How did you do your measurement?  What equipment did you use?  Was it consistent from tube to tube?

[quote author=granger.frederic]
When i tried your design, It worked, but for a professional use, i think that the silence was essential, and the capacitances were softening a bit too much the transients...
[/quote]

The same claim again:  you are saying the "doubled capacitances" are causing a shift in frequency response.  What is the circuit action?  What frequency response measurements have you performed?  Do you have any Spice data to back this up?

[quote author=granger.frederic]
Your PCB material is maybe good enough in the beginning but after several years of ambient moisture , nobody knows...
[/quote]

Again, more assertions.  After how long?  Under what conditions?    Assuming what PCB material?  Assuming what exact formulation of solder mask?  What data/evidence do you have to substantiate a claim that after several years of moisture, then it will be worse?  Can you point to any studies?  Any studies by anyone?

If we discuss facts and data, then all of us become better informed and more knowledgable.  If we back up our assertions with nothing but subjective opinions then none of us can progress forward, and we might as well be talking about "Do you like Edward, or Jacob?". 

And to Ricardo's point, if we want to convert our subjective opinions into actionable theories, then experiments are in order.

 

[granger.frederic]

Matador,

what a stupid pseudo...

i actively dislike the bullfighting, and that's exactly what you're trying to begin here : a fight

excuse me to have another opinion and to debate in this forum !

i'm too busy to answer you with all the measures and references, but i you check all my posts since one year in this forum, you will have a good sample. good luck!

my advice is : do your business and profit selling PCB clones, and have a nice day ...!

as far as i'm concern , the tone begins to displease me and i leave definitively the place

good luck everybody....

 

[fazeka]

I dunno man. Matador's asking some valid specific questions. I, for one, would love to hear what you have to say in response to those in an effort to better understand as a student of all of this knowledge. You either answer/contribute constructively or you don't. I can't state it any clearer.

 

[kcatthedog]

ok that is what I don't get,,, I do think it is fair to ask each of us to substantiate our opinions or for us all to "man  up" and say ok maybe I was being just a little teesnsy , weensy opinionated, which is ok too cus we are all human and are probably here,
cus we are passionate about the gear, its authenticity, history and modern day recreations.

I think these are values we share and a concern for quality and value for money

don't think we have to bicker or attack one another at all

guess I will just try to appreciate my lowly little mk-u4y7and blue cap; it will be tuff ,,guess I will just have to man up and bare with it

Help !

 

[granger.frederic]

Are you kidding ?

i 've answered to half of matador's question in this topic : just read !

the rest is in the other topics for most

about Spice : Matador could you give us a MODEL for a 408a , i'm terribly curious ...?

To explain my attitude:  whatever i say , max or matador will prove the opposite because i will go against their investment.

Did Matador or Ioaudio ever give detailed specs about their products ?

they're answering to a question by another question......!!!!!!!!!!!!!!!!!!!!!

the other main reason is that i don't want to help or debate with this kind of people.period.

And for people who prefer not to hear me , my advice is to put a MK47 and a true U47 ,  record and compare !

For those who like my different point of view please PM me like you do at this time.

Cheers

 

[EEMO1]

can't believe this shit is still going on.

 

[Matador]

Around these parts, the burden of proof is on the accuser, but you know what, fine.

Most people here know about a resistive voltage divider, and the magnitude of the output voltage is a function of the top and bottom impedances:

So most people know the output voltage Vout wrt. the input voltage Vin: it's just Vout = Vin*(Z2/Z1+Z2).

Of particular note is that if Z1 and Z2 are purely resistive, then there is no frequency component in this equation.  If Z1 = Z2, Vout = 0.5 * Vin for all frequencies...whether or not Vin(t) varies at 1Hz or at a gazillion Hz.  In fact, if you need a 0.5 multiplier, there are an infinite number of Z1's and Z2's that will achieve this.

Now what if Z1 and Z2 are not purely resistive, and are in fact purely capacitive?  The equation stands, however Z1 becomes Xz1 and Z2 becomes Xz1:  in other words, the Z1 and Z2 components now have a frequency component.

Since the exact math bores people:  at a high level, we can take the Laplace equivalent circuit, and see that both the numerator and denominator both have a term that is proportional to 1/sC, which means I can simplify by multiplying by the constant term s/s.  In short, this means that the frequency component cancels out of the above equation.  This means that the input to output voltage ratio is constant over frequency, much like the resistive voltage divider.  The exact equation reduces down to Vout = Vin*(C1/C1+C2).  Notice the difference!  Unlike the resistive divider which increases the ratio by increasing the bottom component, for the cap divider the ratio increases by increasing the top component.

So what is the actual difference you might ask?  The output impedance of the circuit does change over frequency, unlike the resistive voltage divider.  If the top and bottom caps are 100pF, then at 100Hz it looks (and behaves) like a resistive divider with both resistors equal to  15Mohm.  If the frequency increases to 1MHz, then it looks (and behaves) like a resistive divider with both resistors equal to 1.5kOhms.  It's like a resistor divider where both resistors scale equally as a function of frequency.

So lets bring this back to microphones:  in a microphone, the top cap is the capsule, and the bottom cap is the input capacitance of the tube.  In reality, the bottom capacitor is the parallel combination of the grid resistor and the input capacitance.  From a frequency response point of view, it doesn't matter:  the frequency terms still cancel.

By the capacitive equation above, something interesting emerges:  if there were a way to make the bottom capacitance bigger, then the ratio would cause the overall output to reduce.  This is the standard way to make a pad:  you add a capacitance effectively in parallel with the tube input capacitance, and the overall output level drops, but the frequency response doesn't change.  For example, look at the U67:  there is a switch that adds 500pF attached to the output of the capsule, leading down to the backplate, which is sitting at an effective AC ground.  This effectively increases the input capacitance of the tube (as seen by the capsule), and provides the pad function.  Miller effect does the same thing:  the bottom cap increases, which causes the level to drop at all frequencies.  I would bet that one could construct a pad with a tetrode, by switching its potential between the cathode and the plate.  This would engage and disengage the Miller effect and cause a pad-like effect (all other things being equal, of course).

If this isn't clear enough for anyone to believe, I can also provide a Spice simulation which shows this effect quite clearly.

To summarize:  the Miller effect of the tube does cause a change in level, but with purely capacitive sources, it doesn't cause a change in frequency response.

 

[marcus4audio]

Sounds great. Thanks for simple elaboration. Now I wonder why my U67 sounds so bad with that  500pf pad switch?  Any scientific explanation?