Measured VF14 characteristic

[o3misha]

bypass or decoupling capacitors have an impact on sound in theory.
i admit that, no problem.

Cheers
Fred

C1 in U47 is working as c13 in AKG c12, though they placed in opposite sides of capsule. C1 also is filtering cap, -  its just second role. But c1 is much bigger in value 10nF vs.c13  1000pF. Thats why the influence of C1 is less than c13. Everybody, who knows how capacitors work (capsule in our case) admit it- c1 is signal cap. Neumann used best possible caps for those times, BTW.  My goal is to make it clear: we can use different types of c1 to correct a bit sound going from capsule. It is real thing, not just theory. I gave all  facts based on the laws of physics and my own test result to prove understand that influence.
In the main, I think you think correctly: lamp, capsule and the transformer - are incomparably more important than any other details of the microphone.

Moby, I uploaded the link, see my previous post.

 

[marcus4audio]

germanium transistors must run at low temperature if we don't want some noise.
a sometime important miller effect, is also limiting for a mic application...
wait and see ...

I'm looking forward to it. BTw, I never heard Anrdeas fet version of Vf14 but I'm sure it sounds good.  I also wonder how he managed that fet crcuit?

Maybe making the spice sim model of Vf14 can help with this? Is it possible to make it from numbers and curves we have here? Someone skilled for this?

 

[lampas]

I believe that you all know for this:

http://ethanwiner.com/U47-FET.html


Does someone try this circuit in real U47 ?

 

[granger.frederic]

I know this schematic
I've tested it in another mic : Neumann U87i (when i disconnect the feedback  )
I think it sounds good but a bit too clean for the U47
input/output capacitances in a silicon Jfet are much lower, thus the miller effect is much lower also, than a underheated VF14m in triode mode.
the distortion profile is probably different and arrives with more brutality.

Germanium transistors have a miller effect close to a tube: narrower bandwidth, so they produce a different sound than silicon transistors :more analog,vintage,round.

but the sidebacks are: thermal noise , input impedance , rare these days , few datasheets

 

[granger.frederic]

i looks like no spice mode for germanium tr. exists... :-\

 

[gyraf]

You can't get a germanium transistor to do the high impedance of a condenser front-end amp. Historically, tubes were the only way until fet's came around.

Jakob E.

 

[Melodeath00]

Oh, it's not you fault, this thread was already polluted by various personal attacks : deleted and modified posts, "you're a liar"...etc etc

Once again, Mr Grosser is an excellent tech, but it's too bad that his substitutes are for genuine U47 only.

I've experimented a lot of tube in my U47n, and today i think that the 13CW4 with a bias at 0,65V is the best substitution i've tested.

I've tested EF14 (with my tweak and without), EF12 , UF14 ,parallel 6028, EF800, EF844S (rare tube), 7598, 6SJ7, 6072A (plate follower and CCDA ), EF42 and 13CW4.

for me, the 13CW4 is better, followed by EF14, then EF12.

13CW4 cuts through a mix (like EF14) and has the softness in the highs (like EF12)

the only issue is that it's hard to pick a 13CW4 usable in a U47.

Cheers

Hi Fred, have you tried an EF800 with a different anode voltage? that would lower the output impedance, making it a better match with the BV8. That should give you a better high end and transient response, which are the problems you mentioned with EF800.

You also mention it has less low-end. Some people use 1G with an EF800, perhaps to increase the low-end and make it more like a VF14?

 

[marcus4audio]

i looks like no spice mode for germanium tr. exists... :-\

Well its true, but it's possible to make it. It's bit tricky to include all the parameters like noise, distortion and temperature dependance but it can be really helpfull to have basic model of Vf14 and possible discreete parts. That way it will be easier  to achieve same voltage, current, gain...the rest of the details will be possible to tweak on the bench
I used to know guy who is really skilled for this but I can't find his contact anymore :/ anyone here??

 

[granger.frederic]

You can't get a germanium transistor to do the high impedance of a condenser front-end amp. Historically, tubes were the only way until fet's came around.

Jakob E.

it seems like it exists some silicon germanium mosfet but i can't find a commercial reference
i'm on it...

however, we can easily obtain high impedance input (1000 Mohms) on bipolar transistors, with some complex circuits

http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1968/High%20Input-Impedance%20Amplifier%20Circuits.pdf

 

[ricardo]

however, we can easily obtain high impedance input (1000 Mohms) on bipolar transistors, with some complex circuits

http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1968/High%20Input-Impedance%20Amplifier%20Circuits.pdf

This is an important & useful article, worth careful study.

However, the BJT examples fall short of what's required by at least a magnitude.  Also the noise of the BJT examples is VERY poor for condensor mike use.

The FET circuits are good with a few resistor changes.

 

[granger.frederic]

the article was written in 1968
we could achieve today, a two or three stage circuit, with a large bandwidth,high input impedance and low noise...
even with some Ge-T
i will try to spice that asap

someone to upload the VF14 spice parameters, please ?

 

[bezen4uk]

the article was written in 1968
we could achieve today, a two or three stage circuit, with a large bandwidth,high input impedance and low noise...
even with some Ge-T
i will try to spice that asap

someone to upload the VF14 spice parameters, please ?

The nature of current noise in BJT didn't change much since 1968 - trust to ricardo.
The best current noise density you can achieve with multistage BJT is 40-100fA/sqrt(Hz), while single FET has 0.2fA/sqrt(Hz).

Regarding VF14 spice model - I tried to create the model by Norman Koren's soft (BTW, thanks him for that cool soft).
Usually it works very good, but this time I receive funky curves.
I should try another model core.

 

[marcus4audio]

-
Regarding VF14 spice model - I tried to create the model by Norman Koren's soft (BTW, thanks him for that cool soft).
Usually it works very good, but this time I receive funky curves.
I should try another model core.

That will be great

 

[granger.frederic]

there's for example mpsa18 : 6,5nV and i'm sure we could find much better
http://www.onsemi.com/pub_link/Collateral/MPSA18-D.PDF

i want to try some alternatives circuits , even with AOP...to be open minded...

but i hope we'll find a spice model for the VF14...
http://www.normankoren.com/Audio/Tube_params.html

 

[bfox99]

Dear Misha,

i think that something is wrong with your schematic, or, i can't understand it ....!

in the AMI schematic, the output transformer seems to have a specific winding to put at ground potential, the AC coming from the cathode, thus decoupling it...

I doubt that the BV8 has been designed for that ...but maybe all that exceeds my knowledge ...

This schematic is using a very clever trick to apply negative feedback from the other end of the transformer and sending it to the cathode.

The cathode is not driving the transformer. The cathode resistor is acting as the return path for the transformer and the voltage generated across the cathode resistor by the signal injects negative feedback into the circuit.  This results in wider bandwidth, lower distortion and a little less gain than connecting the transformer to ground.

BF

 

[o3misha]

Neumann used negative feedback in many microphones, thats why I proposed to try it in clone. I think that using transformer via this method we can get more natural sound, and we can avoid to use cathode cap.

 

[granger.frederic]

you've truncated my posts

i've said that this trick impose a positive feedback:

the anode/cathode are out of phase
then C2 inverts it,so it becomes in phase with the cathode
thus the feedback is positive (or near because of the phase rotations inside the transformer)


please, correct me if i'm wrong

 

[o3misha]

I think this is negative feedback, not positive. Occurs substraction of part of out signal from input signal. This is a negative feedback. Advantages of such a transformer use: expanding bandwidth, reducing the phase distortion, reduced influence differences in the characteristics of light, and thus - improving repeatability microphones. Cons: a slight decrease in the level of output, mitigating the signal fronts.

 

[granger.frederic]

i can't understand why it could be a negative feedback...

however, in a tube stage, the miller effet is important, and the feedback is near useless.
that's why we can't find a feedback in C12, U47 etc...
That's less obvious with FETs, and Neumann always use feedback in U87,U47fet...

An exception: the U67

 

[o3misha]

Did you try to use transformer such way? It is absolutely different sound in comparisson with usual selfbias. Negative feedback is very useful for raising headroom and getting better frequency response. Try it,-you will never regret. You will get also better low end. Anyway, U67 is not alone with feedback in schematic - there are CMV563 and its bro UM57 (why positive feedback???).
Distinguish between positive and negative feedback. Negative feedback input signal changes so as to counteract the change of the output signal. This makes the system more resistant to accidental changes. Positive feedback, on the contrary, increases output signal.