St*der D19 Mic Pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Samuel Groner

Well-known member
Joined
Aug 19, 2004
Messages
2,940
Location
Zürich, Switzerland
Hi

As St*der has started to publish some schematics, I found this interesting mic pre:
D19_MicAD_Op_Serv47.jpg

Could someone give me a hint as how the first active stage works? It looks to me as if it were an inverting stage and the transformer and the mic set the source impedance and thus the gain. But this cannot be true as this would result in a rather un-flat frequency response, to say the least.

Wait, the transformer is somehow included in the feedback loop; does this contribute to the solution? They seem to switch the turns ratio as well.

Thanks for help!
Samuel
 
It least it's a new approach!

But using an opamp, any time, in the inverting configuration, means adding 6dB of noise gain. That's inherent in the topology of the inverting mode.

Perhaps those lost 6dB is not an issue here, but I very much doubt it.

I suspect say this design brought on primarily to facilitate remote control of the gain. By putting the switches in the virtual earth point, they avoid any significant voltage swing across the switch.

It's an interesting approach but certainly not optimal from a noise point of view. Inverters have no place as the first stage in mic preamps - period.
 
As Jakob said, it's in the Meta, but for your convenience:
ftp.studer.ch

Finished downloading yesterday, 3.7 GB all in all. The 990 console is really worth a look at.

I feel like the line could have been faster considering that I'm only one or the other kilometer away from the factory...

Mats, I'm not sure if this acts like a simple inverter; that would give very messy frequency response - i.e. the transformer output impedance looks like some hundred pF to ground at high frequency - this would give a tremendous ultrasonic shelving EQ, wouldn't it?

It must have to do with the transformer beeing inside the feedback loop, I guess. Hm, maybe we get some sort of bootstrapping here? Or R151/C135 correct the HF response? :? Help!

At least I got the feeling that it is not possible to simply rebuild this with a different transformer.

Samuel
 
sorry for beeing off topic:
anybody looked at the tube section of the D19MicValve? (same MicPre section btw).
I like the parameters: Bass Warmth, Angel Zoom, Valve Drive/Clip ...
 
[quote author="Samuel Groner"]Mats, I'm not sure if this acts like a simple inverter; that would give very messy frequency response - i.e. the transformer output impedance looks like some hundred pF to ground at high frequency - this would give a tremendous ultrasonic shelving EQ, wouldn't it?Samuel[/quote]

I have not really looked at this very closely but with the '+' input being tied to common, that really don't give you many options.

But a simple inverter it is not :?
 
Number of odd things from a noise standpoint, including the recycled common out of a section of the MC33078, which imposes its voltage noise---can't understand that rationale. But that's after a bunch of gain probably so it's not important I guess.

I suspect part of the function of the transformer feedback winding is to control the impedance presented to the mic, besides setting the gain of the stage. It would be nice to know the turns ratios.

Note the switchable ~100Hz 2 pole highpass on the output.

An odd one.
 
whats going on after the second transformer (t7)?

If you could dump all that, you'd have a real simple project there. Anyone know what kind of transformers they used in those consoles?

dave
 
[quote author="soundguy"]whats going on after the second transformer (t7)?

If you could dump all that, you'd have a real simple project there. Anyone know what kind of transformers they used in those consoles?

dave[/quote]

The first trafo is just a common-mode choke. Yeah, it would be good to know about T7---I suspect it was made custom for Stud*r.
 
By accident I found an article by which discusses this input arrangement; it's in Gayford's "Microphone Engineering Handbook". I think PRR once posted a scan of these pages, but I can't find the link right now.

As a short conclusion, there is no significant noise disadvantage with this type of input amplifier.

No good?

Samuel
 
That material is interesting as well for the discussion of providing a lower-than-thermal-noise termination via feedback.

Gayford is hard to find---no copies for sale that I can see :sad:

EDIT: and five people in line ahead of me on Amazon!
 
Thanks clintrubber, that's the one.

How is gain determined for fig. 8.35a? And could someone shed some light on why source impedance does not influence gain (except for loading effects)?

Thanks...

Samuel
 
[quote author="Samuel Groner"]Thanks clintrubber, that's the one.

How is gain determined for fig. 8.35a? And could someone shed some light on why source impedance does not influence gain (except for loading effects)?

Thanks...

Samuel[/quote]

The magnitude of the gain is determined by the current---the transformed source voltage (unloaded) divided by the sum of the transformed source impedance and the synthetic input impedance---times the feedback resistor R1. The sign is negative, but of course the trafo secondary winding connections can be reversed if necessary.

When the source impedance is zero the current is due entirely to the voltage divided by the synthetic input impedance given by the formula shown. The R1 term cancels out and the voltage gain is -(R2 + R3)/R3, except for sign like that for a standard noninverting voltage amp.

If there is finite source impedance, the transformed impedance is added to the synthetic impedance and the attenuation is just like you would get with a more conventional resistive load in parallel feeding a voltage amplifier. For example, if the synthetic Z is equal to the transformed source Z, the gain will be cut in half---the current is cut in half since the transformed voltage is driving twice the Z.

The second figure, 8.35b, is a somewhat more conventional approach, but it assumes a black box with a well-defined inverting gain and a high input impedance of its own. Since we are accustomed to making such boxes out of op amps with feedback and input R's, but would then have to buffer the input R to make the input Z dominated by the external resistor in the figure, the first circuit does have its advantages.

Note that things begin to fall apart when the frequency-dependent open-loop gain of the amplifiers falls a lot. The circuits work best when the open-loop gain is constant, as opposed to the more typical 6dB per octave falloff. As well, the voltage noise of the amplifier has to be small for the equivalent noise in the synthetic impedance to be small.
 
Thanks for the answer - took me some time to think it all through.

The circuits work best when the open-loop gain is constant.
This cries for a discrete implementation - how do you guarantee stability with such an approach?

The voltage noise of the amplifier has to be small for the equivalent noise in the synthetic impedance to be small.
So the amp needs a lower OSI than with a more usuall approach? What would be an optimum turns ratio for the input transformer given a specific amp and OSI?

Samuel
 

Latest posts

Back
Top