negative Feedback in a Schoeps circuit?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

maye4

Active member
Joined
Mar 9, 2013
Messages
43
Try to tame harhness in a chinese mic with Schoeps circuit.

I want to add a negative Feedback as neumann does.

But there are 2 signals come out of jfet ,and they have different phases...As a new modder,I'm stucked...

Any ideas?Thanks.
GenChinaMic.jpg
 
Interested to hear clever peoples ideas .....Im mess with a KM84 circuit.... I have read that a 3.3pf ish cap between FET drain & capsule connection is for neg feed back......could a cap be added to your circuit in a similar way ? ....

Interested to read more......
 
Do some research on the Neumann U-87 and MXL 2001. Use their schematics to figure out the feedback cap and try a similar value in your mic. You may need to tweak the value depending on what you want from the mic (brighter vs. darker).

-James-

P.S.- You will lose output by implementing a feedback cap. So, modder beware...
 
OK,I'll rephrase my question.
I know neumann's feedback circuit.But here the situation is completely different.After the FET,there are two signals which have opposite phase.It's the phase bothers me...
You see ,you can't just apply negative feedback to anti phased signal...
And I am not sure if it's ok to apply negative feedback before jfet...

below is neumann style negative feedback.
220975d1298229845-its-mxl-v67-mod-thread-again-mxl-v67.gif
 
You might try just increasing C3 and C7. It looks like these are already rolling off some highs.

C6 provides a ground path for the node where these caps meet, they act as shunts on the signal before the drivers.

If you really want to use feedback at the FET stage, you can do what Gary suggests and add a small cap from drain to gate. (You don't need a second NFB path on the source leg, because the signal there will reflect the drain signal anyway.)

There is actually some positive feedback between the source and gate, by virtue of the divider node that the gate resistor is connected to.

 
MagnetoSound said:
You might try just increasing C3 and C7. It looks like these are already rolling off some highs.

C6 provides a ground path for the node where these caps meet, they act as shunts on the signal before the drivers.

If you really want to use feedback at the FET stage, you can do what Gary suggests and add a small cap from drain to gate. (You don't need a second NFB path on the source leg, because the signal there will reflect the drain signal anyway.)

There is actually some positive feedback between the source and gate, by virtue of the divider node that the gate resistor is connected to.

Thank you very much.This is very helpful!

Other than what you said,wil it help to bypass the biasing resister using a capacitor?Just as C4 in v67 circuit.
 
There's quite a noise penalty using the Neumann feedback method like MXL V67.

There's a better method to make your Chinese mike sound like U87 in my MicBuilders directory under "ChinaMod+U87hybrid".
 
I have done some measurements on this circuit: http://www.groupdiy.com/index.php?topic=51769.0
The best way to 'tame' the HF boost, is to add a capacitor between the drain and the source of the FET.
Try 33 or 47 nF first, listen and increase or decrease the value.

You might try just increasing C3 and C7. It looks like these are already rolling off some highs.

Those capacitors are mainly there to block (V)HF. You would have to increase them to an unrealistic high value to get an audible effect on the HF response.

 
Ricardo can find your "ChinaMod+U87hybrid". not sure where to look in metas ? thanks
 
IMHO the fact that the "ChinaMod+U87hybrid" modification only influences the drain, causes an unbalance to the output.
I think it would be better to use an R-C (of just a C) filter between the source and the drain.
In this case the balance is maintained.
 
RuudNL said:
I have done some measurements on this circuit: http://www.groupdiy.com/index.php?topic=51769.0

I completely overlooked this thread, apologies to the OP.


The best way to 'tame' the HF boost, is to add a capacitor between the drain and the source of the FET.
Try 33 or 47 nF first, listen and increase or decrease the value.

You might try just increasing C3 and C7. It looks like these are already rolling off some highs.

Those capacitors are mainly there to block (V)HF. You would have to increase them to an unrealistic high value to get an audible effect on the HF response.

Yes, agree you might as well do it before the blocking caps to get smaller value (and only one cap).

 
RuudNL said:
IMHO the fact that the "ChinaMod+U87hybrid" modification only influences the drain, causes an unbalance to the output.

It's just a different approach. What you do to the drain affects the signal current through the FET. It is reflected at the source.

 
What you do to the drain affects the signal current through the FET. It is reflected at the source.

No. If you want to see it, simply build the FET stage and connect an oscilloscope to the drain and the source.
You will see that if you increase the frequency, the (AC) voltage on the drain will drop.
The voltage on the source however will remain the same, because it is a source follower. ( A ~ 1 x)
This causes an unbalanced signal.

 
Thank you,guys.That's the community spirit!

Finally...Thanks to ricardo's files,I understand what u87's feedback circuit is all about!!!

It acts like a dynamatic EQ!
 
RuudNL said:
What you do to the drain affects the signal current through the FET. It is reflected at the source.

No. If you want to see it, simply build the FET stage and connect an oscilloscope to the drain and the source.
You will see that if you increase the frequency, the (AC) voltage on the drain will drop.
The voltage on the source however will remain the same, because it is a source follower. ( A ~ 1 x)
This causes an unbalanced signal.

Oops, yes you're right of course - it becomes a common-drain stage at higher frequencies. You would need to use either two caps to keep the FET impedances balanced, or a single 'cancellation' cap from source to drain as you already suggested.

However ... it's perhaps worth pointing out that, even with the unbalanced (or rather, uneven) drive signal of ricardo's circuit, the output is still impedance-balanced nonetheless, even if the signal is not fully differential, and any disadvantage is debatable, no?

 
IMO the Schoeps circuit is not suited for capsules that require some kind of frequency correction (e.g. K67 style capsules).

The Schoeps circuit is basically unity gain. Okay, the signal is doubled by the phase splitter (+6 dB), but there's no amplification in either leg. So where's the gain you want to reduce via negative feedback?

Yes, you can roll of some highs by increasing the caps in front of the emitter followers or the output caps (which are pretty heavy at 22n) But that's not the same. That way, the preceding stage has to drive a heavy load at higher frequencies. Putting a cap between source and drain, too, is a heavy load on the FET. So what? Well, try recording a tambourine or the jangle of a bunch of keys. The highs will sound distorted.

By contrast, reducing the highs via negative feedback the U87 way will not compromise clarity but actually increase high frequency headroom a bit.
 
I agree with Rossi:  the Schoeps circuit is excellent for letting the true sound of the capsule come through.  I would switch to a K47-type capsule and leave the circuit wide-band.
 

Latest posts

Back
Top