Studio Projects C1 circuit mods ONLY (keep original capsule)?

[Paul678]

    I recently got a C1, and it sounds excellent.  Definitely my best mic for acoustic
guitar so far. 

      For vocals, it might be a tiny, tiny bit on the bright side, but I say might, because I usually
end up adding a small 10kHz presence bump to my AT4047 anyways. 

    So when I use the C1 in a mix, it's likely I won't have to EQ the high end at all.

    STILL, I'd like to know if there are any CIRCUIT MODS ONLY (there is someone doing mods without
swapping capsules), that would make the high or low end adjustable a bit.  Certainly the low
end might benefit from higher value series caps C4 and C3:

        http://twin-x.com/groupdiy/displayimage.php?album=465&pos=0

    This schematic is correct, right?  I have the older model, but I'm guessing the
rest of the circuit apart from the pad and HPF are pretty much the same.

        C3 and C4 are shown here as 100nF, or 0.1uF, so surely there is room
for larger caps here.

        Any advice greatly appreciated....

 

[RuudNL]

This schematic is correct, right?

No. The FET doesn't get a supply voltage. There is at least one connection missing.

 

[Paul678]

This schematic is correct, right?

No. The FET doesn't get a supply voltage. There is at least one connection missing.

Do you have a link to a correct schematic?

And can you recommend any good mods for this mic?

 

[EEMO1]

Look at the various MCA SP-1 mods? looks like about the same based on the schematic.

 

[Paul678]

Look at the various MCA SP-1 mods? looks like about the same based on the schematic.

Yes, it's basically a Schoeps transformerless circuit.

C3 and C4 are shown here as 100nF, or 0.1uF, so surely there is room
for larger caps here, to increase the bass response. Also, I assume R8 and C7
make up an RC time constant, for the negative feedback a la the U87 circuit.
They appear to be 47k Ohms, and 1000pF. So the resonant frequency calculates
to: 1/(2*pi*(RC), and this gives us about 3.386 kHz for the lower cut-off.
At parallel resonance, the impedance is high, so I assume for higher
frequencies, it allows the negative feedback.

And I know C20 and C21 are mainly for RF interference, but can these shunt caps
be increased so that they filter some of the high frequencies too?

Thanks for reading, everyone....

 

[RuudNL]

I am sure that there are errors in the circuit diagram.
One would expect a connection between the junction of D1 and D2 and C5, otherwise the FET stage and the DC converter would never get a supply voltage.
But if this is the case, the whole circuit around Q3 is pointless, because there is nothing done with the output of this transistor...
So the only thing I can think of is that there is an other error.
I could think the curved red line might be the correct situation.
But personally I wouldn't trust the schematic at all!
Maybe somebody else could come up with a better circuit diagram?

 

[RuudNL]

In the meantime I found a better schematic:

This version has some (other) small errors:

- Low shelf and -10 dB are reversed on the swith
- the diodes across the 220 uF capacitors are zeners, not 1N4148
- D1 and D2 might be zeners too

 

[Paul678]

In the meantime I found a better schematic:

This version has some (other) small errors:

- Low shelf and -10 dB are reversed on the swith
- the diodes across the 220 uF capacitors are zeners, not 1N4148
- D1 and D2 might be zeners too

Thanks much, Ruud. 

I did verify that the connection you drew from D1/D2 to C5 was correct on my mic.

I also put 0.47uF in parallel with C3 and C4, and it seems to increase the bass response
a little bit.

I also added about 10nF in parallel with C20 and C21, and it DEFINITELY took the
higher frequencies away....too much really.  I knew the caps would be too big, but
I just wanted to see the effect.  I will try 5nF (for a total of 6nF), or maybe less, and will see
if that will be enough.  It's funny, but if you take too much of the harshness away,
you don't get the sibilant detail.....so you have to have a tiny bit of harshness, at least to my
ears.

Adding 1nF parallel across C7 seemed to make the mic distort a bit.  Then I tried
paralleling 47k across R8, and it seems to tame the high end a bit.    Now that you have
found a better schematic, maybe you can enlighten me about R8 and C7?  They
appear to be more important when the "low shelf" is turned on, at least just
looking at the schematic.....

 

[RuudNL]

As far as I can see the effect of R8 / C7 is the following:

The signal at the input (base) of Q3 is taken from the source of the FET.
The transistor inverts the phase, so if the switch is in the right (=low shelf) position, this will introduce negative feedback.
If the R (with the question marks) is indeed 1 G.ohm, the effect can never be too big.
The feedback resistor is 1 G.ohm, but the input impedance at the gate of the FET is in the order of 500 M.ohm. (R17//R18)
The higher the frequency, the lower the reactance of C7.
So the higher the frequency, the lower the feedback...
This would mean: more gain at higher frequencies!

Maybe other microphone specialists have a different opinion? I really would be interested to read about this!

 

[Paul678]

As far as I can see the effect of R8 / C7 is the following:

The signal at the input (base) of Q3 is taken from the source of the FET.
The transistor inverts the phase, so if the switch is in the right (=low shelf) position, this will introduce negative feedback.
If the R (with the question marks) is indeed 1 G.ohm, the effect can never be too big.
The feedback resistor is 1 G.ohm, but the input impedance at the gate of the FET is in the order of 500 M.ohm. (R17//R18)
The higher the frequency, the lower the reactance of C7.
So the higher the frequency, the lower the feedback...
This would mean: more gain at higher frequencies!

Maybe other microphone specialists have a different opinion? I really would be interested to read about this!

At least one person on the Yahoo site thought that C7/R8 had no
effect when the switch is in the normal, or flat position.

However, you can see that even when the switch is in the
"normal", or flat position, R8 and C7 are still
part of the circuit, connected between C21 and C20.

But I do have to admit, upon listening to the recordings, that
changing R8 and C7 didn't do too much (I only recorded the
normal position).

Someone else said: "To control treble you could increase C20 and C21 or you could add a cap from
base of Q1 to base of Q2. Start with 1nF and season to taste, depending on your
capsule's HF peak."

I'm gonna try this right now!

 

[RuudNL]

However, you can see that even when the switch is in the "normal", or flat position, R8 and C7 are still part of the circuit, connected between C21 and C20.

Yes, but the output signal of Q3 is then getting nowhere...
(The junction C20/C21 is 'ground' for AC voltages.)

 

[Paul678]

However, you can see that even when the switch is in the "normal", or flat position, R8 and C7 are still part of the circuit, connected between C21 and C20.

Yes, but the output signal of Q3 is then getting nowhere...
(The junction C20/C21 is 'ground' for AC voltages.)

Yes, you are right.  The AC short to ground should have been
obvious with that 47uF shunt!

Ok, thanks to everyone.  I'm pretty happy with the sound of the mic.

The capsule is like 90% of the sound, but it's still fun to tinker a bit....

 

[Paul678]

Ok, I've listened to the before after recordings,
and to be totally honest, I'm not sure increasing C3 and C4
to 1uF increased the bass response noticeably ( i decided to go all the way!).

It's quite possible the original values, 100nF, already maxed out
the low end, right?

If so, I'll just put the original, factory caps back in, as they are much smaller.

 

[HellfireStudios]

What were you recording? The low frequency increase is usually due to a movement of the cutoff frequency inherent in the design. So only audio containing low frequency energy will benefit from the increased bandwidth.

-James-

 

[Paul678]

What were you recording? The low frequency increase is usually due to a movement of the cutoff frequency inherent in the design. So only audio containing low frequency energy will benefit from the increased bandwidth.

-James-

  Oh, that's a good point.  I was only recording my voice, but I was hoping
it would not just extend the bandwidth, but also bring up the lower mids, but
unfortunately, it sounds pretty much the same.  Perhaps next time, I should
record a bass drum?

However, the cap across the bases of the bipolars has definitely had a
good effect on the excess sibilance, and I have yet to add caps in parallel
to c21 and c20.

 

[RuudNL]

Most of the time the sibilance is caused by a sharp peak around 8 KHz.
Increasing the values of C20 and C21 causes a HF roll-off, but the peak will remain.
Did somebody ever experiment with a series L-C circuit, connected between the drain and the source of the FET?
(Tuned to the peak frequency, maybe with an extra resistor in parallel with the C to 'tune' the Q-factor?)

 

[Gus]

IIRC the R. Schulein microphone article had something about an added RLC network for EQ.

 

[Paul678]

Most of the time the sibilance is caused by a sharp peak around 8 KHz.
Increasing the values of C20 and C21 causes a HF roll-off, but the peak will remain.
Did somebody ever experiment with a series L-C circuit, connected between the drain and the source of the FET?
(Tuned to the peak frequency, maybe with an extra resistor in parallel with the C to 'tune' the Q-factor?)

The L/C ratio is one of the factors that determines the "Q" and so selectivity. For a series resonant circuit with a given resistance, the higher the inductance and the lower the capacitance, the narrower the filter bandwidth.

Frequency of the notch = 1/(2*pi*sqrt(L*C))

So what values should I start with?

 

[RuudNL]

It is the resistance of the inductor and the losses in the capacitor that determine the Q-factor.
That is why I suggested to add a resistor to the circuit.
When you are using good components, the Q-factor will probably be too high. (Notch too small.)
Since you can't change the resistance of the coil, the best way to reduce the Q-factor is adding a resistor in parallel with the capacitor.
But this will also introduces a DC current, so an extra (relatively large) capacitor in series with that parallel resistor will be necessary, to block any DC that might disturb the FET circuit.
And about the frequency: you already gave the formula...

 

[3nity]

From my expérience with sp mics capsule is too bright, your sounds improve over exchanging parts.