"single ic and input trafo preamp"...

[rodabod]

Try the closest value you have, or use a combination in parallel to add up to the desired capacitance.

 

[Harpo]

I would probably try to decrease the 150pF across the 15k feedback resistor to 47pF.
The LPF with 150pF would have -3dB cutoff at ~70kHz, giving phase error of ~15.8° @ 20kHz.
A 47pF would bring this error down to 5° @ 20kHz. YMMV.

 

[oragall]

TNX... i try with a more close value of 800pf and change the 150pf to 47pf or other value more lower than 150pf.

Have any formule to determine the phase error?

and the basspas how db per octave is? eight???

 

[Harpo]

TNX... i try with a more close value of 800pf and change the 150pf to 47pf or other value more lower than 150pf.

Have any formule to determine the phase error?

and the basspas how db per octave is? eight???

Cutoff frequency where amplitude has dropped by -3dB is 1/(2*PI()*R*C), giving 70735.5Hz for 15k and 150pF.
Phase response for this LPF is ARCTAN(testfrequency/cutofffrequency)*180/PI(), giving 15.79° for cutoff 70735.5Hz tested at 20kHz.
At testfrequency 20kHz this has an amplitude response of LOG10(1/(Root(1+Power(testfrequency/cutofffrequency))))*20, giving -0.334dB.

With 47pF this -3dB cutoff goes up to 225kHz, giving response @ 20kHz of 5.06° and amplitude drop -0.034dB.
A NE5534 should work stable up to this range.

Your 2000uF cap in the shunt arm would have a HPF -3dB cutoff @ 1/(2*PI()*R*C), giving 0.53Hz for 150R (=worst case max.gain setting).
Phase response for this HPF is -ARCTAN(cutofffrequency/testfrequency)*180/PI(), giving -1.52° for cutoff 0.53Hz tested at 20Hz.
At testfrequency 20Hz this has an amplitude response of LOG10(1/(Root(1+Power(cutofffrequency/testfrequency))))*20, giving -0.003dB.
A 470uF or 680uF cap is probably big enough for your application. YMMV.

 

[oragall]

wow tnx again ...
if i comprend the values of schematic its good ... only the capacitor of 150pf need change...
but if change the value to 47pf  the LPF inecrease to 225751hrz ... maybe its very much. .. or its good? but if change the 150pf and the resistor increase to 45K... the frecuency down to 75K aprox...
???  how its the correct form to mod this????




the 2000uf its good.. and don understand what you say with the 470uf capacitors...

 

[Harpo]

wow tnx again ...
if i comprend the values of schematic its good ... only the capacitor of 150pf need change...
but if change the value to 47pf  the LPF inecrease to 225751hrz ... maybe its very much. .. or its good? but if change the 150pf and the resistor increase to 45K... the frecuency down to 75K aprox...
???  how its the correct form to mod this????

the 2000uf its good.. and don understand what you say with the 470uf capacitors...

As said YMMV. You neigther will hear 225kHz nor 70kHz, but this would be the LPF cutoff frequency where level would have dropped by -3dB with a 6dB/octave slope. You might be interested, how much this filter affects the audible range as you most often want your preamp behave as linear as it can do and a phase shift by 15° might get more noticed than a level drop by -0.3dB @ 20kHz.
In your schematic you have drawn a 22pF cap for compensation between pin 5 and 8 of your NE5534, as this op-amp is not unity gain stable. Looking at the datasheet of this op-amp, my suggested 225kHz should keep this amp in safe operating areas for your designed voltage gain of 100. If you increase the feedback resistor from 15k to 45k and keeping shunt resistor @ 150R (setting the max.gain), your voltage gain will increase to 300 but your frequency margin drops to about 80kHz. (Your drawn 150pF might have been chosen for another op-amp or a DOA such as 1731 or 2520 to keep it stable.)

If you like your 2000uF and can make it fit to your pcb (probably should read 2200uF as a more common availiable value), keep it. This value might have been chosen for a lower value shunt resistor. Needn't be that big if size matters, even more as you most often don't have dialed in a max.gain setting.

 

[oragall]

if you say ... and the formulas say...
the pre its OK! only change the 150pf to 47pf cap and the fracuency of operation the pre and IC is determine for the input transformer... and maybe for space and dont need a 2000uf because i use a variable resistor of 10K rev log.. I change the 2000uf to 1000uf capacitor not need more.

 

[oragall]

first page with the shematic and pcb actualized with  corrections.

 

[oragall]

woow!! ...  finally end the proyect and de first tests sound amazing. No test of condenser because problems of phantom (still not have the ideal power trafo).

the trafo its edcor pc 600/15k series.

The only problem its the gain, the pre not have good gain, but need replace a 150 ohm resistor to 50 ohm, to modify the gain in 9.5db.

XD

The order of test its:
Sm57 preamp firestudio
Sm57 new preamp
I5 Preamp firestudio
I5 new preamp
and the last 2 demos is the sound of geen pre.


http://www.goear.com/listen/722fe56/preamp-prueba-


photos soon....

 

[Harpo]

The only problem its the gain, ...

Connecting your 150R shunt arm resistor to the opamps inverting input (pin2) instead of output (pin6), as shown in your schematic, will fix it.
(from your pcb, left side leg of this 150R connects to left side leg of the 47pF cap).

 

[oragall]

undersand i test the change.....
but what its de diferencie ????

 

[tv]

On the PCB, you have connected the 150 Ohm resistor to the output pin instead to the inv. input.

IMO you have the power-rails labeled wrongly on your PCB layout, as well.

 

[oragall]

have a new pcb.
the old pcb is small and cant integrate the trafo on the board.

 

[oragall]

pics

what its de diferencie between insert the gain in the right, acording Harpo?

 

[Harpo]

what its de diferencie between insert the gain in the right, acording Harpo?

With 150R connected to the opamps output instead of opamps inverting input its only a voltage follower with gain of 1 (no amplification, just an impedance buffer) and the 150R in series with the 10K rev.log pot loads down the output, so the only gain will come from your transformers stepup ratio. When you connect this shunt arm resistor (in series with the pot and cap) to the opamps inverting input instead, you can vary the voltage gain of the opamp in range between 1+15K/(150R+10K)=2.48 or log10(2.48)*20=+7.9dB with pot wiper in ccw position, and 1+15K/(150R+0K)=101 or log10(101)*20=+40.1dB with pot wiper in full cw position. Add the stepup ratio of your transformer on top giving the total gain of your preamp.
BTW hard to see, but zooming in the pic of your pot connections, you seem to only have connected the outer pot ends (pot wiper doing nothing, only a 10K fixed resistor).
Facing the pot, pins pointing down connect the center pin with the pin on the left side end of the pot. Turning the pot (wired as rheostat) clockwise will reduce the resistance between left+center connection and the right side end of the pot, hence increasing the gain of the opamp.

 

[oragall]

understand... the pot is wire ok.. (I hope you understand)
Change que 150r to 50r and the diference its good. Now have a decent gain.
The other problem its de hummmm ... the circuit have a horrible humm :S .. now i heck the circuit to treat elmilimate or reduce.


14:45pm : the hum is because the trafo and the Power supply, if move the preamp, the hum is reduced, I put ferrites but it was the same, so I took them

maybe the psu can affect, the psu its the same of keith ssl 9k

 

[Harpo]

Using a shielded mic wire between XLR and mic transformer and shielding the mic transformer should help your hum, picked up from your radiating mains transformer, hum frequency probably dominant at mains frequency 50/60Hz. Get your preamp in a shield connected metal box (think faraday cage) and keep the supply/mains transformer external in a separate box (build it safe, this is connected to mains power) should help further.

 

[oragall]

tnx harpo .. now made the second channel and the box. XD

 

[oragall]

pliss !!!! which means the C.T. in the picture???  ??? ??? ???

 

[Harpo]

Center tap. You probably won't use this connection for your build.