Guitar preamp op amp tl072 input noise

[Ale]

Good morning,
  I'm trying to build a guitar preamp with a tl072. I'm learning very much trying and looking for schemas everywhere on the net, but I coulden't make any of them work properly. I really don't understand what's the problem with it, so I'm posting you the schema I made. It works, but I still have lot of noise from the input (if I ground the input the noise disappear, so I think it's not the power source). Further more, the guitar is not the problem, because it is very silent one having a faraday cage on the pickups.

Can you help me to solve this problem, please?

Ale

[Andy Watson]

R1 and R2 are very high values for the function they perform - I would reduce them to 100k or less and use the other op-amp as a unity gain buffer to generate GND.

Pin 3, the positive input of your op-amp should have a return path to ground for the bias currents.

Also, where are you getting the 12V from? Some power supplies - especially wall-warts, can couple significant mains interference into your circuit.

[danadak]

C7 will experience polarity reversal with signal, so should be non polar,
or two polars back to back +  - -  + or -  +  +  - connection.


https://electronics.stackexchange.com/questions/161548/does-an-electrolytic-capacitor-degrade-each-time-it-receives-reverse-voltage


Also the G control circuit changes its frequency response with pot (RV1) setting.
Is this what you want ?



Regards, Dana.

[sevenTech]

Hi Ale,
I don't think it is good practice to force vgnd (half of the power supply) to cable shielding...Most common is to connect input filter (resistor and capacitor) to virtual ground so that the signal will be moved ("offseted") to the working area of op-amp (around half of the power supply voltage (6V in your example)) ...Then on the output filter the signal will be moved ("offseted") back to the potential of the cable shield.  As I started to draw, I've also shown unity gain buffer as Andy stated - in case if you are not familiar yet. Those R1 and R2 shall be reduced as noted before. If you will use this buffer, they can be relatively high thus reducing power consumption. Without unity gain buffer it should work too, but R1 and R2 should be waay lower values then. But often if you use dual op-amp package, you will just use one of them for such purposes 

BTW: Unused op amps in the package shall be properly terminated - otherwise it can add some noise.

[Zero999]

Good morning,
  I'm trying to build a guitar preamp with a tl072. I'm learning very much trying and looking for schemas everywhere on the net, but I coulden't make any of them work properly. I really don't understand what's the problem with it, so I'm posting you the schema I made. It works, but I still have lot of noise from the input (if I ground the input the noise disappear, so I think it's not the power source). Further more, the guitar is not the problem, because it is very silent one having a faraday cage on the pickups.

Can you help me to solve this problem, please?

Ale

There's no decent path for the input bias currents. It's only working because the TL072 has tiny bias currents, so the leakage through C5 is making it work, (if you replace the TL072 with the NE5532, it won't work) but it will be very noisy. Move R6 to the other side of C5, so it connects directly to pin 3.

[TheUnnamedNewbie]

R6 of 10k seems a bit on the low side, I believe most guitar amplifiers have input impedances on the order of 1 M?. Ofcourse, need to make sure you can still bias that opamp.

[Ale]

Wow, thank you guys. So the schema should be something like this right? I didn't know that I should use the op amp to get a proper VGND.
Anyway, the power source is a instrument power supply which should be very stable. It has many 9V, a 12V and a 18V output.
Dana, I would like to change the gain through RV1, is that correct?

[Zero999]

How did you choose the value for C6?

It will have an impedance of 100k, at frequencies above 15.9Hz, which will be the upper cut-off frequency, when RV1 is set to the maximum value of 100k.

[jm_araujo]

It seems to me that the jack connections are reversed in the schematic. Tip is signal and sleeve is ground.

[Ale]

Yes, jack connections were upside-down, I corrected them.
Now the powering part is working fine, I can hear a little noise when jack input is disconnected, but when I connect it all shout down and I hear nothing.
I follow the instruction of SevenTech, I used gnd of power supply for the shield, is that correct?

Thanks again 

[jm_araujo]

You should look at the cut off frequencies of the RC filtering, as Hero999 said.

At the output R4/C7 have a cutoff of 159Hz which lands you right in the middle of the guitar scale. You should increase C7 at least a magnitude.

[IanMacdonald]

I presume you mean hiss. Guitar inputs are prone to hiss because they are relatively sensitive and high impedance. (a few mV at 47k or so) Under these conditions surprisingly it's the resistors that almost always generate the hiss, not the transistors. Using good quality metal film types, especially for high values in the signal or feedback path, will make a huge difference. 

[Zero999]

You should look at the cut off frequencies of the RC filtering, as Hero999 said.

At the output R4/C7 have a cutoff of 159Hz which lands you right in the middle of the guitar scale. You should increase C7 at least a magnitude.

You've got the decimal place wrong. The cut-off is 15.9Hz, so it's no surprise it's not working. Reduce the capacitor to 100pF.

[jm_araujo]

2nd line of my text was referring to the output high pass (DC blocking) filter, and the math was right for it (.1u C7 with 10k R4)

[Zero999]

2nd line of my text was referring to the output high pass (DC blocking) filter, and the math was right for it (.1u C7 with 10k R4)

Oh I see, yes, no wonder hardly any signal is getting through. I'd remove R3 and make C7 100µF.

[Ale]

Something like this? I will try it soon fisically to test it.

[jmelson]

You are feeding +/- 12 V directly to the pins you mark as +6 / -6 V.  Did you mean to have a resistor there?  How about a Zener to drop the 12 V to 6 V?

You have no DC path from the non-inverting input of the op-amp.  This is likely the cause of most of your problem.  You say "noise" but I'm guessing you really mean distortion, as the output may be floating at least several Volts away from ground with the floating input.  Connect a relatively high value resistor to ground from pin 3.

Also, the TL072 is not known for low noise.  You might try my favorite, the OP275, direct drop-in pin for pin compatible.

Jon

[sevenTech]

2nd line of my text was referring to the output high pass (DC blocking) filter, and the math was right for it (.1u C7 with 10k R4)

Oh I see, yes, no wonder hardly any signal is getting through. I'd remove R3 and make C7 100µF.

R3 must be present, if just removed it will be unity gain without possibility to change gain. If shorting to virtual ground it will saturate. (Vout = (1+ (R2/R1)) * Vin)
As to that output filter, I will just increase R value to 100K. C can be 0.1uF... (fc = 1/(2*pi*R*C) gives 15.9Hz). Non-polarised 100uF caps are pretty bulky and pricey.

[sevenTech]

Yes, jack connections were upside-down, I corrected them.
Now the powering part is working fine, I can hear a little noise when jack input is disconnected, but when I connect it all shout down and I hear nothing.
I follow the instruction of SevenTech, I used gnd of power supply for the shield, is that correct?

Thanks again 

I think that this circuit should work...Just fix the output filter. I don't know why you hear nothing...Recheck wiring if there isn't some problem :-) And remove C6 for simplicity. You dont need that for now.

[alexanderbrevig]

jmelson +6V is only 12V when measured relative to the 0V and not the vitrtual  ground. No issues here:)

[jmelson]

jmelson +6V is only 12V when measured relative to the 0V and not the vitrtual  ground. No issues here:)

Oh, I see what he's doing, using a single +12 supply to make + and - 6 V from a synthetic analog ground!  Yes, now I got it.

Jon

[Audioguru]

The extremely low value of 10k for the original input resistor loaded down the signal (especially high frequencies). Now that the input resistor is a normal 1M the output from the magnetic pickup will be much higher (especially high frequencies) and then the gain will be much too high. Change R3 to 10k.
C6 muffles all high frequencies. Remove it.

[Zero999]

2nd line of my text was referring to the output high pass (DC blocking) filter, and the math was right for it (.1u C7 with 10k R4)

Oh I see, yes, no wonder hardly any signal is getting through. I'd remove R3 and make C7 100µF.

R3 must be present, if just removed it will be unity gain without possibility to change gain. If shorting to virtual ground it will saturate. (Vout = (1+ (R2/R1)) * Vin)
As to that output filter, I will just increase R value to 100K. C can be 0.1uF... (fc = 1/(2*pi*R*C) gives 15.9Hz). Non-polarised 100uF caps are pretty bulky and pricey.

Sorry I meant R4, not R3. The problem, whatever load the amplifier is connected to, goes in parallel with R4, so there's no point in having it there. The AC coupling capacitor just needs to be as big as possible.  Don't use non-polarised capacitors. Use two 220µF aluminium polarised capacitors connected back-to-back.

There's still the issue of C6 cutting high frequencies.

[Ale]

I modified the schema, is now correct?

[danadak]

RV1 at one extreme will short the output of the amp to the output of the
ref buffer. TL072 is protected from output shorts, but thats a bad idea
to rely on that in your schematic.

Also I assume RV1 is for G adjust. Therefore remove the Vgnd connection
from - input, connect the unconnected end RV1 to a R to ground. The R will
set max gain, prevent amp from going open loop.

Regards, Dana.