Mysterious Resistor?

[Shahriar]

Hi
I have made a preAmp From a datasheet from Texas Instrument that the PDF is Attached, But i Think somthing is missing because when my hand touched the
Capacitor (Cx) accidently then i found that i have to Solder a big value resistor between the
output of the Capacitor and SGND (Rx), but there is a little bit Hiss, at first why it is needed? If I don't put the Res The circuit will be mute unless i shout in Mic...secondly I have done many tests and a resistor less than 820K will mute the Circiut as it was at first
how can i calculate the Best value for Cx and Rx to have the less Hiss, the Mid voltage in schematic is 6V to to bias the AC signal to the center of the
operating voltage range to prevent positive or negative clipping
of the audio signals.

your help is appreciated in advance...

View attachment 35999

 

[Shahriar]

Here is the Original PDF that i substitute TLC72274 with NE5532
and i didn't understan the resistor R24(20K)

Texas_Mic_Mixer.pdf

 

[audioguru2]

Hi Shahriar,
The NE5532 is different from most opamps because it has protection diodes, back-to-back, across its inputs. Opamps that don't have those diodes don't need Rx. Of course, with Rx at a certain lower value or less, the opamp's output is forced to the positive supply, and therefore mutes. The high value of Rx and the current through it causes hiss (resistor thermal noise). I would use low-noise TL072 opamps or other low-noise opamps without input diodes instead of NE5532s and not require Rx.

TI's mic mixer circuit can be used with dynamic, 2-wire electret or 3-wire electret mics.
Without jumpers on JP1 and JP2 as shown, it is set for a dynamic mic that doesn't need a bias voltage. The dynamic mic connects to TIP and ground.
A 2-wire electret mic has an output FET as a common-source configuration. Its bias voltage of 0.66 X Vdd is from U1a through R7 which is the drain resistor. The mic's FET's drain connects to the TIP. JP1's middle pin is jumpered to the TIP so that R7 is connected to the drain. The FET's source is connected to ground.
A 3-wire electret mic has an output FET as a common-drain source-follower configuration. Again its bias voltage of 0.66 X Vdd is from U1a through R7 which is the drain voltage supply (not used as a resistor, just as a voltage source). The mic's FET's drain connects to the RING. JP1's middle pin is jumpered to the RING to connect to the drain. The FET's source is connected to the TIP. JP2 is jumpered to provide R24 as the FET's source resistor.

 

[GPG1]

The NE5532 is damaged. Replace it.

 

[MP1]

Shahriar, ignore the comment posted about not needing Cx. All op-amp circuits should have this capacitor. This capacitor will block the DC component, but also serves as a filter for the input frequency. The correct combination will help eliminate frequencies you do not want to amplify. Let me know if you need the calculation for this.

Isn't it great when you ask a question about a particular component and someone tells you to use something else? Personally, I think if someone can not help with the question answered, they should keep quiet and allow someone with the answer to post it, rather than adding more confusion and a complete re-design.


MP

 

[audioguru2]

MP,
All opamp circuits do not need an input coupling capacitor. Try amplifying a small DC voltage through a coupling cap. It won't work!

Why are you talking about Cx? Of course it is needed.
Shahriar was asking why he had to modify the circuit by adding Rx, to adjust the messed-up offset voltage of the opamp that he probably zapped with a static spark. The opamp's input diodes should have protected against damage, that is what they are for, unless they also got damaged.
Where is the "complete re-design"?

 

[audioguru2]

Hi Shariar,
Hey, isn't this the same IC that got hot in your other post?
The one that was trying to operate without a negative supply?
Then was when its input protection diode got fried.
Injecting some mains hum with your finger was probably too much voltage for the input without a protection diode, therefore messed-up its offset voltage.

 

[Shahriar]

Dear Audioguru & MP

dear MP, There are so many guys in this forum with lots of information that want to share with others like you, Audioguru, Hotwaterwizard, ... and i really appreciate all of your correct or wrong comments...

I replaced NE5532 with TL074 and because they are pin compatible & i had used IC socket it was a piece of cake...
But what happened? Circuit did NOT work without 1Mohm Resistor, I checked the DC voltage of Output: Vout=6.04v without 1M res, Vout=6.32v With 1M res As expected...

I found a schematic (that is attached), if you take a look you see R4=1Mohm, The author has said this: "Resistor R4 provides a ground reference for the pre-amp" not anymore comment, I sent an email to him and waiting for reply.

View attachment 35559

 

[MP1]

Shahriar, ignore the comment posted about not needing Cx.

Good, the person who posted went back and corrected. The post had Cx earlier. I guess this was supposed to be Rx instead.
BTW- when refering to all op-amp circuits, I had in mind audio circuits, since this was the subject matter. I did not make myself clear and left out the word audio. It is obvious that if you are amplifying a DC signal you cannot put it through a capacitor. I do not think anyone misunderstood this since I mentioned blocking DC component, but here is the confirmation in case it is needed.

MP

 

[audioguru2]

Hi Shahriar,
Of course R4 and R5 are needed in your new circuit to provide a ground (mid) reference voltage for the non-inverting opamps. Without them, the output voltage of the opamps could be anywhere.

Your other original circuit had an inverting opamp that was already Mid (ground) referenced since its non-inverting input was connected directly to Mid.

All these opamps are simply DC voltage-followers of the Mid voltage.

Sorry, but you are confusing:
1) Wasn't your replacement a TL072 (Not TL074)?
2) Which circuit didn't work without a 1M resistor? (The original one?)
3) Where did you connect the 1M resistor? (From input to ground? Not to Mid?)
4) Which output voltage changed just a little when you added the 1M? (The output voltage of the same opamp?)
5) Since the output voltage was correct without the 1M resistor, why did you add it and why did it make the circuit work? The original circuit should work fine without it.

 

[Shahriar]

Hi

here the answer of your questions...

1- Sorry, TL072 is correct, I misspell it.
2- Yes, in the Original circuit i replace NE5532 with TL072, I have not build this

second circuit i just post it to see maybe The 1Mohm(R4) resistor do the same as

the 1Mohm I put in the original circuit.
3- In the Original circuit, I replaced Rx with 1M resistor. (one pin between Cx an

10k Resistor and the other pin to the SGND)
4- In the original circuit, the Output voltage of 'U1A' changes a little, I wrote this

to tell you that the OpAmp does NOT saturate with or without Rx (1Mohm), in fact

Rx(1M) resistor does not play any roll in DC Voltage of output.
5- This is the main question!!! This is why i call it Mysterious Resistor.

dear audioguru, can you explain some more about the Reference voltage in the

second circuit i posted, why output will go anywhere without it?

Shahriar

 

[audioguru2]

Hi Shahriar,
I don't know why your original circuit doesn't work without Rx as a 1M resistor. The circuit is biased properly (with the non-inverting input connected to Mid) and its output voltage is correct. The opamp should work fine without Rx. As you measured, Rx doesn't seem to do anything (it just pulls-down the input with only 6uA, which causes the output to pull-up only 6uA). You measured the output voltage to increase 280mV, which is exactly 6uA through the 47K feedback resistor.

In your 2nd circuit, the opamp is operating as a non-inverting AC amp and as a DC voltage follower. In a DC voltage follower, there is 100% negative feedback for DC so the output "follows" the input. The Reference Voltage of 0V is given to the non-inverting input by R4. Without R4, the DC voltage of the input could be any voltage and the output voltage will be the same. The 1st circuit is similar, because the Reference Voltage of Mid is connected to the non-inverting input. The DC output voltage will be the same.

 

[Shahriar]

*** Thank you very much for your always complete replies ***