Selecting corner frequency for IC op amp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Potato Cakes

Well-known member
Joined
Jul 1, 2014
Messages
2,258
Location
Nashville, TN
Hello, everyone,

I am swapping out IC op amps and changing feedback resistor to adjust gain specific to my current project. I would like to know I would go about determining a corner frequency for the new op amp? I can figure out what this was for the original chip, but I don't know if there would be any problem basing this off of the original schematic.

Thanks!

Paul
 
I was kind of figuring as such, but there is one value I'm getting that is a bit perplexing. On a makeup gain amp for a channel fader, the original schematic has a 22k resistor and 10pF, which gives a corner frequency of 723,798Hz. To give me the gain I want, I need to use a 51R resistor, and to keep that same frequency I would need a 4n3 capacitor, which seems quite large for any feedback capacitor that I've seen for line level audio applications.There is also a 10k resistor that connects to both feedback cap and resistor on the inverting input side and goes to ground and I'm not sure how to take that into account. I'm also not sure if I'm way over thinking this.

Thanks!

Paul
 
Instead of 51r 4n3, why not zero Ohms and an infinite cap?

WTF do you want a gain of 1.0051? How is that different from Unity?
 
PRR said:
Instead of 51r 4n3, why not zero Ohms and an infinite cap?

WTF do you want a gain of 1.0051? How is that different from Unity?

I chose 51R because this is where my trimmer ended up when I got exactly unity gain. My other inquiry was dealing with the initial design. In the attached image are the two different op amp layouts that are what I'm dealing with. The one on the left has 22k and 10pF for the feedback and also a 10k resistor that ties to the inverted input and the feedback network and connects to ground. Calculating just the feedback network gives a corner frequency of 723,798Hz, which way higher than I any other network that I've had to do this calculation. However, I know I'm missing a calculation for the 10k that goes to ground. Trying to do reading on my own I found that for this type of situation that I'm supposed to be doing something with the gain (1+R2/R1) in addition to f(0)=1/(2πRC), I just don't know what. I'm sure that the original designer for these circuits did not intend to have a corner frequency so high for a TL072. The circuit on the right calculates more like I am expecting and does not have a R2 component like the one on the left. The corner frequency is 234,170Hz and for the change in resistor to give me the gain I need (37k) I need a 20pF cap (18 to be exact), which makes more sense to me. Both of these op amp circuits are for make up gain for faders used on the different types of channels in the same console.

After all of that rambling, how to do apply (1+R2/R1) in regards to the circuit on the left? And for infinite capacitor, is this removing the cap and leaving that space empty?

Thanks!

Paul
 

Attachments

  • Feedback example.jpg
    Feedback example.jpg
    35.1 KB · Views: 21
PRR said:
WTF do you want a gain of 1.0051? How is that different from Unity?

Also, I hyper obsess over details and my volt meter reads AC down three decimal places. 0.002VAC difference sometimes really bothers me, especially when I know I have the tools and components to make it exact. I know it's not necessary, but most of the time it is difficult for me to ignore even the minuscule differences. Which may make for a fun discussion over in the Brewery section.

Thanks!

Paul
 
Potato Cakes said:
I was kind of figuring as such, but there is one value I'm getting that is a bit perplexing. On a makeup gain amp for a channel fader, the original schematic has a 22k resistor and 10pF, which gives a corner frequency of 723,798Hz. To give me the gain I want, I need to use a 51R resistor, and to keep that same frequency I would need a 4n3 capacitor, which seems quite large for any feedback capacitor that I've seen for line level audio applications.There is also a 10k resistor that connects to both feedback cap and resistor on the inverting input side and goes to ground and I'm not sure how to take that into account. I'm also not sure if I'm way over thinking this.

Thanks!

Paul
You may be overthinking this...

The path bandwidth is the cumulative effect of every HP and LP pole in that path...

When selecting a NF capacitor across a gain stage it is probably more about stability than the frequency response of that one stage. For example if there is stray capacitance at the op amp - input, that can cause lag, that the lead from a feedback capacitor can swamp out and neutralize.

Sorry I do not mean to make this more complicated... in general a pole that is sufficiently above the bandpass of interest will be OK.  That said if there are ten op amp stages in the audio path all in series, and each is -3dB at 200 kHz they will combine to be -1dB at 20kHz.  So a corner frequency for one stage that is -3dB up around 700 kHz is not that crazy.

JR

PS: bus
 
I believe I follow what you're saying and now this part of the design makes sense. There a number of other IC op amps in the signal so cumulatively this one guy won't matter much. And I know I'm usually overthinking most of the custom projects I take on as I just don't know enough to do things simpler and more elegant like I know my work could be.

Thanks!

Paul
 
The fader recovery amp is a fixed gain stage. Gain us the ratio of the resistor to ground and the inverting input to the feedback plus 1, just like a mic pre. The resistor to ground doesn’t change the feedback corner frequency.

The original design has gain in hand at the fader - 10 dB - so you do not want unity gain. Set the fader to its unity gain point and then redo the trimmer exercise.

 

Latest posts

Back
Top