Op Amp interference between CH1 & CH2 when used for different tasks?

[Ted Krotkiewski]

Hi Friends!
Apologies for troubling you with such a basic question. I am sure it has been answered many times before, but when searching I couldn't find it.

QUESTION 1:
I am experimenting with some Passive inductor based and active EQ circuits where there is a few Op amp unity gain buffers and Op amp gain buffers etc.
So, I am a little puzzled wether I can use ONE Dual channel Op-amp for two completely different tasks that occurs in completely different places in the circuit.

- For example, with a OPA2134 (dual channel), Can I use CH.1 for Input unity Gain buffer in the very first stage of the circuit and then CH.2 as an output gain buffer in the last stage of the circuit? - Will there be "Crosstalk" or other sort of interference that we want to avoid when they take place at such different tasks?
Or would I be better in using two single channel OPA134 IC´s instead?

QUESTION 2:
Is a TL072 or a OPA2134 just two completely independent TL071´s or OPA134´s channels in one package to save money and space? or is it more to it that I am missing?

For example, I see many people use Dual channel IC´s, when there is only one channel being used (No extra function pins being used either) So I wonder wether they might differ performance wise between dual channel or single ch?

The IC´s I will be using are NE5534/32 and OPA134 / OPA2134

Thank you for your time!
//Ted Krotkiewski

 

[abbey road d enfer]

So, I am a little puzzled wether I can use ONE Dual channel Op-amp for two completely different tasks that occurs in completely different places in the circuit.

X-talk between sections of multiple opamps is a weel documented parameter. For a TL072, it's 100dB.
This is for both sections operating in their linear region (not clipping, not slew-rate limiting), so I would recommand some caution in this respect.

  with a OPA2134 (dual channel), Can I use CH.1 for Input unity Gain buffer in the very first stage of the circuit and then CH.2 as an output gain buffer in the last stage of the circuit?

Yes. I wouldn't recommand using one half for a signal and the other half for another signal, though, or if one of the stages had very high gain (>40dB). The hearing process can detect x-talk at extremely low levels.

-Will there be "Crosstalk" or other sort of interference that we want to avoid when they take place at such different tasks?

As long as you make sure to obey the preceding recommandation, you're good.
Not so "different tasks"; they carry two versions of the same signal.

Is a TL072 or a OPA2134 just two completely independent TL071´s or OPA134´s channels in one package to save money and space? or is it more to it that I am missing?

Apart from the power connections, they are supposed to be 100% independant, but they are on the same dye, which creates thermal interactions.

For example, I see many people use Dual channel IC´s, when there is only one channel being used (No extra function pins being used either) So I wonder wether they might differ performance wise between dual channel or single ch?

The only reason is that by buying more of the dual they get a better price and their inventory is more manageable.

The IC´s I will be using are NE5534/32 and OPA134 / OPA2134 

Beware that a 5532 is not two 5534's. The 34 has better noise specs, and is uncompensated, which allows adjusting performance for high-gain applications.
This is kind of an exception, most of the duals/quads are identical in performance to the single.

 

[scott2000]

Beware that a 5532 is not two 5534's. The 34 has better noise specs, and is uncompensated, which allows adjusting performance for high-gain applications.
This is kind of an exception, most of the duals/quads are identical in performance to the single.

Is the 5533 considered a dual 5534?

 

[abbey road d enfer]

Is the 5533 considered a dual 5534?

It is supposed to be. However, I have never seen one. I see some on E-bay, they are quite expensive compared to a pair of 34's.

 

[scott2000]

It is supposed to be. However, I have never seen one. I see some on E-bay, they are quite expensive compared to a pair of 34's.

ATI used them  in some of their older distribution amps.... I think the DA 2016 iirc....

 

[JohnRoberts]

As Abbey already shared, crosstalk between amplifiers in dual or quad packages is well described in data sheets. 100 dB sounds like a typical spec. Be aware these are typically specified wrt the input. So noise gain (closed loop gain) of each op amp will degrade that separation spec. 

It is generally good practice to not put wildly different signals in two halves of one dual. Why tempt fate?

JR

 

[Ted Krotkiewski]

Thank you @Abbey and everyone else for answering my question! 

Beware that a 5532 is not two 5534's. The 34 has better noise specs, and is uncompensated, which allows adjusting performance for high-gain applications.
This is kind of an exception, most of the duals/quads are identical in performance to the single.

This was new to me. I am very happy you pointed this out, as I would propably have scratched my head for days not knowing this.

Ok.. While on the Opamp Topic, I have one more question that I would like to ask:

Lets plot my circuit has a few Unity Gain Buffers and some Gain buffers. Lets also plot I only have 2520 DOA´s and TL072 opamps to use.
In terms of "affecting the sound" are they equally altering the sound as audio passes through them, no matter their function, or is it so that the opamps doing Gain amplification will affect the sound more than the others only doing unity gain?

The reason I ask is, Since it will be quite expensive having 2520´s on everything, It makes more sense in my head having 2520´s only on the Gain amp duties, and TL072´s on the unity gain buffers.

-Or will it be useless implementing a 2520 DOA in my circuit if there are other "Weak links" (TL072´s)

Again. sorry for my neewb questions...

Thank you for your time!
//Ted

 

[abbey road d enfer]

Thank you @Abbey and everyone else for answering my question! 
This was new to me. I am very happy you pointed this out, as I would propably have scratched my head for days not knowing this.

Ok.. While on the Opamp Topic, I have one more question that I would like to ask:

Lets plot my circuit has a few Unity Gain Buffers and some Gain buffers. Lets also plot I only have 2520 DOA´s and TL072 opamps to use.
In terms of "affecting the sound" are they equally altering the sound as audio passes through them, no matter their function, or is it so that the opamps doing Gain amplification will affect the sound more than the others only doing unity gain?

The reason I ask is, Since it will be quite expensive having 2520´s on everything, It makes more sense in my head having 2520´s only on the Gain amp duties, and TL072´s on the unity gain buffers.

-Or will it be useless implementing a 2520 DOA in my circuit if there are other "Weak links" (TL072´s)

Again. sorry for my neewb questions...

Thank you for your time!
//Ted

IMO, a 2520 is justified when :

• low-noise performance is needed (beware that the noise perf of the 2520 is similar to that of a 5532, i.e. much better than a TL0xx)
• High drive capability, e.g. for driving transformers, particularly the 1:2 or 1:3 used in API, or a large number of aux pots

TL072 can be considered a weak link compared to a 2520, but also compared to a 5532.
For applications that don't need heavy weightlifting, a 5532 is as good as a 2520.
If low noise is necessary, a5534 (single) is about 3dB quieter than a 5532, but there are newer opamps that have better noise performance, such as LM4562.