I found this mixer at http://www.circuit-finder.com/categ.../mini-stereo-mixer-mic-line-for-pc-audio-card which had several features/considerations that suited my needs. I only missed on the op-amp because I don't have any. I have some 4558, 358 and TL072 and some more of the common type (99% of my parts come from dumpster dives). If I just ignore the chip specified by the schematic, and use some 358 chips instead, how would this affect the operation of this mixer? Is it possible to adjust any component values or make very minor changes that would allow using the above chips?

I am not going to be using the section for mic input (upper left), and instead wish to make 2 of the line inputs (instead of just one as shown in the schematic).

Thanks for reading this post.

 

The 358 is a grand old part, but it is old. Compared to the 272 it has higher noise, higher distortion, and way less gain-bandwidth. It would suck as a mic preamp, but the line input stages are unity-gain so the feedback loop should close ok. C13 and C16 should be increased; I don't think a roll off at 72 kHz gets you anything. 220 pF moves the corner freq to 15 kHz, where it will actually do something useful to high frequency noise.

With a single 5 V rail, the output voltage swing range will be limited. Basically, it is 0 V to around 3.5 V or 4.0 V, depending on the current load. The circuit has the audio signal centered at 2.5 V, but this is not the actual center of the output voltage range. Increase R31 to move the "midpoint" down to somewhere around 1.75-2.0 V.

You can reduce the 358's output stage crossover distortion by adding a resistor from the output to the negative rail (GND in this circuit). This causes the output to run as a Class A stage, but it also increases battery drain and increases the chip's operating temperature.

ak

 

GREAT! Thanks so much for the reply.

C13 & C16 ===> 220pF will do!

Adjust R31 to between 1.75 and 2.0 volts. Yup, I should be able to handle that. If I'm really feeling ambitious, I'm thinking about measuring how high the swing actually reaches. I don't have a scope, but I could use an Arduino Pro Mini to monitor the output voltage, something like: READ voltage, IF voltage exceeds variable increment by small amount, REPEAT

What value resistor should I use? Is the output the chip's output or the line output? (so I could just place the resistor across the RCA jack?) Battery drain won't matter since it will always use a power supply. If the chip fries, no big deal (it was salvaged in the first place)

If I may pose a quick question: think I heard that substitution should be done with the same type of chip, meaning that if the 272 is a FET op amp then use the 358 (also FET) instead of the 4558 (BJT?) Do I have any of this right?

Thanks again for your help!

 

You want to add the Class A bias resistor to the output of every opamp.

In my world, the LM358 is a bipolar dual opamp.

For your circuit, the opamp type should not matter.

ak

 

Sorry about that, I should "fact-check" before I "confusion-display". Thanks for the heads up! I'd like to bury this thread now, but before I deny its existence, could you tell me what the value for the Class A bias resistor is in this case?
Thanks again!

 

If the circuit is driving a normal downstream audio circuit like a speaker amplifier with a 10K input impedance, then something in the 1K - 3.3K range should be fine.

I'd like to bury this thread now, but before I deny its existence,

Why?

ak

 

I've found forum "sticky" posts which basically say:
If all it takes is a quick Google search to get an answer, then posting the question is wasting everyone's time.

So really I shouldn't have asked this:
If I may pose a quick question: think I heard that substitution should be done with the same type of chip, meaning that if the 272 is a FET op amp then use the 358 (also FET) instead of the 4558 (BJT?) Do I have any of this right?

since Googling would have shown that even the premise of the question is invalid. Also, posting stuff like:
I'd like to bury this thread now, but before I deny its existence,
isn't helpful to anyone who might be reading.

I do very much appreciate your help, and I will be sure to include Class A bias resistors - in the 1K - 3.3K range - to the output of every opamp.

Thank you

 

Hello,

Sorry to say, but the LM358 is also a BJT opamp:



Bertus

 

Yes, in post #7 I expressed my regrets for not having Googled before posting. Hopefully I will have learned a lesson here