OpAmp demystification

[@xi@g@me]

Hello all! I'm new to op amps, and trying to fix a PCB for a friend. It's an audio problem, there is no apparent sound on both channels. I'm following the schematics to find the probable point of failure, but I'm not able to understand correctly this part of the circuit (and mainly, what I need to expect on both sides)
Here is the part of the schematic that I'm stuck on (full schematic here):


(Note there is a series capacitor with a parallel resistor on each channel before mixing, leading to think this is a passive high pass filter)
Top opAmp is for the right channel, bottom for the left.
I can't even tell if the output is to be inverted or not, but what I see that each channel has what looks like a voltage divider circuit (R99, 101 and 102 on right channel, R103, 104 and 105 on right channel, each 22k). I also see that there is a parallel capacitor that may be used as an active low pass filter.
For the feedback circuit, I'm not used to see the output connected to both the inverting and non inverting input (though there is a series capacitor on the positive feedback loop - is it the stabilize the output?).

Anyway, if you could help me understand this part of the circuit better, I'd be grateful

 

[bertus]

Hello,

Have a look at page 29 of the attached PDF.
It will show you the following circuit wit explanation:

Your circuit has an extra RC on the input.

Bertus

 

[@xi@g@me]

OK, so this is indeed an active low pass filter, with no gain (and no inversion). Thanks for the help This handbook may become handy, even though I'm not familiar with the s domain yet.
The other RC circuit is a passive low pass filter for a different frequency?

 

[bertus]

Hello,

Here is an other PDF on opamp circuits.

Bertus

 

[AnalogKid]

For multi-pole active filters, the corner frequency of each pole usually is different. You can see that in your component values. The circuit in #2 and your circuit are different. Your circuit is a true three-pole filter, not a two-pole filter with an extra R-C tacked onto the front.

Side comment: The LM324 (and its dual brother, the LM358) has notoriously poor audio performance. The input stage is noisy, the voltage amplifier stage is slow, and the output stage has relatively high crossover distortion. All three of these issues are consequences of its exceptionally low (for its time) operating current. Life is choice.

OTOH it is cheap and nearly indestructible. Great for an intercom, modem, or phone application, but not so much for digital audio. What is the device you are working on? Make/model/photo - ?

ak

 

[@xi@g@me]

This is a SNES, PAL version. There is another pair of op amps right after these ones, with a gain of 2.4 (I think), so I guess this would increase the SNR a bit. I never played a SNES game on something else than a TV set, so I can't tell for the sound quality

 

[@xi@g@me]

The way this is handled is quite simple. The sound goes out from the DSP with only one channel (and a clock), and is fed into a codec chip (stereo 16 bit DAC). It then goes through the active low pass filter (+ I guess passive high pass), is mixed with cartridge audio and external audio (from the port under the console), and has a mute system (for DSP audio only). The resulting audio signal is then fed into the AV multiout connector.
I tried to see what was on the AV multiout connector using the oscilloscope and I get constant ~4V on both channels (the mute system works, then). As the signal is the same on both channels, I suspect the DSP to output of opAmp chip section. Most probably one of the chips, otherwise one of the channel would work differently.

 

[@xi@g@me]

I'm late here, but just for a not I was able to fix the problem. The ref capacitor for the codec chip was bad... My colleague told me he recaped the whold board... but actually he forgot only one cap and of course this was the faulty one.
Thanks for your help and for the resources