simple opamp boost (aka noob moment)

[cloudscapes]

I think I'm losing my mind. I'm building a small active utility mixer and I've got the passive input/panning circuit working just fine, but when it comes to actually pushing the signal through opamps and hopefully getting a small boost, apparently I'm incapable of doing that. This is one of those "okay, 10th try, now lets try..... it doesnt work. of course it doesnt work! how silly of me to think something I'd do would work! especially something as simple as this!" 

I've tried these three:
http://www.generalguitargadgets.com/diagrams/mixer_sc.gif
http://experimentalistsanonymous.com/diy/Schematics/Buffers%20Switchers%20Mixers%20and%20Routers/Simple%20Mixer.pdf (mostly the same)
http://experimentalistsanonymous.com/diy/Schematics/Buffers%20Switchers%20Mixers%20and%20Routers/Simple%20Mixers.gif (bottom-left)

they all have something in common. when no power is applied, I get a faint output signal. when power is applied, absolutelly nothing. probing around, I find absolutelly nothing comes out of the very first opamp output . the opamps are fine. I tried 4558s and tl072s both from fresh batches and pulled out of working projects. I'm using a good battery. I've redone the whole thing multiple times on breadboard (easy because it takes about 4 minutes, so simple), minus the part that actually does work, panning/fading. caps are good and oriented properly. I also checked the holes I was using on the breadboard just to be sure (at that point I was desperate), though I had just cleaned up an earlier project on it that I know was working.

pin voltages for second schematic example
1 7.94
2 2.33
3 4.10
4 0
5 4.10
6 6.12
7 1.3
8 8.61

I should be getting something from pin 1 on all of these schems, correct?

[PerroGrande]

Well, it isn't surprising that you've got one thing in common between all those schematics...  They're all inverting configurations that have a connection from input to output that doesn't *need* the active component.  Some sound "leaking" through a config like this is normal.

Troubleshooting -- step 1...

Using RG's simple mixer -- the first schematic you linked...

Step 1.   Remove R10 and re-check your voltages on the IC.  This will isolate the two stages of the amp.

From the voltages you posted, the "first" op amp (pins 1,2, and 3) is pegged at its "positive" rail limit.  Not too surprising, since pins 2 and 3 are nearly 2 volts apart.  An op amp will work very hard to make the two inputs show equal voltage. (Likewise for 5 and 6)

If after removing R10 the voltages on 2 & 3 are dissimilar, check your wiring, pin orientation, etc.

If after removing R10, the voltages on 2&3 are nearly the same, use an audio probe to check for sound on IC pin 1... (Or just take pin 1 as the mixer's output for testing).

If all is good from pin 1, try adding a 10uF capacitor in series BEFORE R10, and then re-install R10.  This will AC-couple the two stages.

Re-check voltages and audio.

[brett]

Hi
the pin voltages for 1,2 and 6 and 7 should also be at half the supply voltage.  I notice also that the bias voltage (4.1V) isn't half of the supply voltage (8.61V).  So there are significant problems.
You don't happen to have the LED connected to pin 7 instead of the supply, do you?
The voltage on pin 7 is very low, and I'm guessing it is close to the problem.
cheers

[cloudscapes]

thanks for the help guys

the bias voltage was an error on my part. sorry about that. it's 4.12. and I haven't connected the LED. figured I'd keep things simple untill they work. I've redone RG's schematic (the voltages I posted were for the second schem). it still doesnt work, but here are the troubleshooting results:

I've removed R10. pin voltages are

1 - 7.96
2 - 0.85
3 - 4.11

4 - 0

5 - 4.10
6 - 4.30
7 - 4.30

8 - 8.62

I hear the output from pin 2 (just checking my audio probe works) but hardly anything from pin 1. its incredibly faint. after a good night's sleep and in bright light, I can confidently say that the pin orientation and wiring is correct. I've also removed the panning/fading part of the circuit to avoid confusion and now just feed audio straight to pin 2. I tried a fresh 4558 and a tl072 pulled out of a working project.

[PerroGrande]

Yeah -- you've got some wiring/connection problem of some sort. 

Pin 3 should be half of the battery voltage (4.3).

Whatever the wiring problem is, it is in the first half of the circuit --- i.e. the stuff connected to op amp pins 1, 2, and 3.

[brett]

Hi
have another look at the input to pin 2 and check that ther's not a DC connection to earth.  (e.g. one of the input capacitors might be faulty).
A low voltage on pin 2 appears to be making the output of the first op-amp "latch up" to nearly the power supply voltage.
cheers

[cloudscapes]

again had a look at the connections. there's really nothing to goof up with! there's hardly any wiring to speak of. one resistor between pins 1 and 2, and an extra wire on 3. just to be sure, I redid the circuit on another part of the breadboard. same results, both for audio and pin voltages. I've replaced all aprts, caps, jumper wires, resistors...

then I wondered what the pin voltages would be if I unplugged the input (a potential source of current).I did that and found pins 1 and 2 are 4.30. I'm truly stumped. I cant even begin to understand why!

for both output and input, I use my soundcard's breakout box and set both the outputs and inputs to be at -10db. I've used this system for testing lots of things and it has always worked! in a way I refuse to believe that the card's the source of the problem because if the card works independantly and with other effects I plug it into, the for all intent and purposes the mixer design itself is faulty and incompatible with real-world setups.

but I know that's not true. its just the frustration speaking.

for kicks, I used my guitar as input instead. same results. as soon as I have an input, even a passive input like a guitar, pin 1 jumps to 7.9 and 2 down to 0.8, both from 4.3

[flo]

Please measure the resistance between the tip on the input jack and pin2 of the opamp without having anything jacked-in. It should be very high (like unmeasurable high).

The resistor between pin 1 and pin 2 can be temporarily jumpered (bridged) for testing. The first opamp now has a gain of 1 so its more a buffer than a booster. The DC voltage of pin 1 should off course be equal to the DC voltage of pin 2. Both pin 1 and pin 2 should be equal to the Vbias (~4.5V) that is also on pin 3.

[PerroGrande]

I think Brett's point may be spot on.

In an inverting configuration, the point where the feedback resistor meets the input resistor is at virtual ground -- driven there by the op-amp's desire to make the two inputs the same.  On a traditional bi-polar power supply configuration, the non-inverting pin (#3 in this case) is grounded.

However, in this case, the inverting input pin is tied to 4.5 volts -- half the supply.   In our circuit, the "virtual ground" is actually at half the supply voltage.  This isn't a problem unless the "virtual ground" and the "real ground" get confused with one another, so to speak.  If this were to happen, the op amp would latch to the positive rail in an attempt to make the two inputs the same -- which is what you're experiencing.

One thing to try:

On the input potentiometer (R1 on RG's diagram), instead of connecting the third lug of the potentiometer to ground directly, connect it to ground through a 10uF cap (- side toward ground of course) and see if that changes anything.

[cloudscapes]

flo, there's no input jack. I'm feeding input straight into pin 2 with alligator clips.  there is input as I can touch pin 2 my audio probe and I hear the audio clearly. on pin 1, however, nothing.

PerroGrande, there's no potentiometer. I already assessed that the "mixer" parts of the circuit worked flawlessly passivley and removed that part to try and debug the actual buffer/booster opamp. I'm feeding audio directly into pin 2 to keep the thing as simpel as possible and try to debug. do you want me to recconect a pot?

update: okay, I recconected C1 and R5 and am feeding the input into C1 instead. it's extremelly overdriven but now it works. I dont understand why it didnt work before when I had the full circuit up, and I understand even less why it seems to need a 0.1uf cap between the input and pin 1. to me it makes about as much sense as why hooking a hose to a tap wont spray water out the end unless I actually bend the hose a little in the middle (seriously, thats how I'm seeing it). I was feeding input straight into pin 1 because in my mind that's "the path of least resistance" (even if its not a resistor), so the best possible condition. obviously there's still a whole lot I dont understand..

I'll rebuild the rest of the circuit (different fader pots and all) and report back.

I am a moron.

[PerroGrande]

You *have* to have C1 in the circuit.

This eliminates any DC offset that may exist on the incoming signal AND isolates the input from "real" ground.

Oh -- and Pin 1 is the OUTPUT pin from your op amp. 

Pins 2 and 3 are inputs.  One is an "inverting" input, the other a "non-inverting" input.  R5 and R9 on RG's circuit are involved in feedback, which controls the gain of the op-amp (which would otherwise be insanely high, as well as adding other beneficial behaviors to the circuit. You need both to be in place for the circuit to work.

[cloudscapes]

thanks for the explanation about C1. and about opamps which I'm not very familiar with yet (usually a copy parts of a circuit and it just works)
and yeah I knew pin 1 was output. I was reporting that there was nothing comign out of it, but of course at the time I didn't have C1.

its a bit clearer now

[flo]

http://en.wikipedia.org/wiki/Operational_amplifier_applications#Inverting_amplifier
You also need a resistor between C1 and pin2 (Rin). Together with the resistance between pin1 to pin2, it determines the gain of the "inverting amplifier" (Rf).

[brett]

Hi
we all learn a lot as we go along.  Studies show that trial and error is one of the best ways for adults to learn (adults are usually too impatient to read manuals, textbooks, etc.  We want action and results!)

So, you now know about inverting op-amps, which atre the backbone of lots of circuits.  That's great.  While you're at it, you might as well figure out how non-inverters work as well.  Many principles are common to both types of circuit.

OT: For what it's worth, at the end of each day, I think about 2 things: how many things I've learned and how many people I've been kind to.
have an excellent day!