Hi, I am trying to put up an opamp module for my hobby project.
(instrumental amp)
I found a very good paper, explaining about "constant power crossfader":
- http://www.geofex.com/Article_Folders/panner.pdf
Two sources blend providing constant level at all positions.
I am trying to rework it with noninverting opamp configuration.
Even tried to simulate it, but it seems it does not work properly
.
(instrumental amp)
I found a very good paper, explaining about "constant power crossfader":
- http://www.geofex.com/Article_Folders/panner.pdf
Two sources blend providing constant level at all positions.
I am trying to rework it with noninverting opamp configuration.
Even tried to simulate it, but it seems it does not work properly
.
Why? They are the same price.
The circuit is symmetrical. Or should be. Your re-draw is not symmetrical.
I have re-drawn what I think you thought. However I do not think it is constant-power.
Attachments
I simulated the inverting curcuit.
it works, but it seems there is -3db dip in pot's mid position.
Resistors are calculated according to the pdf link above.
Not really 'constant' level blend.
Is there any option for constant level blending with one pot?
As for the inverting opamp, I generally dislike it..
Isn't it more noisy, as if we want a unity gain, we first attenuate the signal and then amplify it?
Isn't it problematic if we want to decrease a bit the input signal?
We cant just use voltage divider on the input.
I also don't understand how we leave the input floating to ground. There is no resistor between input of the circuit and ground.
P.s. For the 'constant' crossfade circuit - what the heck is the input impedance...
it works, but it seems there is -3db dip in pot's mid position.
Resistors are calculated according to the pdf link above.
Not really 'constant' level blend.
Is there any option for constant level blending with one pot?
As for the inverting opamp, I generally dislike it..
Isn't it more noisy, as if we want a unity gain, we first attenuate the signal and then amplify it?
Isn't it problematic if we want to decrease a bit the input signal?
We cant just use voltage divider on the input.
I also don't understand how we leave the input floating to ground. There is no resistor between input of the circuit and ground.
P.s. For the 'constant' crossfade circuit - what the heck is the input impedance...
Last edited:
I apply the simulation of the noninverting circuit. All resistors according to the pdf doc:
- http://www.geofex.com/Article_Folders/panner.pdf
3db dip at the mid position is twice the power, right....
- http://www.geofex.com/Article_Folders/panner.pdf
3db dip at the mid position is twice the power, right....
-3dB plus -3dB is about unity. Or 0.5Watts plus 0.5Watts is 1.0Watts.
It is very possible you do NOT want "constant power". Speech/music signals, and perception, is not simple pure-tone math. For some situations you may want the center to be louder than either side alone. To avoid clipping, two signals must be mixed at -6dB (constant voltage) to be *sure* they won't sum to an overload. Because exact peak coincidence is rare, a -5dB or -4dB mix may be acceptable.
As for your other concerns: Don't be prejudiced. Inverters placed *after* other system gain may not add hiss (hiss should be determined at the system input and the rest of the system run at higher signal level). An inverter does not need a resistor to ground on its input (why would it?). The plan in R.G.'s paper _IS_ very widely used in professional systems. If you are still learning where the resistors go, then just plagiarize, then tweak the resistors a little.
And listen. SPICE won't tell you if a cross-fade/blend is ear-pleasing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.