Why does this inductor break this circuit ?

[MathWizard]

I have a 3-stage BJT audio amp for an electret mic, for inputting into an AM TX. And I'm trying to get the bandpass filters between stages to work better for audio. They are LC and RC , set up for 60Hz to 15-20kHz. Some of the circuit is on a protoboard, and the filters and emitter bypass caps are on a breadboard.

And I have a lot of FM radio everywhere at every node, around -63 to -55dBV most of the time. In open air my probe gets around -75dBV while the GND is still on the circuit GND. Without the GND clipped on it's back around -65 to -55dBV. (but my AWG and other scope probe were still GNDed on that circuit)

1 thing I just tried was putting a 1uH inductor in series from the PSU to my circuit Vcc point. It only has 0.2 ohm resistance, and I still have my 5V DC. The total circuit current, when idling, with the mic disconnected, and no oscillations under 100MHz anyways, the idle current is 14mA, which is about what it should be with the LED on the board.

But with the 1uH inductor from the PSU, my idle current drops to 8mA, and the audio won't work either, it's gone from upto 3-4Vpp, to nothing, just the -55to-65 dBV noise floor.

[fourfathom]

Can't say for sure but that inductor might be working with the stray or intrinsic capacitance and turning Q1 into an oscillator. This would likely be in the Mhz region.

[MathWizard]

I'll have to check all the DC voltages, but in LTS, nothing seems to happen just from putting a basic inductor in the basic noise free, mostly parasitic free models.

And check the Vcc with AC coupling.

[antenna]

why the R10-R11 divider?  Seems kind of backwards to add a voltage divider to the signal before trying to amplify it.

R10 being 50k means that filter network must try to transform that high impedance to a low impedance to drive the input of that first amplifier (which looks like about 2k).  And that filter isn't doing that. Someone here once told me that audio is about driving a high input impedance with a low impedance signal, but this looks like high impedance signal is driving a low impedance input

those emitter bypass caps are probably creating so much gain that the larger signal after the amplifier couples back to the input causing your oscillations.  With several amplifier stages all doing the same, and probably talking to each other like it talks to your probe in the air, I think that would make an excellent noise source!

is it all that necessary to use emitter bypass caps at all?  Why not just control the gain with the collector/emitter resistor and reconsider the signal impedances?

[RFDx]

And I have a lot of FM radio everywhere at every node, around -63 to -55dBV most of the time. In open air my probe gets around -75dBV while the GND is still on the circuit GND. Without the GND clipped on it's back around -65 to -55dBV. (but my AWG and other scope probe were still GNDed on that circuit)

Engage the 20MHz BW limit on your scope, use the probes with the ground spring instead of the GND-wire.

1 thing I just tried was putting a 1uH inductor in series from the PSU to my circuit Vcc point. It only has 0.2 ohm resistance, and I still have my 5V DC. The total circuit current, when idling, with the mic disconnected, and no oscillations under 100MHz anyways, the idle current is 14mA, which is about what it should be with the LED on the board.

But with the 1uH inductor from the PSU, my idle current drops to 8mA, and the audio won't work either, it's gone from upto 3-4Vpp, to nothing, just the -55to-65 dBV noise floor.

Not one decoupling capacitor on VCC to be seen far and wide. In LTspice obviously not a problem, in reality not so much. If you introduce a single 1uH inductor in the VCC line, at RF all VCC connections of the three high gain stages are now suddenly high impedance while being connected together with no isolation between them.