Phase inverter noise issue

[knutolai]

So I'm putting together different phase-inverter circuits for a feedback looper pedal and came upon a strange noise issue with this NPN circuit:


The SPDT switch switches between a inverted and a non-inverted version of the signal fed into it. I get what seems to be RF noise when It's flipped so that the signal in non-inverted. Could this be a issue solely because I've put the circuit together on a breadboard (no shielding)? I just find it odd as it only affects the sound when its in the non-inverted position.

I was wondering if there is anything wrong about the way I have biased the transistor. I've only really been working with op-amps before this. Any suggestions or obviously visible flaws?

[dwmorrin]

It should work as shown.  What is V+? 9V? Are you using battery, power supply?
Measure the voltages at the collector and emitter if you want to see if the bias is good.
Collector is ~2/3V+, emitter is ~1/3V+.
Stick a metal box below your breadboard and connect it to circuit ground to eliminate shielding from the equation.
What is driving the circuit?  Are you hearing radio... seeing noise on the scope?

[knutolai]

Yes 9 volt. Im using a power supply with filtering caps. I'll do those measures after dinner, as well as the shielding you suggested. thx for the suggestions!

I think the noise could be radio, but I cant hear anything distinguishable in it. It just reminds me of the sounds I sometimes get from my cheapo plastic behringer pedals. And I have no money for a scope  should wish for one for christmas!

[R.G.]

It is possible that the circuit is oscillating when you load the emitter. Emitter followers like to oscillate under some conditions of loading.

See what happens when you put a 1K resistor in series with the emitter before it hits the output cap.

It's also possible that stray wiring stuff is giving you a problem. Try a 0.1uF ceramic from V+ to ground as close to the ends of the collector and emitter resistors as you can get it.

[PRR]

> anything wrong about the way I have biased the transistor

Yes. You put the Base at half the supply. Assume 8V for easy-math, base is at 4V. The emitter resistor then drops nearly 4V. The same current flows in collector resistor, which drops the same voltage, 4V. There is "zero" Volts across the transistor. Super-small signals (even air-sneaks) WILL distort.

I did the math too fast. The emitter will sit 0.6V low, so the collector is 0.6V high of half-voltage. That's only 1.2V across transistor--- it will pass small signals but fuzz easy. Also I've neglected base current drop in the 2*220K base resistors. With 9V supply you may have 3V at emitter, 3V at collector. Works, but a bit shy of what it could be.

The *rough* right-answer for this type phase-splitter is 1/4 supply at input. Say 330K and 100K. Then trim for base current. Also for load current. The right answer may be 1/4 to 1/3 of supply. Almost never "half".

I don't see why the emitter output should be exceptionally susceptable to RF. A long-long line can bring in enough RF to rectify the emitter junction, not the collector junction, but that's a lot of RF.

> put a 1K resistor in series with the emitter before it hits the output cap.

Always a wise thing.

[knutolai]

Thx guys! I'll test this out as soon as I get back to my workshop.

> anything wrong about the way I have biased the transistor
Yes. You put the Base at half the supply.
The *rough* right-answer for this type phase-splitter is 1/4 supply at input.

Is it different for NPN buffers? I've kinda based the circuit (and the drawing ) on the basic NPN buffer from Jack Ormas "basic buffers" page:

The signal applied at the input in my tests was very weak, so that probably explains why I didn't have any distortion issues.

[PRR]

You need room for output(s) to swing.

That buffer has one output.

Your phase-splitter has two outputs.