reverse BJT conduction on peltz oscillator?

[vivi-d]

It looks like the current thru Q1 is positive and negative. How is this possible? I thought BJTs only conducted one way.

What's going on?

[mawyatt]

That's conduction thru the Base Collector PN Junction when the NPN Collector swings below the Base by a Junction Drop. This is shown as sourcing current from the NPN Collector rather than the NPN Collector sinking current in normal bias condition.

Best

[vivi-d]

I see. That makes sense. Is there any current gain with this base-collector conduction?

[mawyatt]

I see. That makes sense. Is there any current gain with this base-collector conduction?

No current gain in this case. However, there's always a however one can operate a bipolar transistor upside down, or inverted by swapping the normal Collector and Emitter terminals, so in this configuration there is current gain altho it's much lower than normal operation.

Long ago folks used this inverted configuration as a saturated switch since the saturation voltage is very low, sometimes below 10mv, where in normal saturated operation it's well above this. With modern FET switches this transistor inverted operation has generally been forgotten.

Best 

[Zero999]

BJTs do work in reverse. The hFE is much lower and the saturation voltage a little lower. Here's a bidirectional level circuit which uses the reverse mode to translate 5V logic levels to 3.3V and the forward mode to go from 3.3V to 5V.