In every particular case you can find solution, where reverse bias will eliminated.
For example:
View attachment 250872View attachment 250873

ok, so in this case , the Electrolytic capacitor is positively biased.

 

Positively biased or negatively biased is relative to something not specified, in any case it does experience reverse bias.

 

Electrolytic capacitor is reverse biased,
when voltage on anode is negative relative to cathode
or voltage on cathode is positive relative to anode.

 

Positively biased or negatively biased is relative to something not specified, in any case it does experience reverse bias.

for right design Electrolytic capacitor can avoid reverse bias, as the simulation shows.

 

Sorry about the omission. I wrote: in any case it does experience reverse bias, but one word was missing. Should be:
...in any case it does NOT experience reverse bias.

 

Sorry about the omission. I wrote: in any case it does experience reverse bias, but one word was missing. Should be:
...in any case it does NOT experience reverse bias.

yes, the design is very cute.

 

The collector voltage on the transistor that is conducting does not go to ero, but does drop to some low voltage if the transistor is fully saturated. and so the reverse voltage will mostly serve to remove the chathe on the capacitor, and once that charge is removed the circuit will switch. so the apparent reverse bias is not a high current state, and will not be a serious problem. And the final circuit in post #20 removes the problem completely.

 

In every particular case you can find solution, where reverse bias will eliminated.
For example:

I wonder where the short spikes at the start of the LED currents come from?

With an estimate on 25 pF Base / Emitter capacity and 10 KΩ Base resistor on opposite transistor, I get a time constant of a quarter of a microsecond.

And with the stated resistors, I also think that the peak currents should be higher than 5 mA

 

I wonder where the short spikes at the start of the LED currents come from?

With an estimate on 25 pF Base / Emitter capacity and 10 KΩ Base resistor on opposite transistor, I get a time constant of a quarter of a microsecond.

And with the stated resistors, I also think that the peak currents should be higher than 5 mA

If those short current spikes (post #20) actually exist, they can only come from discharging the capacitors discharging against the base of the transistor that is switched off. So I am suspecting that they may be a simulator software artifact, only present in simulation, not in reality.

 

@Danko
How did you get your oscillator to start?

 

@Danko
How did you get your oscillator to start?

It was not start with TS's 470 and 5.1k resistors but with 200 and 10k works just fine.

 

I wonder where the short spikes at the start of the LED currents come from?
With an estimate on 25 pF Base / Emitter capacity and 10 KΩ Base resistor on opposite transistor, I get a time constant of a quarter of a microsecond.
And with the stated resistors, I also think that the peak currents should be higher than 5 mA

Classic multivibrator

Current spike
C1 current goes not only through capacitance of Q2 B-E, but through capacitance of Q2 B-C too, or maybe it is Q2 Miller effect...
________

 

The collector voltage on the transistor that is conducting does not go to ero, but does drop to some low voltage if the transistor is fully saturated. and so the reverse voltage will mostly serve to remove the chathe on the capacitor, and once that charge is removed the circuit will switch. so the apparent reverse bias is not a high current state, and will not be a serious problem. And the final circuit in post #20 removes the problem completely.

somebody says reverse bias is not good, somebody says is ok. lol

I wonder where the short spikes at the start of the LED currents come from?

With an estimate on 25 pF Base / Emitter capacity and 10 KΩ Base resistor on opposite transistor, I get a time constant of a quarter of a microsecond.

And with the stated resistors, I also think that the peak currents should be higher than 5 mA

Classic multivibrator
View attachment 251091View attachment 251093
Current spike
C1 current goes not only through capacitance of Q2 B-E, but through capacitance of Q2 B-C too, or maybe it is Q2 Miller effect...
View attachment 251085________View attachment 251088

I think C1 is big enough to absorb current from base-emitter or base-collector capacitors. don't know why it suddenly has a -50ma current. I think the current should go from power to R2 to C1 and Q1, to encharge C1, and let point b open Q2. I'm not sure if this enchargement can make current spike of Q1. or how C1 current comes into being.

 

Many thanks to @Danko for the additional measurements.

 

When Q1 is conducting, it's base current is higher, and so C2 charges up thru the 200 ohm resistor. Then when Q2 conducts, C2 discharges thru the Q2 collector because the base of Q1 is at 0.7 volts, or close to it. Thus the very short current spike.

 

When Q1 is conducting, it's base current is higher, and so C2 charges up thru the 200 ohm resistor. Then when Q2 conducts, C2 discharges thru the Q2 collector because the base of Q1 is at 0.7 volts, or close to it. Thus the very short current spike.

when Q2 conducts, the base of Q1 is about 0.3V

 

when Q2 conducts, the base of Q1 is about 0.3V

OK, still a low enough voltage to have the capacitor discharge.

 

when Q2 conducts, the base of Q1 is about 0.3V

when Q2 conducts, the base of Q1 is about 0.3V

sorry it's collector about 0.3V.

 

For the simulation, where changes are easy, try adding a 1K resistor in series with each capacitor, and see what that does to the current spike.

 

In fact, current Ib(Q2) is sum of currents Ie(Q2) and Ic(Q2) of minority charge carriers,
earlier stored in base and flowing from base region under impact of reverse
base bias. At 8.5ns base region is devoid of stored charge and Q2 is OFF.