2N2222 astable multivibrator not starting on it's own

Thread Starter


Joined Jul 13, 2017
Hi all,

I've build the astable multivibrator with 2N2222 transistors. As you'll find many examples of on internet.
All works, but I'm still wondering why it won't start oscillating after cold start.

I've read other threads and they mention for example "Since no two devices are identical, one or the other will have slightly higher gain than the other. The one with the highest gain will start to conduct slightly more than the other. That slight difference results in the collector of that transistor starting to drop voltage slightly." as why it would start oscillating.

However mine just won't start on it's own, I have to pull one side to ground for it to start.
I've attached some images from my oscilloscope. Probes are on the + side of the capacitors.

What I noticed is that both charge up 'till 188mV initally (InitialPower.png), which is the bottom voltage when it's oscillating. Then I hit the 'reset' button (afterReset.png) causing the 2nd capacitor to go to 0V and in turn causes a little spike in the first capacitor. Then it starts oscillating (the other two graphs).

Reason for my question is that all sites describing on how to create this circuit, don't mention this behaviour. That made me question my setup.

The circuit is build according to 'circuit.png', see my circuit in circuit2.jpg. With:
C1/C2: 100μF
Q1/2: 2N2222
R1/4: 1k
R2/3: 10k

Hopefully someone is able to give me some insights in what's happening.
And if more info is needed, just let me know!




Joined Mar 30, 2015
if more info is needed, just let me know!
It would be helpful if you posted a schematic instead of a wiring diagram/cartoon.

With a schematic, we can glance at it to get an idea for how the circuit will operate. We can't do that from a wiring diagram.


Joined Mar 30, 2015
Tnx for the feedback. As requested the schematic.
That circuit is flawed because the base current from the transistor that's on will likely be sufficient to cause the LED that should be off to be dimly lit; or more than dimly if it's a high efficiency LED.

The only thing that looks like it could be a problem to me is the size of the capacitors.

EDIT: This is the more traditional way to draw that circuit:
This circuit requires diodes anti-parallel to the base-emitter junctions to keep them from breaking down.


Joined Jun 4, 2014
If you are running this circuit from 5V as your diagram appears to show, try connecting the positive of the capacitors to the transistor collectors.

Audioguru again

Joined Oct 21, 2019
There are thousands of electronic circuits on the internet (Instructables and on You Tube) designed by 10 years old kids who know nothing about electronics. Even their soldering is horrible.

Thread Starter


Joined Jul 13, 2017
Your schematic and photo of the Mickey Mouse solderless breadboard shows the capacitors wrongly connected to the LED and its resistor instead of being connected properly to the collectors of the capacitors.
Tnx Audioguru and AlbertHall.

Right ... completely overlooked that one ... when looking too often to the same schematics made me blind for that small difference. Not yet sure in what way this change makes the difference but I think is has to do with the voltage difference (across the LED)/the speed of which the capacitors are charged.
Any insights on this is much appreciated.

I've been changing some values. My circuit as basis:
Changing R2/3 to 100k: Circuit starts oscillating real slow (but it does start on it's own). Then when I lower C1/2 to 10uF, it also starts on it's own but is oscillating faster (as expected). Also, in this case there's no (obvious) difference in whether the caps are before or after the LED.


Joined Jun 4, 2014
Not yet sure in what way this change makes the difference but I think is has to do with the voltage difference (across the LED)
Running it at 5V, the LED voltage is a large proportion of the supply voltage and so the voltage across R1/R4 is much less than the supply voltage resulting in less feedback around the oscillator.


Joined Aug 21, 2008
Take the signals to the capacitors from the collector instead of the LEDs.

You do not need to add diodes if you are only using 5 volts as the power supply. 6.5 volts and up with a 2N2222 I might worry about long term harm to the gain of the transistors.