# Polyphase multivibrator (real vs SPICE simulation)

i have a following schematic for three-phase multivibrator:

SPICE simulation shows that it should work (11 volts peaks on collector nodes):

However, when i make this circuit on breadboard it does not work, only left and right LEDs are on, and no fluctuations occur. I have tried to google polyphase multivibrator circuits but non of them seem to work (i always use PN2222A). Is it PN2222A transistors fault? Thank you for help.

UPDATE 1 : As pointed out by Andy Aka i did make a wiring mistake (forgot to connect middle transistor's emitter to ground) . Circuit started to oscillate, however only left transistor visibly cuts off, middle is roughly half off, and right transistor only slightly dims. Ring oscillation is present.

Oscilloscope shows additional 600kHz oscillations at the collector of the right transistor.

• Breadboards have awful parasitics, there is cross capacitance between each row, parasitic resistance between wire and contact. These are not built into your model so what you are modeling does not reflect what you have built on the breadboard. In addition, breadboarding can be prone to errors in implementing the circuit. Wire breaks and improper positioning can lead to problems. – Voltage Spike Apr 4 at 18:19
• Thank you, i naively assumed that for such circuit these effects are of minor importance, since expected oscillations are around 1 Hz. – Asphir Dom Apr 4 at 19:27
• Make sure you use electrolytic capacitors in LT spice – Voltage Spike Apr 4 at 21:27

I decided to build this circuit on the breadboard also

As you can see I decided to remove $$\5k\Omega\$$ resistor.

The LTspice result:

And this is what I saw on my Rigol

Not bad as you can see.

• I am using LTSpice for Mac, and i could not reproduce your simulation result, with PN2222A removing 5k resistor gives no ring oscillations. I tried using BC547B like you, but still no oscillations. Collector voltages are always at 0. – Asphir Dom Apr 5 at 10:23
• Do you know how to set the initial condition for a capacitor in LTspice?\ – G36 Apr 5 at 10:25
• I did not do that, yes. – Asphir Dom Apr 5 at 11:38
• Thank you, putting charge on one of the capacitors, produces oscillations. – Asphir Dom Apr 5 at 12:39
• Unfortunately i could not reproduce your result on a breadboard, and get the perfect collector curves. Using PN2222A transistors this circuit does not work. I changed transistor to BC547 and got oscillations only at start, they quickly die out and only two right LEDs are ON. I then added 5k resistor back, it does help to sustain oscillations but oscilloscope curves are nowhere near to perfect ones you have shown. Only the middle transistor seem to follow LTSpice simulation. Any suggestions what i am doing wrong? – Asphir Dom Apr 7 at 19:27

However, when i make this circuit on breadboard it does not work, only left and right LEDs are on, and no fluctuations occur.

If somehow the circuit has started up in a benign manner, there's every chance that it won't form a ring oscillation. When this happens C1, C2 and C3 might as well be out of circuit and, what you'll find is that all the transistors are partially biased on.

Hence, when you say left and right LEDs are on, this tells me that you have likely made a wiring error or you have used electrolytic capacitors for C1, C2 and C3 and they (or at least one of them) has become reverse biased and is/are no longer working as you'd expect it/them to.

• Thank you, indeed i made wiring mistake, forgot to connect middle transistor emitter to ground. Circuit started to work, but not exactly as intended. Only first led visibly switches off, while second goes half way off, and third one(the right one) only very slightly dims, never really going off. I connected oscilloscope to collector of the right LED (the one with smallest oscillations) and observed 600kHZ oscillations happening time to time (once per cycle?). – Asphir Dom Apr 4 at 19:21