# Is an LR Phase Shift Oscillator possible?

All the phase shift oscillators I've seen use RC stages each of which produces a 60degree phase-shift. Is it possible to inductors instead of resistors? My understanding is that inductors also cause a phase shift, but I've seen no designs which use inductors. I tried to simulate the following circuit, but while I can get it to resonate I can't seem to find a set of values/configuration that oscillates:

If you look at the collector resistor of 100 and the emitter resistor of 10K. The transistor stage has a gain of 0.01 ("negative" x100 gain). It is actually worse than that because the input bias resistors are also 100. To get a possibility of oscillation, the overall loop gain must be greater than 1.

Since you are just playing with simulation to learn something, without doing an analysis, these values might help get the thing going.

Set the gain of the transistor stage to around 10, i.e. collector resistor 10K, emitter resistor 1K.

10 ohm series resistors to 10K, all other resistors to 50K.

Once again, these are just guesses without doing any calculation.

• hmm, I tried these values and tried adding decoupling caps, but after fiddling with it for an hour no luck. I guess I may try the Opamp version next and see if I have more success. Commented Dec 19, 2014 at 14:03

Why use inductors when capacitors are cheaper and more "perfect" as components. Yes an RL phase shift oscillator can be made but not with the circuit you have shown - if this oscillates then it's despite the problems you'll get with the 10k resistors in series with each inductor.

For a single phase shift network to be valid XL will approximately equal 10 ohms (as per the series resistors in your diagram) but, you have 10k resistors in series with each inductor and so the criteria that is aimed for in a phase shift oscillator is scuppered before you connect it to the scope.

• Yes, I can see it's not a good idea to use inductors in this application. But I'm trying to understand things a bit better. Aside from the component values is the configuration I've shown correct? I'm currently just trying to get it to oscillate in simulation, but no matter what values I try, I can't. Commented Dec 16, 2014 at 15:13
• Connect the inductors straight to ground, and put the 10K resistors where the 10R ones are now, and it'll be close. (Oh and increase the base bias Rs fro 100R to 10k or 100k).
– user16324
Commented Dec 16, 2014 at 16:28
• And... connect the collector to the phase shift network via a capacitor to stop it being dc loaded. Commented Dec 16, 2014 at 16:37
• If you "only" are trying to understand "things a bit better" why not investigating an opamp based phase shift oscillator realized with R-L highpasss or L-R lowpass stages? Such a circuit is simple to design and much easier to understand because there are (a) no bias problems and (b) no parasitics to take into account (influence of input and output resistances for the amplifying device).
– LvW
Commented Dec 16, 2014 at 16:41
• @camil I undid the edit because I don't like to use "10K". I like to use the correct "10k" i.e. the smaller letter k. Commented Dec 18, 2014 at 12:17