# Do Quadrature and Bubba oscillators count as phase-shift oscillators?

I need this information for my school report. Do these oscillators count as phase-shift? I found some references here at the Texas Instruments website, but I'm not quite sure: https://www.ti.com/lit/an/sloa060/sloa060.pdf http://www.ti.com/lit/an/slyt164/slyt164.pdf

"The Bubba oscillator (Figure 9) is another phase-shift oscillator" "The quadrature oscillator is another type of phase-shift oscillator"

I think, the common classification is as follows:

1) Oscillators based on a BANDPASS as frequency-selective feedback path (zero phase at f=fo) require a positive amplifier in the loop (Example: WIEN oscillator). That means: Both parts (amplifier and feedback path) must not exhibit any phase shift at f=fo. Therefore they do NOT belong to the class of phaseshift-oscillators.

2) There are other oscillator types based on an inverting amplifier (180 deg phase shift). In this case, the frequency-selective feedback path must also produce 180 deg at f=fo in order to produce a loop phase of 0 deg at f=fo (Barkhausen condition). For this purpose, the feedback path is realized as a third-order lowpass or 3rd-order highpass circuit (Hartley, Colpitt, Pears, Clapp).

3) As a special case within the class of phase-shift oscillators, there are some oscillator types which produce the required 180 deg phase shift using two integrating stages (at least two additional opamps). In this case, the circuit produces at two different integrator ouput nodes two different signals, which are off in phase by 90 deg (sin and cos). Because these two signals are "in quadrature" these integrator-based oscillators are called "quadrature oscillator".

4.) A special case within this class of "quadrature oscillators" is the BUBBA-oscillator. Here, instead of two integrating stages (each with 90 deg phase shift) four passive RC-blocks are used (decoupled with a buffer) - each producing 45 deg. of phase shift at the desired frequency fo. However, again it is possible to use two opamp outputs for providing at the same time two signals "in quadrature".

All oscillators satisfy the Barkhausen stability criterion to function which is controlling the phase to become positive feedback at gain>=1.

So in broad terms all oscillators require phase shift even by voltage gain with inversion (180 deg).

But historically we only used the term for RC phase shifts including Bubba and his brothers and sisters.

• Tony Stewart, I think that all relevant textbooks classify the Hartley-, Pierce- , Colpitt- and Clapp-Oscillators as typical phase-shift-oscillators. The BUBBA-oscillator (and all the derivatives) are a special class of integrator-oscillators. – LvW Jun 15 '18 at 10:12

Here is a list of oscillator types that utilize phase shift to obtain oscillation: -

• All of them (apart from relaxation oscillators)

And in case there is some confusion a fuller list is: -

• Wien bridge
• Hartley
• Colpitts
• Pierce
• Armstrong
• In German "Wein" is fermented grape juice and "Wien" is the capital of Austria (and also the last name of the inventor of the Wien bridge oscialltor). – Curd Jun 15 '18 at 9:27
• Andy aka, regarding the WIEN oscillator: This oscillator type works with a positive amplifier (zero phase) and a bandpass (also zero deg. at f=fo). Hence, it is very uncommon (wrong?) to say that also the WIEN type oscillator belongs to the class of phase shift oscillators. – LvW Jun 15 '18 at 9:49
• @LvW my answer said "utilize phase shift to obtain oscillation". I didn't say it was commonly grouped as a "phase shift oscillator". – Andy aka Jun 15 '18 at 9:53
• Andy aka, OK accepted. Nevertheless, in case of WIEN oscillator this would mean: This oscillator utilizes a phase shift of zero for the feedback path. This is correct, of course - however, my concern was that somebody could misunderstand your classification (because the question contains the wording "phase-shift oscillator"). You see what I mean? – LvW Jun 15 '18 at 10:09
• My answer is to try and highlight that all oscillators (excluding relaxation types) rely on phase shift to produce the corresponding oscillation frequency @LvW – Andy aka Jun 15 '18 at 10:15