Oscillator gain

[Atif Hussain Shah]

I am design an oscillator circuit for broad band range i-e from 100MHz to 200MHz.

The resonance frequency of the resonator changes from 100MHz to 200MHz as the Er of the substrate changes. The loss of resonator is around 4dB and its phase is around 150 to 170MHz.

With that resonator, I want to connect an amplifier in reflective mode so that the oscillator circuit completes.

The phase of amplifier is equal and opposite to the phase of Resonator so the phase condition of Bark-Hausten criteria is met but the gain of amplifier is less than the loss of resonator in 150 to 200MHz range. So Second condition of oscillator is not met in range 150 to 200MHz. The ciruit and response of Amplifier is attached below. How can I increase the gain to 4 dB. In full range(100 to 200MHz) of Amplifier.

@big Boss please help

 

[BigBoss]

Dear Atif,
Please refer to Oscillator Design Techniques in ADS and read published papers and textbooks about them.
Because you're on a wrong way and I think you confused all.
There are few oscillator design techniques and you have follow one of them otherwise you'll be stuck in somewhere. I will upload one of my designs here as ADS project file and please examine it attentively.
But first things first, read Ulrich Rohde's Textbook about Oscillator Design. You have to understand at least fundamentals about oscillators otherwise you'll be lost.

 

[Atif Hussain Shah]

Dear Atif,
Please refer to Oscillator Design Techniques in ADS and read published papers and textbooks about them.
Because you're on a wrong way and I think you confused all.
There are few oscillator design techniques and you have follow one of them otherwise you'll be stuck in somewhere. I will upload one of my designs here as ADS project file and please examine it attentively.
But first things first, read Ulrich Rohde's Textbook about Oscillator Design. You have to understand at least fundamentals about oscillators otherwise you'll be lost.

Thank you Big Boss.
I will read the read Ulrich Rohde's Textbook. Please also upload your ads file so that it can help.
Dear Big Boss, i have simulated my circuit in Harmonic Balance and its giving desired responce in 100 to 150 Mhz range.
I am using s parameter simulation to see the magnitude of S11>1(because S11= Gamma_Transistor) IN full range of transistor.

I am attaching the resonator responce and meeting point of phase condition to give a better idea what i am doing.

Resonator Response(at a single Er bc i change the resonator response by changing Er :its useful for my application)

Oscillation Meeting condition

Harmonic Balance simulation

Now for some cases my resonator loss is very high i-e S11 is around -4dB while my amplifier has gain of around 2dB so in that case Harmonic balance simulation does not converge. So thats why i am saying to increase the gain of amplifer to 4dB so that Harmonic balance simulation converges.

--- Updated May 31, 2022 ---

Thank you Big Boss.
I will read the read Ulrich Rohde's Textbook. Please also upload your ads file so that it can help.
Dear Big Boss, i have simulated my circuit in Harmonic Balance and its giving desired responce in 100 to 150 Mhz range.
I am using s parameter simulation to see the magnitude of S11>1(because S11= Gamma_Transistor) IN full range of transistor.

I am attaching the resonator responce and meeting point of phase condition to give a better idea what i am doing.

Resonator Response(at a single Er bc i change the resonator response by changing Er :its useful for my application)

Oscillation Meeting condition

Harmonic Balance simulation

Now for some cases my resonator loss is very high i-e S11 is around -4dB while my amplifier has gain of around 2dB so in that case Harmonic balance simulation does not converge. So thats why i am saying to increase the gain of amplifer to 4dB so that Harmonic balance simulation converges.

I HAVE USED inverted phase of amplifier so that we can see where phase condition of oscillator meets.

 

[BigBoss]

Look at an eye first here.

Synergy Microwave Corporation - Published Technical Articles

synergymwave.com

Then research ADS examples after that search google to find valuable papers.

 

[BigBoss]

Here is the archived 2.4GHz Oscillator project for ADS 2022U2.
Unarchive it and look how I did arrange the simulation controllers.
Don't do HB, Transient and Small Signal Oscillator Analysis simultaneously and attribute different Datasets for both.
Look at Small Signal analysis results and see how S11 encircles 1+j*0 point that gives at the same time the number of RHS Poles.
Read and work my friend, the best way to have knowledge.

File on MEGA

mega.nz

 

[Atif Hussain Shah]

Here is the archived 2.4GHz Oscillator project for ADS 2022U2.
Unarchive it and look how I did arrange the simulation controllers.
Don't do HB, Transient and Small Signal Oscillator Analysis simultaneously and attribute different Datasets for both.
Look at Small Signal analysis results and see how S11 encircles 1+j*0 point that gives at the same time the number of RHS Poles.
Read and work my friend, the best way to have knowledge.

File on MEGA

mega.nz

Big Boss i am very thankful for your help and support.

 

[biff44]

i would try to screw around with the existing circuit.
technically i would not call it getting "more gain", so much as getting the reflection coefficient magnitude to be >>1 above 150 MHz and of the right phase.

for instance, what would happen if L5 was 20 nh?

 

[Atif Hussain Shah]

i would try to screw around with the existing circuit.
technically i would not call it getting "more gain", so much as getting the reflection coefficient magnitude to be >>1 above 150 MHz and of the right phase.

for instance, what would happen if L5 was 20 nh?

yes you are right. "to get the reflection coefficient magnitude to be >>1 above 150 MHz and of the right phase". is what i want.

 

[Atif Hussain Shah]

i would try to screw around with the existing circuit.
technically i would not call it getting "more gain", so much as getting the reflection coefficient magnitude to be >>1 above 150 MHz and of the right phase.

for instance, what would happen if L5 was 20 nh?

biff44​

I have a confusion in one more thing. The phase condition of Reflection Coefficients meet at 138MHz. So i expect that the Oscillator oscillates at 138MHz but In Harmonic Balance simulation, oscillating frequency is around 134MHz. i am attaching the graph of both. I dont know why in harmonic balance simulation, oscillating frequency is shifted from the expected? Biff44 can you help?

PHASE CONDTION MEETING POINT OF REFLECTION COEFFICIENTS

HARMONIC BALANCE SIMULATION RESULTS

 

[BigBoss]

134MHz vs 138MHz It's normal and pretty good.
Increase the Harmonic Number and check the Oscillator's Collector/Drain current in time domain to able to understand about the health of your oscillator.

 

[Atif Hussain Shah]

134MHz vs 138MHz It's normal and pretty good.
Increase the Harmonic Number and check the Oscillator's Collector/Drain current in time domain to able to understand about the health of your oscillator.

BigBoss i have increased the number of harmonics to 7. Collector current in time domain is attached. But i dont know if it is good or not?

 

[BigBoss]

Not bad, increase more. For instance make it 15 or more.
Your oscillator seems working but it needs some optimizations. ( my belief )
Also do a transient analysis to observe oscillation starting point in time domain until the oscillations settle down.
And use small time steps in transient analysis and compare HB results.

 

[Atif Hussain Shah]

Not bad, increase more. For instance make it 15 or more.
Your oscillator seems working but it needs some optimizations. ( my belief )
Also do a transient analysis to observe oscillation starting point in time domain until the oscillations settle down.
And use small time steps in transient analysis and compare HB results.

BigBoss thank you for your reply and help.
i will do transient analysis.
You said optimization is required. i want to ask which parameter should i optimize?
And My Transistor has 1 Emitter(with two emitter pins). Can i use just only one emitter pin and left the other emitter pin unconnected ?

 

[BigBoss]

BigBoss thank you for your reply and help.
i will do transient analysis.
You said optimization is required. i want to ask which parameter should i optimize?
And My Transistor has 1 Emitter(with two emitter pins). Can i use just only one emitter pin and left the other emitter pin unconnected ?

If your transistor has 2 Emitter pins, you should connect those pins together. This is related to Package and of course package parasitic elements.
Optimization is done for;
-High Output Level. ( I expect higher Output Level @ 135MHz. Let's say around 13-16dBm@50Ohm
-Harmonics are very high. Optimization is necessary at this point.
-Phase Noise ( not known )
-Drift by Temperature ( not known)
-Modulation Feature ??
-Pulling by measurement
-Pushing

Did you read Rohde's textbook ? Or Grebennikov's ? You'd better read before going on.

OK, I have pointed out some important points on oscillator design. The rest, is up to you.

Addendum : The Most Important Thing : The Layout
You have to design a appropriate layout after that you should simulate this layout in a planar EM simulator such as Momentum then place the realistic components with proper models.
You will see your oscillator-maybe- wouldn't work. Black magic !!.

 

[Atif Hussain Shah]

If your transistor has 2 Emitter pins, you should connect those pins together. This is related to Package and of course package parasitic elements.
Optimization is done for;
-High Output Level. ( I expect higher Output Level @ 135MHz. Let's say around 13-16dBm@50Ohm
-Harmonics are very high. Optimization is necessary at this point.
-Phase Noise ( not known )
-Drift by Temperature ( not known)
-Modulation Feature ??
-Pulling by measurement
-Pushing

Did you read Rohde's textbook ? Or Grebennikov's ? You'd better read before going on.

OK, I have pointed out some important points on oscillator design. The rest, is up to you.

Addendum : The Most Important Thing : The Layout
You have to design a appropriate layout after that you should simulate this layout in a planar EM simulator such as Momentum then place the realistic components with proper models.
You will see your oscillator-maybe- wouldn't work. Black magic !!.

I am very thankful to you Big Boss.
I have designed the layout, performed em simulation and then co simulate it in ads. The results are attached.

Schematic

CO-Simulation

CO SIMULATION RESULTS

In cosimulation, the frequency has been shifted from 134 to 132.6 MHz.

--- Updated Jun 9, 2022 ---

I am very thankful to you Big Boss.
I have designed the layout, performed em simulation and then co simulate it in ads. The results are attached.

Schematic

CO-Simulation

CO SIMULATION RESULTS

In cosimulation, the frequency has been shifted from 134 to 132.6 MHz.

Big Boss i am reading the books. Thank you for providing.
One thing, i also worry is that power level of 2nd harmonic is very high and that should be reduced.

 

[BigBoss]

What is Term G2 ? What it's for ?

 

[Atif Hussain Shah]

What is Term G2 ? What it's for ?

Term G2 is 50ohm termination. My load is at 50ohm

 

[BigBoss]

Term G2 is 50ohm termination. My load is at 50ohm

You have already a Load at Collector ?? It's impossible to get Oscillation signal from Coaxial Resonator side in this configuration. I think you're doing something wrong.

 

[Atif Hussain Shah]

You have already a Load at Collector ?? It's impossible to get Oscillation signal from Coaxial Resonator side in this configuration. I think you're doing something wrong.

Coaxial cable is not the resonator. It connects the resonator and transistor. Resonator is encircled by black colour.

BIGBOSS above circuit is in colpit configuration. Why you say that its impossible to oscillate?
Look at the configuration i have attached from Rohde Book below

 

[BigBoss]

I'm asking why there is a port PortG2 ? You load the resonator and if the Quality factor drops down a certain level, the oscillator will not oscillate. You're doing something wrong buddy. Please examine your task carefully.