Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Oscillation in RF power amplifier

Mabrok

Full Member level 4
Joined
Mar 1, 2017
Messages
212
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,477
Hi. How I can Identify the oscillation of RF power amplifier in the signal shown on spectrum analyzer? Is it by observing the spikes on the signal (in other words, for the PA to be stable the signal should be smooth without any spikes), Am I right?

Attached photo from my lab's spectrum analyzer for the reference. Thanks in advanced
 

Attachments


barry

Advanced Member level 5
Joined
Mar 31, 2005
Messages
4,985
Helped
1,089
Reputation
2,188
Reaction score
1,074
Trophy points
1,393
Location
California, USA
Activity points
27,283
1) Your amplitude is too high, you're saturating the specturm analyzer. Adjust your attenuation.
2) You apparently have a second harmonic. Maybe it's from overdriving the SA. See [1].
3) Your bandwidth is too wide to really tell much about the signal, other than that harmonic.
4) A perfectly "stable" amplifier can have spikes if it's non-linear.
 

    Mabrok

    points: 2
    Helpful Answer Positive Rating

Mabrok

Full Member level 4
Joined
Mar 1, 2017
Messages
212
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,477
1) Your amplitude is too high, you're saturating the specturm analyzer. Adjust your attenuation.
RF signal generator is off and I have used 40 dB attenuator. My device capability is 40 dBm power at the most. The signal band was set from 10 Hz to 6GHz, but the interested band designed band only 3.4-3.6 GHz. So, How the stable amplifier signal should looks like (based on the given picture)? Thank you
 
Last edited:

barry

Advanced Member level 5
Joined
Mar 31, 2005
Messages
4,985
Helped
1,089
Reputation
2,188
Reaction score
1,074
Trophy points
1,393
Location
California, USA
Activity points
27,283
I don't understand your question. What do you mean "identify the oscillation"? It's an oscillation at about 2.782GHz at over 40dBm. It's right there. Zoom in on it if you want more exact measurement.

If there's no input signal, then your PA is generating OVER 40 dBm, since it's saturating the SA at 0 dBm. Again, ADJUST THE ATTENUATION OF YOUR SA.
 

Mabrok

Full Member level 4
Joined
Mar 1, 2017
Messages
212
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,477
I don't understand your question. What do you mean "identify the oscillation"? It's an oscillation at about 2.782GHz at over 40dBm. It's right there. Zoom in on it if you want more exact measurement.

If there's no input signal, then your PA is generating OVER 40 dBm, since it's saturating the SA at 0 dBm. Again, ADJUST THE ATTENUATION OF YOUR SA.
I mean how I can know that there is an oscillation in the signal (in general for RF power amplifier)? What are the signs that I observe on SA spectrum so that I can say that I have oscillation at xx certain frequency?

One more thing oscillation at 2.782 GHz is out of band (3.4-3.6 GHz interested band), Does this out of band oscillation affect my design response?
 

barry

Advanced Member level 5
Joined
Mar 31, 2005
Messages
4,985
Helped
1,089
Reputation
2,188
Reaction score
1,074
Trophy points
1,393
Location
California, USA
Activity points
27,283
How do you know there's an oscillation? You have no input, but the the output is >40dBm at 2.782GHz. If that doesn't tell you that you've got an oscillation, nothing will.

Does it affect your design? You are blasting out 10Watts of a signal you don't want!!! Just because it's not in your "interested band" doesn't mean you can ignore it. If flames were shooting out the back of your car, would you ignore it because you're not standing behind your car?
 

    Mabrok

    points: 2
    Helpful Answer Positive Rating

Mabrok

Full Member level 4
Joined
Mar 1, 2017
Messages
212
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,477
How do you know there's an oscillation? You have no input, but the the output is >40dBm at 2.782GHz. If that doesn't tell you that you've got an oscillation, nothing will.

Does it affect your design? You are blasting out 10Watts of a signal you don't want!!! Just because it's not in your "interested band" doesn't mean you can ignore it. If flames were shooting out the back of your car, would you ignore it because you're not standing behind your car?
So, Does it mean the amplifier have to be no oscillation( stable) from the smallest to the highest possible frequency of SA (In my case 10 Hz to 13 GHz)?
How much the amplitude of the signal should be to ensure the stability (based on the attached picture)?
 

Easy peasy

Advanced Member level 5
Joined
Aug 15, 2015
Messages
2,779
Helped
1,005
Reputation
2,010
Reaction score
1,017
Trophy points
113
Location
Melbourne
Activity points
15,488
you need to be very careful you don't destroy the front end of your spectrum analyser as you clearly don't have the training necessary to use it properly ....
 

barry

Advanced Member level 5
Joined
Mar 31, 2005
Messages
4,985
Helped
1,089
Reputation
2,188
Reaction score
1,074
Trophy points
1,393
Location
California, USA
Activity points
27,283
I think you are missing some seriously important knowledge. If your amplifier oscillates at ANY frequency, it’s unstable. There’s no minimum amplitude to “ensure the stability”. If there’s no input signal, the output should be zero (other then noise).
 

Toggle Sidebar

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top