How to decide the input power of the RF source and the frequency space for LNA?
In a two-tone IP3 measurement the level of the tones at the input of the LNA shall be less than P1dB(output) ? LNA_Gain
The equation OIP3 = Pout + (IM3/2) is valid only when the LNA is not in compression.
Frequency spacing usually is the double of the channel spacing of your system, but this is not mandatory. Depends by your system requirements.
Hi,
1. Usually freq space it's about 1-2MHz (maximum 5MHz) otherwise biasing of amp will influence... never mond long story!
2. About input power you want to check LINEARITY of lna so you must be at linear part of yhe amp! How you check if you are in linear part you increase 1 dB on signal generator and f1 increase also by 1dB BBBUT the side band will increase only by 3dB (of cource it's true with two signal)
Anyway if you need anything else or papers about this essu tell me
David
Thanks very much! But there is a range of linear region,Should i decide a certain point for measurement? Do You mean as long as it's in linear,the measrement is right?
Hi,
You want to check linearity of your amp. I am correct?
If I am correct you must work in linear part!
David
I think the below sentance is true
Do You mean as long as it's in linear,the measrement is right?
Yes yes yes :D
As long measure is linear the test is correct!
David
I have another problem with the IP3 simulation.I use two single LNA to make a Balanced LNA.But the OIP3 is only 3 dB higher than the single one.however it says that it should be 6dB in the book.What's the matter?
The OIP3 of the balanced LNA is improved by a factor of 3dB compared with a single LNA.
Is IMD3 improved by a factor of 6dB？
Does anyone have some paper about the merits of Balanced LNA compared with single LNA?
It depends on the splitter and combiner used for the balanced configuration, i.e., 90 degree or 180 degree. Please check out "Practical Rf System Design" by William Egan.
From book "Practical Rf System Design" by William Egan, Chapter 6 ? Architectures that improve linearity
?Third harmonics and third-order IMs that are near the harmonics (at frequency sums) generated in the two identical amplifiers are ideally eliminated when 90° hybrids are used to combine them. Third-order IMs near the fundamentals (at difference frequencies) are not reduced?
And from the same chapter, 6.1.2.4, summarizes the characteristics using 180 degree hybrids.
Hi,
Read the folowing paper about increase IP3 in balanced Amp
You wiil see that when you using balanced Amp how much you can improve IP3.
In this paper you wiil find exact (simple) formula to this!
http://www.mimixbroadband.com/PDFfil...app%20note.pdf
David
So, as I said, the improvement of OIP3 for balanced LNA is 3dB
Please check out "Practical Rf System Design" by William Egan.
Thanks for your help! :)
hi, i was searching for OIP3 information and i came across this thread.
I also have a similar question for the Pros here. I came across this situation when i work on LNA, all the while i'm thinking, as long as i very far below the P1dB point,then my OIP3 should be accurate because i'm in linear region, but i did two experiment:
1)two -10dBm input power, and the delta_IM3 i'm getting is around 79dB
2)two -20dBm input power, the delta_IM3 i'm getting is still around -80dB!
I know if i deliberately choose -10dBm as my input power level,then surely i getting a higher OIP3 according to the maths, but i wish to understand, why is it if i put -20dBm input power, the delta_IM3 product i'm getting is similar?
Do anyone came across this situation before?
Thanks in advance Pros.
James
What's your performance of your LNA? Gain? P1dB? Freq Band?
How do you mix two tones together? Which component? Does the isolation good enough?
Is the spectrum analysis good enough? If you use two -10dBm, assume the gain is 30dB, the output is 23dBm. That can make spectrum analysis distortion. So that's the IM3 of SA, not your PA.
Because OIP3 is almost constant until the amplifier enters into saturation regime little by little.So if you plot OIP3 vs Pout, you'll see a straight line and then this lines will start to drop off.
OIP3 is not a real metric, it's just a figure of merit of the nonlinearity of an active NTI device.