Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics 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

LNA DESIGN AND SIMULATION

nithinp

nithinp

Junior Member level 2
Joined
Dec 26, 2023
Messages
20
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Location
India
Activity points
112
Hi everyone, I'm designing a low noise amplifier (LNA) for 4-5 GHZ (C band) for my final year major project with a specification of
Gain:60dB or greater
NF: <3dB
Operating voltage:+12/15/20
OIP3:+20dBm min
Design process:130nm/90nm
I'm having problem with designing it mainly choosing Id and gm, and also the high gain requirement , also the various parameters like Cgs, fT, gamma, alpha, Cdb.
Please help
 
D.A.(Tony)Stewart said:
The goal of your education is that you will learn how to learn to find the answers on your own. You seem to be lost.
Click to expand...
Im a beginner can you help please by suggesting how to proceed , i donot know about parasitic parameters of mosfets, i know the characterization of mosfets but my mentor said that the operating voltage of the industry lna is +12v,but typically vdd for 130 or 180nm mosfet is 1.8V
 
Last edited:
I'd say google it, or use GPT4, (elektrodabot) I'm sure I have a book on this subject

12/15/20 seems rather high for an LNA. How about just 5V and go shopping for an LNA at Skyworks now. Learn & Master ADS. Create a block diagram with all the specs in a list for testing.

A design might look something like this:
1. First Stage: Common-source with inductive degeneration for low noise figure and moderate gain.
2. Intermediate Stages: One or more common-source or cascode stages to further increase gain while keeping noise and linearity in check.
3. Final Stage: Designed for linearity (OIP3 > +20 dBm), possibly using a different bias point or device geometry to achieve the desired linearity without compromising too much on noise or gain.

For the first stage, you'll want to focus on:
- Noise Matching: Adjusting the source inductance for optimal noise match which typically occurs at a source reflection coefficient that is different from the one for power match.
- Impedance Matching: Designing input and output matching networks for the 50-ohm standard, while also taking into consideration the stability of the amplifier.
- Biasing: Selecting the proper Id to achieve a good balance between low noise figure and acceptable power consumption. Typically, a moderate Id is chosen to reduce the noise figure without drawing too much current.

For the subsequent stages:
- Gain Stages: Ensure each stage is properly matched to the preceding stage to maintain good gain flatness and avoid mismatches that could lead to reflections and instability.
- Inter-stage Matching: This can also help in controlling the bandwidth and compensating for any gain variations across the band.
--- Updated ---

I found this. Did you:?

High Linearity - Skyworks | DigiKey

Skyworks' white paper titled "Ultra-Low Noise Figure, High Gain Amplifier with High Linearity".
www.digikey.ca www.digikey.ca

Order Kit and sample IC chip to compare with your Test Methods (Calibration)

 
Last edited:
Last edited:
First, 60dB Power Gain is pretty pretentious. Even it will be implemented as RFIC, serious stability problems will occur and prediction of these potential stability problems will not be very predicted. The problems that you face to face are "General Design Challanges" and the Foundry should normally give you some hints about those variables. Because they have already measured and defined which one/model is appropriate for Low Noise which one is adequate for Linear design etc.
Second, you have to get some help Foundry Guys about those issues. This is valid for all aspects such as Linearity, Low Noise, High Power etc.
As a first step, you can check the transistors in different operating conditions you may find Optimum Noise Impedance for first stage.
If you Cadence Spectre, simple Noise Simulation will give you all about Low Noise case.
But if you want to understand the theory behind all these parameters and their optimizations, you should read "Tradeoffs and Optimization in Analog CMOS design", David M. Blinkley ( Wiley Press )
We cannot help here how to optimize the Design.
 
If you can imagine it , you can simulate it.

If you can buy it, you can test it then reverse engineer it and learn faster. Time is money. But not cheap. It is better to box 2 stages in shielded compartments. Then sell it for $5k

1703631884127.png



5 THz GBW is certainly begging to become an oscillator $2,566.67 for 1W 30 dB gain
>$6k 42 dB gain https://www.fairviewmicrowave.com/1.5db-nf-low-noise-amplifier-42db-fmam1075-p.aspx GaN Input
--- Updated ---

BTW you can stare at these the wrong way and generate enough E-field to burn out the front ends, so be ESD wise to <28V so have fun.
 
Last edited:
I'm a Student you think i can afford it, I'm not implementing it physically ,I just have to design and simulate it in some software that's all
 
I had to chuckle when I got to the part about wanting to do
4GHz in 130nm CMOS with a 12V supply.

Pick any two, maybe.

But indeed the whole point of a senior project is that you
would study, learn and work on your own.

Now a valid way to begin would be to trudge on over to the
library where they ought to have IEEEXplore (that's where
the IEEE makes money off the authors it doesn't pay) and
you could research exactly how people have done LNAs
with at least some combination of the attributes you claim
are required.

Then you might have -a- question about a -specific- thing
which somebody would feel like they could and want to
answer.

You don't make CMOS LNAs because you like CMOS LNA
performance for what it is. You make CMOS LNAs so that
you can integrate them with other things. So you do not
care one bit about 12V unless you are working in dGaN.
You care about your fellow travelers' working supply
because you are not special.

You had best keep a more critical eye and ear with this
"mentor". What piece-part LNAs use for supplies has not
much to do with what a 90nm CMOS RFIC can use. You
are directed away from the cell phone bands where CMOS
-may- be used for FEM (for mass market cost) but in an
odd band with no real volume, it's going to be a spec war
and not about the pennies. I doubt you can find a CMOS
LNA on the market with the specs you declare. But that's
another valid way to begin - finding products that meet,
then proceed to figuring out why and how.

Speaking of specs - you want +20dBm OIP3 which is 100mW
which is >3V P-P. And that's not going to happen with a single
or a cascode 90nm or 130nm FET - you'll see BVdss first.
Unless you are going to impedance-transform which I don't
think so, in a RF front end. Never heard of a Doherty LNA.
 
nithinp said:
I'm a Student you think i can afford it, I'm not implementing it physically ,I just have to design and simulate it in some software that's all
Click to expand...
One can earn $200k/y with advanced LNA design experience, but it can take 20 or 5 years depending on your experience you go after
--- Updated ---

48V, 12V, & 5V USB are the industry DC standards everything else is converted
 
Last edited:
I think it makes sense, bcs the data is industry standard
 
Last edited:
GPT 3.5 in addition to 48V, 12V, and 5V, there are several other common DC voltage standards used in various applications. Here are a few examples:

  1. 3.3V DC: This voltage is commonly used in digital electronics, microcontrollers, and low-power devices.
  2. 24V DC: Used in industrial control systems, some HVAC (Heating, Ventilation, and Air Conditioning) systems, and certain automotive applications.
  3. 9V DC: Commonly used in battery-operated devices like transistor radios, guitar pedals, and certain electronic toys.
  4. 1.8V DC: Found in many low-power applications, mobile devices, and some integrated circuits.
  5. 15V DC: Used in certain audio equipment, older laptops, and other electronic devices.
  6. 36V DC: Often used in electric vehicle systems, power tools, and some industrial applications.
  7. 19V5 DC: Commonly used as a standard voltage for laptop power adapters.
These are just a few examples, and the choice of voltage depends on the specific requirements of the devices and systems in which they are used. Additionally, advancements in technology and changes in industry standards may introduce new voltage levels or alter the prevalence of existing ones. Always refer to the specific requirements and standards relevant to the application or device in question.
--- Updated ---

20 dB gain per stage might be a good starting point
 
Last edited:
hey guys, im having trouble characterizing the mosfet, Can you please help, the below image shows various parameters obtained for the transistor. this design is for vdd = 1.8v, gain > 15dB, NF<3dB, f=4-5GHz vdd = 1.8v, NF<3dB, f=4-5GHz, here im mainly having problem with choosing gm, Id , Because i need gain, also can someone help me with adding fingers at the gate terminal to decrease Rg and decrease NF

Screenshot 2024-01-03 155618.png
Screenshot 2024-01-03 155648.png
 
Last edited:

LaTeX Commands Quick-Menu:

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top