LNA Input Overload question?

When the amplifier has a sensitive low-noise front end, normally looking at signals -80dBm or -90dBm, even the best of them will probably get damaged by an input any greater than about -0dBm. This happens when nearby radars are operating.

What is the best way to give the LNA input overload protection without messing up it's normal performance?

At what frequency is your LNA operating. If you are specifically worried about local radar, put in a notch filter to stop it. This used to be a problem around the Southampton area, where the coastal radar seemed to be operating on TV channel 37, which was the default upconverter channel for VHS video tape recorders, tweaking the output channel away from Ch. 37 cured the black bars every 5 seconds or so.
Frank

As above explained, use filters to get rid of out-of-band interference. Otherwise use a limiter but its loss ads o receiver noise figure. For radar interference, use a calibrated antenna for your radio link or another system and avoid radar signal to get to the main beam or side lobes.

Frequency is X-Band - around 9.5GHz I think.
The weather radar returns (and scatter from other radars) are in-band, and there is no prospect of stopping them with filters. What is needed is a mighty front-end that is absolutely overload-proof. Maybe I should use a power driver FET with a nearly good noise figure

I had thought that with a low enough noise figure, there might be room to put a PIN diode across the input, and bias it to short-circuit the input from a voltage derived from downstream to shut things down if the signal got too strong. Maybe even a PIN attenuator.

I am wary of simply using diodes across the input, one reversed .. or maybe a pair in series back-to-back, to limit the signal, because of the other effects they can bring with them.

Do we know of any cute fast-acting limiter circuit arrangements?

I would put in a bandpass filter, broadband enough so it had low insertion loss. If there are not too many jamming signals, a simple lowpass might work too.

After that, a passive limiter makes sense.

biff44 said:

I would put in a bandpass filter, broadband enough so it had low insertion loss. If there are not too many jamming signals, a simple lowpass might work too.

After that, a passive limiter makes sense.

Click to expand...

Most of suggested actions is hopeless. MOre filters and limiters only drives receiver NF up.
Check direction of interferers and use a better antenna if you can.

You could try class A push-pull FETs in your LNA, theoretically they produce no I/Ps (square law devices), but they will just push the overload point down stream until you have enough selectivity to notch out the interference. Using a better aerial - larger dish would be a nicer way if mechanically practicable.
Frank

jiripolivka said:

Most of suggested actions is hopeless. MOre filters and limiters only drives receiver NF up.
Check direction of interferers and use a better antenna if you can.

Click to expand...

I see u do not work on military electronics. How exactly would u suggest pointing the antenna as you taxi around on the airport tarmac

Exactly where did " antenna as you taxi around on the airport tarmac" come from - not the OP?
Frank

I had worked for a weather radar company in China, and had designed X band LNA by myself.
Fr your questions, it seems not an issue. You know when the TX transmits, there is a protection before the LNA, and the radar would control the protection to work so that less power can be leak into the LNA.
Even your leakage is about 0dBm, it is pulse signal, not CW, you should make 0dBm divided by duty cycle. It is a coarse way.
Of course you can test your LNA with 0dBm pulse signal in various cinditions.
And you can also ask the LNA vendor to do so for you.

Another solution is designing the LNA by your self and you can select each stage amplifier so that you can control the IIP1. My LNA has three stages, so that I can design more freely.

- - - Updated - - -

There is a protection diode (waveguide structure) before LNA.