I am using max2837 in the receiver chain.
I am feeding data from signal generator at -100 dBm and i can see baseband data after max 2837.
But when connect lna between signal gen and max2837 baseband data is saturated even when the signal generator is not switched on.

Rf experts may explain it why?

 

I am using max2837 in the receiver chain.
I am feeding data from signal generator at -100 dBm and i can see baseband data after max 2837.
But when connect lna between signal gen and max2837 baseband data is saturated even when the signal generator is not switched on.

Rf experts may explain it why?

A schematic diagram would be helpful along with the gain of the LNA. I'm not clear on where you are claiming the saturation occurs. -100 dBm is a pretty good input level for most communication receivers.

 

Its not aboout what gain of lna am i using.actually i am using 38 db gain. The issue is i am just connecting lna before max 2837.
just powering on lna and max 2837. Then i should see baseband data .ie if i feed 2.001 ghz and lo of max is 2 ghz then i shoud see 1 MHz at the output og max 2837 as i am using mixer option. But i am seeing a saturated signal .is it because my noise floor has increased after lna??

 

So, when you insert the LNA, do you drop the output of the signal generator by 38 dB?

 

When i insert lna even when I am not connecting signal generator then i get saturated output after max 2837.

To answer your question.

If i am feeding from signal generator at -100dBm to max2837 without lna then i get output without saturation.
So what i expect is that if i have a signal at -138dbm with 38db gain of lna. The max2837 gets at -100dbm which should work.
But what i see it works fine at around -70dbm .below that it gets saturated.
There should be a logical explanation to this.
Please let me know

 

quesses
1) LNA have DC co-bias in the output. Kill it whole.
2) LNA have oscillations. Kill it by proper negative loop.
3) Wire between LNA and tuner is receiving of EMI. Kill it short and well defended.
4) Materials of wire makes contact between metals dislike one each another (like copper and alumina) thus the milivolts of thermocouple effect are making a problem. Kill it.
5) The LNA output impedance is badly adjusted to input impedasnce thus the kind of virtual resonant tank is formed. Kill it by impedance tuning.

 

I am using max2837 in the receiver chain.
I am feeding data from signal generator at -100 dBm and i can see baseband data after max 2837.
But when connect lna between signal gen and max2837 baseband data is saturated even when the signal generator is not switched on.

Rf experts may explain it why?

It sounds something wrong with the LNA. I would suggest to isolate the problem. Just connect the generator and LNA, then measure the LNA output with spectrum analyzer. It is rather easy to see if there is self oscillating or other problems.

 

quesses
1) LNA have DC co-bias in the output. Kill it whole.
2) LNA have oscillations. Kill it by proper negative loop.
3) Wire between LNA and tuner is receiving of EMI. Kill it short and well defended.
4) Materials of wire makes contact between metals dislike one each another (like copper and alumina) thus the milivolts of thermocouple effect are making a problem. Kill it.
5) The LNA output impedance is badly adjusted to input impedasnce thus the kind of virtual resonant tank is formed. Kill it by impedance tuning.

I am using series capacitor which should kill DC. impedance matching is taken care. how to kill oscillations or any other issue

 

RE:""how to kill oscillation""

Most effective way is cascode cascade. Here is LNA with about 60 dB at multi-GHz range. Input cascade is modified Elector-Electronics from herehttps://xellers.wordpress.com/electronics/1ghz-active-differential-probe/
After this buffer stays cascode cascade according the https://www.researchgate.net/public...low-noise_amplifier_for_wireless_applications realized on bjt named BFU910X (90 GHz capable) according https://encrypted-tbn0.gstatic.com/...1FXJ8s5MfLQUQOJE3_WIVBidFdkQRzzVoI1fhC_QLK8B5
How far I see, it is stable and fine.
Other contrary different way is named bjt within 2 or 3 series cascade giving 20-40 dB. Then input impedance is 50 Ohms. Like here (note the bjt must be changed to more modern). Or similar to this https://www.electroschematics.com/simple-cable-tv-amplifier/