1
\$\begingroup\$

Wiring amplifiers in parallel in a summing configuration improves the signal to noise ratio because amplifier noise in each LNA is uncorrulated. However, external noise sources will still be a factor because both LNAs receive those noise sources, thus the external noise sources are correlated.

With 2 LNAs, would any of these be useful to reduce the correlation of external noise sources? If so, what might be best?

  • Physically rotate LNA1 90-degrees from LNA2
  • Physically rotate LNA1 180-degrees from LNA2
  • Mount one LNA flat on the board and one vertical or on its side (think horizontal vs vertical polarization)

-Eric, KJ7LNW

\$\endgroup\$
4
  • 2
    \$\begingroup\$ The external noise source is a single entity and can't be "extracted" or converted into two uncorrelated noise sources. \$\endgroup\$ – Andy aka Jul 25 '20 at 7:50
  • 1
    \$\begingroup\$ The answers are no, no, and no, in that order. External noise sources will not correlate out. \$\endgroup\$ – Neil_UK Jul 25 '20 at 9:09
  • \$\begingroup\$ @KJ7LNW are you done with this question now? I ask because I see that you've not accepted answers to any of your previous questions and it's something that ought to be done as a rule because it then indicates to others coming along that the accepted answer has worked for you. It's also something that repays the effort shown by others in helping you out. You haven't been a member very long so you might not be aware of this. Of course, some questions may not have produced suitable answers and therefore you cannot accept an answer but, you should pursue answers that show signs of being acceptable \$\endgroup\$ – Andy aka Jul 27 '20 at 22:59
  • \$\begingroup\$ @Andyaka, thanksfor the tip! I didn't know about the accept button, will click through my existing answers. \$\endgroup\$ – KJ7LNW Jul 28 '20 at 0:27
3
\$\begingroup\$

With 2 LNAs, would any of these be useful to reduce the correlation of external noise sources?

No. If your noise has already coupled into the input of the LNA(s), then it helps nothing.

The mounting of the LNA generally has little effect at all - the amount of electromagnetic noise picked up depends on the length of whatever picks up the noise in parallel to the electric field (or perpendicular to the magnetic field). LNAs are small.

In fact, that mounting method just makes sure you build a "noise sensor" in two orthogonal directions. It sounds more like a method to pick up more noise, than less.

Also, as per your previous question: Two LNAs in parallel are not what you usually do. You use one LNA, to dominate the system noise figure. That makes shielding it easy, too: it's but a few components you really need to shield.

\$\endgroup\$
4
\$\begingroup\$

In general the answer is no but....

If you know the frequency of your wanted signal, you can delay the combined signal+noise to produce the inverse of the original wanted signal. If you then subtract the original signal from the delayed version, you get a 2:1 amplitude improvement in signal.

However, because the noise is presumed to be uncorrelated (and won't therefore "merge" constructively to a delayed version of itself), you will get an SNR improvement. Consider three delay stages like this: -

enter image description here

I have two nodes labelled Vraw and Vout and here are the waveforms: -

enter image description here

The red trace is Vraw and the blue trace is Vout. As you can see, there are some noise improvements. If I tripled the stages I get a further improvement thus: -

enter image description here

\$\endgroup\$
6
  • 2
    \$\begingroup\$ heh. I like how you've build a FIR here :) It has a sinc³(f) shape, though, and that might or might not be what you're looking for, right? \$\endgroup\$ – Marcus Müller Jul 25 '20 at 9:40
  • 2
    \$\begingroup\$ Who knows but it's worth a minor consideration. I'm sure there might be better ways to do it with delay stages - any thoughts @Marcus? \$\endgroup\$ – Andy aka Jul 25 '20 at 9:46
  • 2
    \$\begingroup\$ It is defintely cool! Imagine a simple tapped delay line model: input -> delay1 -> delay2->delay3->delay4. Tap into the input signal and the output of each delay (this might require signal amplfication to combat power spliting at every stage, but lets act as if we can "measure" the signal at each point); amplify each tap differently: input·(-1), ignore delay1 output, delay2 output·2.5, ignore delay3, delay4 output·(-1). Sum the result of the amplified tappings. \$\endgroup\$ – Marcus Müller Jul 25 '20 at 10:16
  • \$\begingroup\$ I'm no expert in this area but i think I've seen what you say i.e. a tapped delay line version of the above. Many years ago. \$\endgroup\$ – Andy aka Jul 25 '20 at 10:19
  • 2
    \$\begingroup\$ (also, if you're really just after a single tone, a single-tap feedback filter would do just as well: output(t) = (1-a)·output(t-T/2)+a·input(t) , but you'll be building a resonant thing and that can be fun if you didn't intend it to be an oscillator) \$\endgroup\$ – Marcus Müller Jul 25 '20 at 10:19
2
\$\begingroup\$

Provide heavily_filtered VDD to the LNAs. That reduces one source of external noise.

And then use a copper or steel shield atop the two LNAs, to exclude electric and magnetic effects.

In general the direction of interfering fields will not be known, so simply orienting the two LNAs in opposite directions will not, in the general case, suffice.

However, I was told, as a much younger engineer, that magnetic fields could not be shielded.

That is wrong.

The induced voltages resulting from changing magnetic fields are simply the displacement currents that attempt to provide nulling charges in response to electric fields.

Thus all the concepts of skin effect become useful in attenuating "magnetic" fields.

About 30 years ago, needing to develop a data acquisition system with 15_bit dynamic range, with about 30MHz bandwidth in the opamps, I took an afternoon to evaluate how to attenuate high frequency wire_wire coupling in the form of 1" square loops.

And I used a commonly available source of aluminum foil: chewing gum wrappers.

The test waveform was 20 nanosecond Trise and Tfall square waves, perhaps at 1MHz. Driving a leaded 51_ohm resistor, leads in 1" square loop and soldered onto Coax cable.

I built two of these 1" loops, to be Transmitter and Receiver/

With 1 volt pulses driving the Transmitter loop (thus 1/51 = 20 milliAmps in 20 nanoSeconds), I measured 0.08 volts in the Receiver loop, those two loops held about 1/16" part.

And after wrapping the foil between a folded sheet of paper, to prevent any electrical contact, I slowly slide the "shield" between the TX and RX loops.

Result? about 30dB drop in amplitude, or about 3 milliVolts on the TEk7904 scope's 7A26 50Ohm Rin 400MHz bandwidth plugin.

Do your own experiments, but think about metallic shields, to greatly attenuate both electric and "magnetic" interference.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.