RF and MW receivers' sensitivity

[Emperor42]

I'm currently working on a novel RF and MW sensor and I wanted to know whether there is a paper that shows the current very best receiver sensitivity for a range of wavelengths in the RF-MW range in dBm preferably so I can directly compare to my device.

 

[skeptic2]

What is the bandwidth of your receiver?

https://www.highfrequencyelectronic...dth&catid=94:2014-06-june-articles&Itemid=189

 

[Averagesupernova]

Some years ago I think you could expect a half a microvolt at the antenna connector could get you about 20 db sinad on an HF amateur receiver. Likely has improved since.
-
https://en.m.wikipedia.org/wiki/SINAD

 

[berkeman]

Here is a nice paper about the sensitivity and noise issues for Radio Astronomy receivers:

https://events.asiaa.sinica.edu.tw/school/20160815/talk/sransom0818.pdf

 

[Baluncore]

In a screened room the sensitivity is limited by the noise floor of the front-end RF preamplifier, probably a MOSFET. But in the real world the cosmic and atmospheric noise = QRN, and noise from switching power supplies etc = QRM, will dominate the receiver noise.
https://en.wikipedia.org/wiki/Atmospheric_noise

There are ways of digging signals out of the spectrum analyser grass.
1. What is the intended application where you need a very high sensitivity?
2. What are the characteristics of the signal modulation?
3. Will you have a narrow band synchronous detector to reduce noise?
4. Could you use an A to D converter followed by an FFT to get 24 dB of conversion gain?
5. Could you power-spectrum-accumulation to stack maybe another 20 dB deeper?

 

[sophiecentaur]

I'm currently working on a novel RF and MW sensor .

What is the bandwidth of your receiver?

The use of the word "sensor" could imply something other than an audio receiver so the Bandwidth has to be known before any answer can be given.
Perhaps the OP could reply with that necessary piece of information.
However, the question may have just been posted out of 'curiosity', in which case the "half microvolt" figure for an audio MF receiver could be near enough.

 

[Emperor42]

The use of the word "sensor" could imply something other than an audio receiver so the Bandwidth has to be known before any answer can be given.
Perhaps the OP could reply with that necessary piece of information.
However, the question may have just been posted out of 'curiosity', in which case the "half microvolt" figure for an audio MF receiver could be near enough.

The bandwidth is roughly sub-kilohertz so around 500Hz.

 

[f95toli]

At which frequency?
There are sensors with single photon sensitivity for frequencies above about 4 GHz, but of course these only work at very low temperatures..
Also, the word "sensor" is a bit ambiguous. In applications where you need good sensitivity the easiest thing to do is (usually) to put a good low noise amplifier before he actual detector; that way you can even use standard square-law detector diodes with very small signals.

I routinely work with MW signals (2-8 GHz) with a level of about -120 dBm (and sometimes way below that) and most of our setup is made up of commercially available equipment.

 

[Emperor42]

Thanks for the input guys. I'm trying to get more info, but I'm a quantum Physicist by trade so I'm trying to wrap my head around how to compare the sensitivity of our sensor (which measures it as an oscillating electromagnetic field at a single point in space), which is measured in Tesla/sqrt(Hz). I think that I can convert it into an amplifier by attaching it to a coil or array, which would give me a sensitivity of -164dBm at 12.6GHz and up to -200dBm for RF (around 100MHz). So I wanted to compare, but I guess its a lot more complicated than I thought. As for our bandwidth given that it is a quantum sensor the bandwidth changes depending on the intensity of the signal so it could go down to as low as nHz.

 

[f95toli]

What does "quantum sensor" mean in this context?
The fact that you measure Tesla/sqrt(Hz) would suggest some sort of SQUID based sensor or perhaps a flux qubit.
If that is the case there are a bunch of articles written about using SQUIDs/qubits as sensors. Just put "microwave quantum optics" into Google scholar.

 

[Averagesupernova]

MW to me means lower frequencies:
https://en.m.wikipedia.org/wiki/Medium_wave
So my previous post about HF receivers probably means very little knowing you are interested in microwave.

 

[sophiecentaur]

MW to me means lower frequencies:
https://en.m.wikipedia.org/wiki/Medium_wave
So my previous post about HF receivers probably means very little knowing you are interested in microwave.

Me too. I read MW as MF. I wish people would declare all their variables, the first time they use them. MW is not a good term to use when μ Wave is meant. You can't even use μW because that's microWatts.

 

[tech99]

Thanks for the input guys. I'm trying to get more info, but I'm a quantum Physicist by trade so I'm trying to wrap my head around how to compare the sensitivity of our sensor (which measures it as an oscillating electromagnetic field at a single point in space), which is measured in Tesla/sqrt(Hz). I think that I can convert it into an amplifier by attaching it to a coil or array, which would give me a sensitivity of -164dBm at 12.6GHz and up to -200dBm for RF (around 100MHz). So I wanted to compare, but I guess its a lot more complicated than I thought. As for our bandwidth given that it is a quantum sensor the bandwidth changes depending on the intensity of the signal so it could go down to as low as nHz.

Thermal Noise Power from a resistor at 300K is -174 dBm/Hz. So for 500Hz you have -174 + 10 log 500 = -147 dBm.
If you have -164dBm in 500Hz, that corresponds to a noise temperature of 6K.

 

[Baluncore]

Fundamentally, the “sensitivity” of today's microwave detectors is so high that it is swamped by thermal noise. For that reason, it is the ease and degree of cooling possible that makes the difference between detectors.

Bandwidth is not critical so long as it is maintained constant during comparisons, or is corrected per √Hz when applicable.

Frequency conversion mixers are significantly noisier than LNAs. Once received, it is necessary to progressively amplify a signal ahead of the rise in mixer and thermal noise as the signal leaves the cryogenic receiver environment and approaches 300K. How and in what form will your signal climb the steps of that ladder?

FETs have a lower inherent noise than junction transistors. How many, and which parameters determine the fundamental noise floor of your proposed device? How do they compare with other devices and materials?

What is the lowest temperature at which your detector will operate, or when does it cease to operate because part of it becomes superconducting. Does your RF detector need to be superconducting?

What is the shortest operating wavelength of your device? Can it be printed as an array of thermal detectors for FLIR imaging applications?

The critical temperature of materials such as Si or GaAs will decide which must be used to make FETs for use at lower temperatures. The question or competition becomes how cold it can be, not how sensitive it is.

 

[marcusl]

Thanks for the input guys. I'm trying to get more info, but I'm a quantum Physicist by trade so I'm trying to wrap my head around how to compare the sensitivity of our sensor (which measures it as an oscillating electromagnetic field at a single point in space), which is measured in Tesla/sqrt(Hz). I think that I can convert it into an amplifier by attaching it to a coil or array, which would give me a sensitivity of -164dBm at 12.6GHz and up to -200dBm for RF (around 100MHz). So I wanted to compare, but I guess its a lot more complicated than I thought. As for our bandwidth given that it is a quantum sensor the bandwidth changes depending on the intensity of the signal so it could go down to as low as nHz.

If you have a SQUID, then you are measuring flux through area (granted it may be a small area) rather than field at a single point in space. How exactly are you planning to convert that to dBm? Sensitivity will depend critically on your conversion equipment (transducer), on impedance matching, and on the noise characteristics of your transducer. At high frequencies, parasitics, skin effect and other phenomena may also be important.
"Flux transformers" are widely used with SQUID's and there is an extensive literature extending over 45 years that you can look at.