RF Level Detection

[Rickomatic]

Hey all,

I'm a master's student currently working on a 5 month internship. My goal is to develop a system that can detect
fluid levels in tubes in a non-invasive manner.
Because the tubes are not always transparant optical level sensing does not work all the time and I have looked
at other detection methods. One of the methods that seems most promising is Electrical Impedance Spectroscopy (EIS).
During my initial experiments I have found that frequencies between 1 to 3 GHz gave promising results. I could succesfully
detect the interface between the air and the liquid (water) in plastic tubes. Also I could detect the difference between oil
and water in the tubes.

My initial experiments were done with a Rigol digital spectrum analyzer which I cannot use for the whole duration of my
internship. So I am looking for a way to replace it...

I did a google search and found the HackRF One which looked promising (10MHz - 10GHz) but this appears to be half-duplex
(I cannot send RF and receive at the same time). Also there are quite a few chip sets for SDR and similar applications but I
don't have the knowledge to judge which one is right.

Is there an RF buff out there who knows a good solution for my needs (blog posts, YT video, etc)?

[Yansi]

How are you supposed to do the job if they can't give you the appropriate equipment?

[hendorog]

LimeSDR is full duplex and 3.8GHz IIRC.

Here are some examples where it has been used as a basic VNA, sounds like that is what you are doing.

https://discourse.myriadrf.org/t/vector-network-analyzer/1038


If you are doing single frequency measurements then you just need something you can get an IQ data stream out of, and a signal generator which can be locked to it. If you are sweeping frequencies then it is harder, as the LO for the SA and the RF source won't keep their phase relationship over multiple sweeps.

Since you said you used a digital SA then I guess you only need a single frequency measurement or else I expect you would have hit that issue?

[cdev]

This seems like exactly the kind of application the "tzcam" is being marketed for.

There is a thread here on it.

https://www.eevblog.com/forum/thermal-imaging/ever-heard-of-the-tzcam/

 Its a millimeter wave imaging device.

[Kilo Tango]

Here is a possible detector :

http://www.dl5neg.de/diodesensor/diodesensor.html

Ken

[cdev]

Is there any reason "RF" has to be involved, besides your conclusion that that's the way you see as being the best way?

You should give us a bit more data, like how much detail is needed for your application, do you simply need to know if he fluid level is above or below some arbitrary value, and so on.

There are likely to be other approaches that work better than RF detection.

What kind of fluid is it? Does it conduct electricity? How massive is it, how viscous?

Watch out that you don't make something that's more complicated than necessary, when something that's simple would suffice.

Ultimately, that's the determination you should make, the most likely to be trouble-free solution.

Not the one that you think looks impressive and "high tech" on the CV but which is ultimately impractical.

[Kalvin]

Alternative to RF (if you have not already looked at these):

Ultrasonic:
http://www.ti.com/sensing-products/ultrasonic-sensing/overview.html

Capacitive:
http://www.ti.com/tool/TIDA-00317

The ultrasonic sensing works even with metallic tubes, and the ultrasonic measurement can also measure fluid flow rate.

Disclaimer: I have not used these techniques, so I cannot unfortunately help you more on these, or determine the suitability for your application.

[TheUnnamedNewbie]

Depending on the precision, accuracy and noise levels involved, there are a few very easy to use (at least should be easy if you are a masters student) power detectors from people like Analog Devices. I've used the ADL5902 with great succes in the past, and designing the board took me a weekend or so? I can even share some board files if you want.

Keep in mind when looking for detectors that there is a difference between a RMS detector and a level detector. Many detectors just measure voltage, and then use the assumption that you are dealing with a 50 ohm system to turn that into power.