Author Topic: rcwl-0516 microwave radar sensor (design issue)  (Read 182 times)

0 Members and 1 Guest are viewing this topic.

Offline fedimakni

  • Contributor
  • Posts: 36
  • Country: tn
rcwl-0516 microwave radar sensor (design issue)
« on: June 16, 2021, 03:40:56 am »

I came across the rcwl-0516 microwave radar sensor and i wanted to build the same device.
While doing some research on the internet i found this page which mention a lot of details about the pcb and as the chip used in this module (RCWL-9196) is not available i tried to build one using BISS0001 which is an alternative as mentioned in the link.
I almost copied the same design with the same length of track the same components and same positions and design the pcb by myself but i am still not able to make the device work. I don't even know if the pcb is 2L or 4L. the one i made was 2L, here's a link of my project
Did anyone tried this one before?

How was your experience. could you assist me with some advice?
Thanks a lot for your help.   

Offline evb149

  • Super Contributor
  • ***
  • Posts: 1843
  • Country: us
Re: rcwl-0516 microwave radar sensor (design issue)
« Reply #1 on: June 16, 2021, 06:04:30 am »
It is not likely that you will be able to simply swap the design to a new PCB, use a new transistor and set of RF-relevant parts and just have it work.

Even using the original model parts simply on a different PCB of the same exact PCB design but made by a different PCB factory might be unlikely to work without re-tuning.

An oscillator has to operate in a very narrow region is not a region of stable operation but is well defined enough that you can still get relatively "stable oscillation" arising from the instability because it should not be too much affected by minor changes in power supply voltage, operating DC current, temperature change, humidity, etc. etc.

It is really a narrow balance and every change of circuit AC / DC parameter and operating condition like temperature will affect the oscillation somewhat.

Oscillators like this one can depend on the RF AC characteristics of the PCB very much --
* the exact dielectric constant of the prepreg vs. frequency
* the exact dielectric constant of the core fiberglass vs. frequency
* the exact dielectric constant of the solder mask vs. frequency and its thickness over the trace profile
* the exact prepreg / core fiberglass thickness over the board.
* the weave pattern of the fiberglass in the outer layers of the board vs. the tracks.
* the drill diameter and exact position of the vias relative to other copper  if they carry or affect RF signals
* the DC current level in the transistor
* the DC operating voltage of the transistor
* the parasitic capacitance and inductance values of the PCB pads and tracks for RF circuits.
* the particular transistor's S-parameters and small signal model vs. frequency in the regions of its operation.

In short UHF oscillator design involves either modeling all such things to some degree of approximation and
manually tuning the circuit to operate in just the right region of instability to have just enough reliable positive feedback gain to counter the circuit losses and oscillate at a suitable frequency / amplitude.
Every one of the above parameters may matter in whether the oscillation happens at all, and if so what
the frequency / amplitude and general stability vs. parameter it will have.

If you're interested in learning about these things I suggest you study oscillator theory and also SPICE circuit modeling and try to make some oscillators of various types and frequencies in a circuit simulator and by
manual analysis.

Once you understand the basics you'll have a better idea as to whether a particular transistor could oscillate
with a given set of impedances vs. frequency attached to its ports and at a given DC voltage / bias current / temperature.  To know about a particular kind of transistor you should have a good model for it and maybe yours will have that or not depending on its usual intended purpose and how well documented it is.

Even then it will take some experimentation.  And ideally you'd have a spectrum analyzer, RF VNA, variable bench power supply, and possibly an assortment of PCBs and transistors of the same type to see how variable / repeatable the performance is vs. different similar circuits.

Without much of the above... yeah maybe you could get it to oscillate somehow somewhere by playing around
but to get it to also operate well as a detector is even less likely.

I would try repeating the circuit in a custom variation at a much lower frequency e.g. VHF until you better understand the process and have a better setup then try for higher frequency / performance.

If you're a student and can get access to ADS or something that way then such may be helpful since there will be
oscillator and receiver and mixer tutorials and so forth you could learn from with it.
But a certain amount could be done with SPICE or Qucs-s or something like that.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo