Author Topic: A "new" kind of dipole antenna?  (Read 1454 times)

WhiteHare

  • Hero Member
  • *****
  • Posts: 1300
  • Country: us
A "new" kind of dipole antenna?
« on: October 28, 2020, 11:30:27 PM »
Attached is a photo of a PCB antenna that I thought you guys might find it interesting.    It's from Version 2 of a flume water meter monitor.  Looks like it's maybe (?) some kind of dipole antenna, though up close the construction is fairly intricate.  I have close-up photos if anyone is interested in having a closer look or would like to comment.

By the way, the Version 1 Flume also used HopeRF RF69 and even an atmega328p MCU, so you all would have been completely at home with it.  It even communicated with an ESP8266 gateway node, much as what some of us here were doing.   For Version 2 they kept the RF69 but switched to an ATSAMD20.  I wonder why?  Aside from that, the main differences between version 1 and 2 were the use of this new, fancier antenna, and a simplification of the moistureproof enclosure (no screws holding the cover on Version 2, as compared to 12 self-tappers holding on the lid of Version 1).

I didn't own a version 1, but I very much like my Flume version 2.  It seems like a simple, yet very nicely engineered IoT device that does pretty much exactly what I want it to: log water consumption down to the nearest 0.01 gallons for every minute of every day, and keep all that data forever, for review whenever I like.  Nifty!

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: A "new" kind of dipole antenna?
« Reply #1 on: October 29, 2020, 10:05:10 AM »
Interesting device, probably a dipole of some kind. Is there any FCC ID on it?

Whats the small QFN chip under the antenna?
And can you tell the tech behind the magnetic field detection? I would guess one of the SOT23 chips is a magnetic sensor but I see some perpendicular through hole things mounted on the bottom side, maybe that's some kind of "radar".

My guess is the SAMD20 has to do with the "machine learning and fancy algorithm" (aka marketing mumbo jumbo for "arduino code") which can turn numbers way faster than an AVR.

I wonder if they use the RFM69 library :-)

WhiteHare

  • Hero Member
  • *****
  • Posts: 1300
  • Country: us
Re: A "new" kind of dipole antenna?
« Reply #2 on: October 29, 2020, 04:27:27 PM »
Presumably, it's some kind of bow-tie antenna:  https://iopscience.iop.org/article/10.1088/1742-6596/450/1/012048/pdf

I'm guessing the QFN is some kind of magnetic sensor, attached is a photo of it.

I wonder if they use the RFM69 library :-)

I wouldn't be surprised if their Version 1 (which you can still buy on Amazon) used your dual optiboot for OTA updates.  If so, maybe it will give you some satisfaction.
« Last Edit: October 29, 2020, 05:44:25 PM by Felix »

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: A "new" kind of dipole antenna?
« Reply #3 on: October 29, 2020, 05:52:41 PM »
A lot of antennas are derivations and variations of dipoles.
Yes I forgot to ask about the SOIC package, is that a flash memory? Then this looks a lot like a Moteino M0 :-)
That QFN marking is not helpful, the many signals to it would be strange for a magnetic sensor unless it has a serial interface (SPI, maybe I2C).
As much as I've tried to detect a magnetic field on my water meter, it was futile. I had to resort to a photoreflective sensor to detect a spinner that yields a high spins/gallon resolution.

WhiteHare

  • Hero Member
  • *****
  • Posts: 1300
  • Country: us
Re: A "new" kind of dipole antenna?
« Reply #4 on: October 29, 2020, 09:52:38 PM »

Yes I forgot to ask about the SOIC package, is that a flash memory?

Close-up attached.

WhiteHare

  • Hero Member
  • *****
  • Posts: 1300
  • Country: us
Re: A "new" kind of dipole antenna?
« Reply #5 on: October 30, 2020, 04:18:34 PM »
At any rate, here's an antenna that might be worth trying:  https://agupubs.onlinelibrary.wiley.com/doi/10.1002/2016RS005973

It's compact, can be etched to a PCB, and yet (if I understand it correctly) it's also omnidirectional.

Cheers!