Author Topic: Snow depth monitor  (Read 2283 times)

MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Snow depth monitor
« on: February 25, 2020, 10:42:19 AM »
I'm getting started putting together a Moteino based sensor to measure the snow depth in my back yard.  My idea is to put together a Moteino and a PowerShield interfaced to a LaserRangeFinder:  https://www.amazon.com/gp/product/B07L8B2FVK/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1.  The LRF would be on an arm sticking out from a tripod support that has solar cells, and a weather-proof box containing the electronics.  I'm hoping to power the system with SuperCap and solar.  Since the sensor is to be out in the snow, I plan to put the solar cell(s) vertically to prevent snow build-up on the solar cell(s).  At the moment the mounting is conceptual; I'll try to make some drawings that I can post another day.

Question 1:  I'm not particularly constrained size-wise, so I can use up to 3 110x92mm 6V solar cells to charge my SuperCap.  Each solar cell would charge the cap through a Schottky diode, and the 5.4V Super cap protected from overcharging with a 5.1V zener diode.  The solar cell(s) are 200ma; so worst case, the zener diode would be dissipating .9V @ 200ma(x3) = .54W.  Do I actually need a fancy charge controller as discussed in some other threads?
My thoughts are that using a MPPT charge controller is counter productive.  When the cap is discharged, you want maximum current at low voltage; when the cap is nearly charged, you want maximum voltage at whatever current you can get.  A max power point controller adjusts the voltage out to maximize the power, so probably not the highest voltage.  Anyway it seems like needless complexity unless the leakage through the zener diode is too high when not charging.

Question 2:  My sensor is peak current of 140ma, "average" current of 110ma at 5(+-0.5)V.  Their use case is some form of continuous measurement, but I would only sample at most every minute.  They don't specify the idle current.  I would like to switch on the power to the sensor shortly before taking a measurement, then turn it off again.  Will something like the circuit in the attached image work?  (Image from https://lowpowerlab.com/forum/moteino/using-moteino-r6-in-extremely-low-power-project/msg26879/#msg26879)  I would use the prototyping area on the PowerShield to build the PFET circuit (I have NDP6020P).
A possible alternative to the PFET circuit would be to use a TPL5110 BoB from LPL to control the power to the LRF; seems overkill.  I think I would need two digital outs from the Moteino, one to turn on the LRF through the Delay input, and one to turn off the LRF through the Done input.  Would the TPL5110 smoke if I feed it 5V from the PowerShield?

I haven't thought much about what to do if the capacitor discharges below the BOD of the Moteino.  I'm going to try to have enough solar power that I never reach that point.

I'll post pictures as I progress.



MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Re: Snow depth monitor
« Reply #1 on: February 25, 2020, 11:39:41 AM »
Here's my current low-power, low-precision, low-tech snow depth gauge.  Four foot stake, 6" into the ground.

It is not nearly geeky enough.  I'm guessing there's about 2.5 feet of snow out there.


Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: Snow depth monitor
« Reply #2 on: February 25, 2020, 02:03:41 PM »
How much precision do you need?
From always trying to find the simplest solution to something here's a random thought - I'm guessing you could just use some resistors wired with incrementally longer wires and back into the ADC to read incremental snow depths. The snow might act as the switch to each ADC pin. Easy thing to try out with some resistors (values totally up for experimentation) and some snow  ;). The downside of this is if it rainns on the stick giving false readings.
Or hey, install an IP camera with the PiGateway and have it send you a periodic snapshot of the stick to your email, lol  :o

Back to your project Qs now:
1) Only experimenting can tell what size solar cell you need. You can just add more cells in parallel if need be. And same goes for capacitors, add more if you need more juice.
Avoiding needless complexity and keeping things simple as you already suggested always works best. Ensure the zener is low leakage. I'm not sure I would not even bother with one though.

2) Just use a PFET directly driven, nothing else, to switch the HIGH side of your sensor, it would require switching inverted logic. Or if the sensor allows, use an NFET to switch the low side, positive logic. You would want to use the TPL to power ON everything not just the LRF. And no, 5v to it would not be a problem, but I would power the TPL through a diode to ensure it will pick up the 3.3V DONE signal.

« Last Edit: February 25, 2020, 02:12:00 PM by Felix »

john4444

  • Jr. Member
  • **
  • Posts: 71
  • Country: us
Re: Snow depth monitor
« Reply #3 on: February 25, 2020, 02:07:41 PM »
Hi MojaveTom,

Your solar-panel / zener  seems ok to me. But, I have not done what you describe so, I comment without any authority.
However, I suspect that you will have issues with switching the ground connection to the LDR.
It is easy for the internal protection diodes to become forward biased through the in / outputs.
You will probably need to switch the positive side of the power to get it to work reliably.

While I was typing this Felix responded.
He is a great source of info.
You can trust what he says.

Good Luck, John
John AE5HQ

Uncle Buzz

  • Full Member
  • ***
  • Posts: 146
  • Country: fr
Re: Snow depth monitor
« Reply #4 on: February 25, 2020, 05:31:45 PM »
From always trying to find the simplest solution to something here's a random thought (...)

Surprisingly, you haven't mentioned your SonarMote that you use to read water level ... This is a typical application for your product!

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: Snow depth monitor
« Reply #5 on: February 26, 2020, 09:32:37 AM »
UncleBuzz,

Ha, how did I forget about SonarMote  ;D
Well certainly worth a try. I wonder how show (of different consistencies, soft, wet, frozen) would reflect ultra sonic sound waves. But then that's also true for IR.
It would require the box to be well sealed, but could run on a LiPo for a long time. Could fail or be inconsistent if the snow is not really flat or starts thawing, but same is true for any device. However we're assuming this stick is placed where snow will be undisturbed so maybe snow will always just be flat regardless of its condition.

MojaveTom just posted a SonarMote 3D enclosure BTW, I linked it on the SonarMote page. I imagine this would be easier to weather seal.

One day when I find time I should redesign these enclosures to include channels for weather seals (flex filament works great for this purpose).

Uncle Buzz

  • Full Member
  • ***
  • Posts: 146
  • Country: fr
Re: Snow depth monitor
« Reply #6 on: February 26, 2020, 10:45:42 AM »
It would probably work :
https://www.geekstips.com/arduino-snow-depth-remote-sensing-with-ultrasonic-sensor/
https://novalynx.com/store/pc/260-705-Ultrasonic-Snow-Depth-Sensor-p1335.htm
https://www.weather.gov/grr/snowsensor

The main challenge is to protect all this electronic stuff from humidity since at least the transducers have to be exposed...

For random thought, variable capacitor or time-domain reflectometry could be fun and low power too...
« Last Edit: February 26, 2020, 10:50:11 AM by Uncle Buzz »

MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Re: Snow depth monitor
« Reply #7 on: February 26, 2020, 12:43:47 PM »
I chose this sensor because it is rated for outdoor use.  My reading of the specs for the HC-SR04 sensors indicated that they wouldn't like being exposed to moisture (but I can't find that spec today), and the environment can get to -10°F as it did this morning.  Sonic sensors are affected by air pressure (altitude, barometric pressure), temperature, humidity; so it can get complicated if you want to account for all those factors.  Altitude can be accounted for by calibrating the sensor in its final location; barometric pressure is probably not a factor since the variations are relatively small.  Temperature here varies from about -15°F to 40°F and the humidity often is over 90%, sometimes foggy during the winter when I would be measuring snow depth.

I guess the lack of environmental specs for the HC-SR04 was my biggest determining factor.  However, I'm willing to give it a try.  Since I have both sensors, I could try hooking my LRF to a SonarMote in a weather proof box.  I have only printed with PLA on my printer, which doesn't hold up well when exposed outdoors; so I would modify a commercial weather proof box.

Hmm, using a SonarMote as the basis for my sensor is an interesting possibility.  It has the 5V switched supply for the HC-SR04 which I could probably tap for my LRF.  I would only activate one sensor at a time to keep the 5V current low.

I found a price on Amazon for a sensor that looks similar to the Novalynx sensor:  ~$150.  Yikes.

Tom

i670

  • NewMember
  • *
  • Posts: 20
  • Country: au
Re: Snow depth monitor
« Reply #8 on: February 27, 2020, 12:19:31 AM »
I've used the JSN-SR04T sensor with more success then the HC-SR04, which definitely didn't like being exposed to the elements over a few months.  The version of the SR04T I bought on eBay appears to be sealed and doesn't have an exposed metal foil to corrode.

MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Re: Snow depth monitor
« Reply #9 on: February 27, 2020, 11:42:13 AM »
I've used the JSN-SR04T sensor with more success then the HC-SR04, which definitely didn't like being exposed to the elements over a few months.  The version of the SR04T I bought on eBay appears to be sealed and doesn't have an exposed metal foil to corrode.

All the specs I can find on this module say that the min operating temperature (°C) is -10.  It regularly gets to -23 here and has been recorded to as low as -40.  I like that the sensor comes with a 2.5m cable; that would make it easy to put out on an arm away from the electronics.

I suspect that the HC-SR04 won't last long in this environment, the sensor you reference would undoubtedly last longer.  Maybe the specified min temperature is just to ensure that the accuracy is within spec, not that the sensor will be damaged.  They are inexpensive enough that an experiment would be worthwhile.  With your experience, I don't think I will even try the HC-SR04 sensor.

Tom

i670

  • NewMember
  • *
  • Posts: 20
  • Country: au
Re: Snow depth monitor
« Reply #10 on: February 28, 2020, 04:12:21 PM »
Looks like there are various specs listed, with one down to -20C.  Likely pot luck how low it will function.  I would think your lipo would struggle, too.

MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Re: Snow depth monitor
« Reply #11 on: February 29, 2020, 10:52:30 AM »
Planning to use super cap and solar.  Just how much of each is part of the experiment.

Tom

MojaveTom

  • NewMember
  • *
  • Posts: 24
  • Country: us
Re: Snow depth monitor
« Reply #12 on: March 07, 2020, 11:08:23 AM »
Here's a cad drawing of what I'm building.  There could be solar cells on three sides, just not the one with the sensor arm sticking out.