Soil moisture data issues

Post Reply
mmelchiori
Posts: 3
Joined: Fri Oct 05, 2018 12:13 pm
Company: DW Informatica S.p.A

Soil moisture data issues

Post by mmelchiori » Fri Oct 05, 2018 3:48 pm

Hi everyone,

we are experiencing an issue regarding soil moisture data interpretation.

Some details about the system hare the following:
- The Smart Agriculture device we are using is a Plug&Sense! SA LoraWan EU
- The soil moisture sensors we plug into the device are P&S! SOCKET B, P&S! SOCKET C and P&S! SOCKET E; we inserted the probes B and E in a watered soil (humus type) at two different depths, and the probe C in a dry soil (humus type).
- The device has been configured using PCS
- The device communicates with a Kerlink Wirnet Gateway using LoRa protocol

The hexadecimal payload we receive is:
- 31273464a60024f445a7d9671346990000a741a84e6a04464acdccbe414c00a534424d7aacc647

and the frequencies measured by the sensors are the following:
- Dry soil (5 cm deep) - P&S! SOCKET C: 9433,9619140625 Hz
- Watered soil (10 cm deep) - P&S! SOCKET B: 7812.5 Hz
- Watered soil (20 cm deep) - P&S! SOCKET E: 8474.576171875 Hz

Then we convert Hz into cbar using the following formula (found in the documentation):
- cbar = (150940 - 19.74 * f / (2.8875 * f) - 137.5)

obtaining the following values:
- Dry soil (5 cm deep): -1.3019342332178918 cbar
- Watered soil (10 cm deep): -0.14623506045506812 cbar
- Watered soil (20 cm deep): -0.6718547653589503 cbar

not consistent with the values we expect (cbar >10 for dry soil and so on).

So we are not able to find the error: based on your experience, are sensors measuring wrong data? Is there a problem in data conversion? Is the formula we use for calculating cbar not suitable for the frequency values measured by the sensors?

Thank you so much for your support

libelium-dev
Posts: 27967
Joined: Mon Sep 28, 2009 1:06 pm

Re: Soil moisture data issues

Post by libelium-dev » Mon Oct 08, 2018 11:22 am

Hi,

That formula is for the sensor we used in that moment. For another watermark sensor, the formula will be similar but you have to follow the same process to calibrate your sensor. For example, keep in mind that the maximum Hz for our sensor was around 7500Hz and you are getting values higher than 7500Hz so it is normal that the values are negative.

Regards

mmelchiori
Posts: 3
Joined: Fri Oct 05, 2018 12:13 pm
Company: DW Informatica S.p.A

Re: Soil moisture data issues

Post by mmelchiori » Wed Oct 10, 2018 3:11 pm

Hello,
therefore should I calibrate the sensors ? Which is the process to calibrate the sensors ?

thanks

best regards

libelium-dev
Posts: 27967
Joined: Mon Sep 28, 2009 1:06 pm

Re: Soil moisture data issues

Post by libelium-dev » Thu Oct 11, 2018 11:02 am

Hi,

You should follow a processes like in the guide but with your sensor. The maximum frequency is get when the sensor is immerse in water and the minimum when it is completely dry.

Regards

mmelchiori
Posts: 3
Joined: Fri Oct 05, 2018 12:13 pm
Company: DW Informatica S.p.A

Re: Soil moisture data issues

Post by mmelchiori » Mon Oct 15, 2018 6:34 pm

Hi,
sorry but I didn't understand how to calibrate the sensors. The moisture sensor manufacturer is Watermark, I've used PCS to configure the Waspmote Plug & Sense, therefore I've just downloaded the binary code generated by Libelium platform. I've not written any code into the device !
Now, since I just receive a frequency value measured by the device through LoRaWAN, how can I do the calibration ?

Thanks

best regards

libelium-dev
Posts: 27967
Joined: Mon Sep 28, 2009 1:06 pm

Re: Soil moisture data issues

Post by libelium-dev » Tue Oct 16, 2018 9:04 am

Hi,

The PCS send data as Hz and it isn't possible send data as cbar. So if you want to know cbar you should perform field test to know the conversion formula of each sensor and be able to convert from Hz to cbar but it have to be done externally, it can't be done in the node using the PCS.

Regards

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest