Hydros-21 Depth Temp Compensation

Hi Dan,

We recently did some internal testing of this issue and will post results soon. We don’t have a “fix”, but we did measure this behavior to guide our own interpretation of data.

Scott

Hi Dan,

I’m curious whether it’s possible to identify Gen 1 vs Gen 2 Hydros21; are they labeled? I’m wondering what I’ve got.

Thanks,

Matt

Hi @dan-wachusett. I am about to order 10 x Hydros 21 units from METER Group. I have been told that they will be Gen 1, but Gen 2 will be released in a few weeks’ time. I’m trying to get some details from METER Group about the upgrade and whether it would be worthwhile waiting for a couple of weeks to get the Gen 2.

I also see that the Gen 2 is not working with modular sensors yet. Any idea how long it will be before it can be used?

James

We recently received a Gen 2 Hyrdos 21 CTD sensor from Meter Group and have confirmed that it works just fine with the Mayfly Data Logger, and using our Modular Sensors library, with just a very small adjustment made to the library.  We’re in the process of updating that now, so you should be good to order them and use them when they arrive.

Thanks @shicks, great to know. Would you recommend that I push back on them for Gen-2 sensors, even if they take a bit longer to get here? It seems silly to rush things through and end up with an inferior sensor.

James

We just got the sensor and haven’t done any comparison tests alongside the Gen 1 sensor to see if there’s a difference, so I can’t comment yet on anything data-wise, we just know that there’s not a communication issue with the Mayfly now that we’ve adjusted the timing of the commands in our library.

There’s an example sketch in the Modular Sensors GitHub Repository with a calculated variable correction for water temperature (really it’s density, so it’s called rho and density in the sketch) that uses data from a temperature sensor. The same sketch has a calculated barometric correction.

https://github.com/EnviroDIY/ModularSensors/tree/master/examples/baro_rho_correction

The Hydros-21 sensor already does internal temperature compensation of the depth data, so you shouldn’t have to provide your own compensation.  The sensor outputs the depth readings in millimeters of water above the sensor, so you don’t even have access to the raw pressure readings like you might with a basic pressure transducer, so you’d have to do any corrections on the sensor’s already-corrected depth output.  You could record temperature with a separate external temperature sensor, but we’ve found that the temperature reported by the CTD sensor is accurate, except in cases of extreme rapid temperature changes, in which case it takes anywhere from 5 to 15 minutes for the CTD temperature to match a DS18b20 sensor in the same water (like when we poured refrigerated water on a room-temperature sensor).  The new Generation 2 Hydros-12 sensor had a quicker response to the temperature change than the old generation.

We do see a variability of about 5mm to 10mm in the depth reading if the sensor is in a constant depth of water but the water temperature changes from about 10C to 30C over the course of a day.  However, that variability depends on the individual sensor, since we tested 10 different sensors all from the same production lot and found that some were more constant across temperature changes than others.  So if micro-scale variations in depth or conductivity matter to your experiment, I’d recommend doing some controlled testing with your sensor in a lab environment where you can change the temperature of the water while recording sensor output to get a profile of how your individual sensor responds to changing temperatures.  And since the conductivity readings of the Hydros-21 sensors are specific conductivity, then temperature sensitivity affects those readings as well, so check your conductivity readings over a range of temperatures as well.  Some of the sensors are more accurate than others, but all of them have been within the design specs of the manufacturer’s stated accuracy levels, but it’s important to find out how precise your sensor is since it varies for each one.

Hi @dan-wachusett – I’d be interested to hear how your measurements work out.

This is what I did for acceptance testing (and characterizing) on a number of different types of depth sensors for a 2013 project – https://www.envirodiy.org/measuring-low-water-in-streams-accurately/

The update since then is the hydrologists I work with use the LT-500 for reliability and accuracy, and also been moving over from the Onset U30 to using the Mayfly.

I wonder, has anybody received a specification for the Hydros-21 Gen 2? .

Typically I find that manufactures create internal releases (generation 2?) as an internal cost reduction, with the same specification. Though every manufacturer is unique, so wondering what Metergroup changes might be :)).    Insitu and Keller (and others)  have different lines of depth sensors for different measurement accuracy and methods.  The challenge for any manufacturer is the basic pressure sensor is similar – a pressure dependent resistor – piezo resistor. Temperature dependency of resistors is well studied, but challenging to work with when bumping up against a specific resistor technology’s limits – so different manufactures have different signal processing and manufacturing calibration for better linearity of the pressure sensor.