Greetings from a newbie to this site:
Last year I constructed an ultrasonic snow gauge. The object was to make one that was completely wire free because several identical units might be placed in locations with no power within a 5 mile radius of the HQ building. I used a pair of XBee-PRO PKG RS-232 RF Modem radios (short range only for the moment) and powered it with solar panels and batteries. Due to the limitations of battery power, I can’t have the sensor or radio modem on full time.
Instead, I use a Parallax BASIC Stamp PLC to contol the system.
(http://www.parallax.com/Store/Microcontrollers/BASICStampModules/tabid/134/txtSearch/plc/List/1/ProductID/290/Default.aspx?SortField=ProductName%2CProductName)
The PLC turns on the sensor, takes multiple readings, averages them, turns on the radio modem, send the readings, and then shuts off the radio and sensor and enters “sleep mode” for about three hours. The way to tailor the load to the battery and solar array is to vary the time between readings - the longer the “sleep”, the less power per day is consumed.
Before 2007-2008 winter set in here the system was up and running. It could accurately and repeatedly measure the distance to the plywood measurement “ground” surface. It could accurately measure the height of my “test snow” - piles of cardboard boxes. However, when the snow arrived, the system could not measure snow depth. It didn’t matter if the snow was soft and fluffy or had a crusty surface, the sensor would respond as if all the ping had been absorbed by the snow (showed maximum range) . It used an EMS sensor (http://emssensors.com/ultrapics/s1225sp.htm), and since this sensor does not have an adjustable gain control, there was no way to experiment with boosting the gain.
This summer (2008) I decided to try some modifications. I used a SensComp Mini-AE sensor (http://www.senscomp.com/specs/Mini-A%20PB%20spec.pdf) which has adjustable gain as well as range zeroing features (which I won’t use). The sensor is mounted in a Davis weatherstation pagoda. This was not an original thought - I borrowed the idea from another snow sensor (http://voksenlia.net/met/snow.php), so I hope that it will adequately function for me. At least I have some confidence that this sensor will work, and its relatively cheap and water resistant when in a housing. It is also internally temperature compensated, so I don’t have to program a speed-of-sound routine. Sheltering it in a pagoda should keep it at ambient temperature without undo sun heating.
Today I began system calibration. I know the the PLC is translating the analog voltage from the sensor correctly since it provides an ASCII output of the analog voltage appearing on the sensor lead - good to 1 millivolt. The PLC will also output the snow depth in inches by subtracting the current reading from the sensor reading of a base measurement table (no snow), thus providing the height of the snow surface from the ground. IN addition, it will also send the current battery voltage so that I can keep an eye on “system health” by monitoring the state of the power supply. If the battery voltage dips too low after several cloudy days, the charge controller will turn off power to the PLC until the batteries are recharged. The PLC will then restart and reboot. The charge controller will also limit the charge rate to a temperature compensated sfae level for the batteries ( two 12 volt 7 amp/hr gel acid batteries).
This system is a rough prototype for a device which could be part of a network of a hub PC storing data sent by several automated gauges in the field. The current XBee short range radios could be replaced by long range units (up to several miles range) which have the ability to be hardwired or programmed to send a unique ID with the data record. It would be easy to set up a net using identical systems and only changing the unique ID code on each unit.
At the receiving end I have an old PC (which also hosts our Davis WeatherLink software) on a LAN running a serial port monitoring program which simply dates all the incoming data transmissions from the snow gauge and maintains them in a log file. I am also running FTP host software on that PC so that any interested researcher can use a browser to download the data files containing snow depth data for future analysis.
Why do all this? I work with a group that is involved in endangered butterfly habitat restoration, and one of the unknowns in our world is the ability of butterfly eggs to “winter over.” It appears that butterfly eggs, laid on the stems of host plants may be protected by the accumulation of insulating snow. While there are volunteers out there reading manual snow rulers all winter in several locations, the use of automated snow gauges the use of might be helpful in assuring some level of measurement consistency.
If all of this works, the next step would be to begin to pick away at the costly items and try to drive the unit cost down to a reasonable level.
I hope some of the experiences here will be of help in other’s efforts to design ultrasonic snow gauges.
Warpdrive