Frequently Asked Questions (FAQ)
Accuracy of Drop Size Measurements
Test of the Disdrometer on Site
Calibration of the Disdrometer
Calibration of the Disdrometer Sensor
Calibration of the Disdrometer Processor
Packing the Disdrometer for Shipment
Length of Sensor Cable
Availability and Data Integrity
Mainly due to the fact, that the sensitivity of the sensor is somewhat dependent on the location of a drop impact on the sensitive surface of the styrofoam cone, the pulse amplitudes of drops of equal diameter will form a distribution around the average amplitude. The standard deviation of this distribution, transformed into drop diameters, is approximately +/- 5% if the drops are distributed evenly over the sensitive surface. The specified accuracy of a drop size measurement of +/- 5% of the measured drop diameter means, that the average measured diameter of a large number of drops of equal diameter, evenly distributed over the sensitive surface of the sensor will be within 5% of their actual diameter.
Distrometer sensors are calibrated at our laboratory using natural drops of known size, produced with a special apparatus designed for that purpose. Larger drops can be produced with a simple jet, but the height of fall must be such, that the drops will reach near terminal fall velocity. A simpler, yet very sensitive method is to make a comparison of rainfall amounts measured with the disdrometer and with an accurate rain gauge. An event with a continuous rain rate of between 1 and 10 mm/h lasting for several hours, such that evaporation from the rain gauge can be neglected and with light winds only, is best. If the disdrometer is in good condition and the total amount of rain is more than about 5 to 10 mm, the difference of the measured rain amounts should not be more than about 15%.
Also check the sensor for signs of corrosion on the metal surface under the preformed plastic foil. If water has entered the unit, the corrosion products that form may impair the free movement of the coils in the gap of the magnets.
A test for the processor RD-80 is to actuate the built in test signal by pushing the button marked “TEST” on the front panel of the processor with the sensor connected. While the button is depressed, LED No. 4 should light up. A processor RD-69 can be tested similarly. If it is connected to an ADA-90 analyser, LED No. 4 of the analyser should light up while the button for the test signal on the rear panel of the processor is depressed. If an AM-90 analyser is used, LED No. 7 should light up. Alternatively a square wave of 1 Volt pp with a frequency of 1000 Hz should appear at the output Ucompr of the processor RD-69.
As a full calibration of the disdrometer with natural water drops of many different sizes, falling at terminal velocities and distributed over the whole measuring range of the instrument is a very pains taking procedure, a calibration method was developed, which gives reliable results with a lesser amount of effort. With this method the sensor and the processor are considered separately. Originally a full calibration as described above was performed with several instruments previously adjusted using the same standard procedure. For every instrument the characteristic of drop size channel number versus drop diameter was measured.
Since the pulse shape of the sensor output is quite variable depending on the drop diameter and on the physical location of the impact of a drop on the sensitive surface of the sensor, the characteristic of drop diameter versus amplitude of the sensor signal was determined indirectly by measuring the characteristic of drop size channel versus amplitude of the sensor signal, using pulses from a pulse generator. With this characteristic it was possible to calculate the characteristic of the amplitude of the sensor output signal versus drop diameter. It was found, that the characteristics of all the sensors that were tested, were smooth and very similar. For this reason it was decided, that the calibration of a sensor with only one size of drops was sufficient.
The averaged characteristic of all the sensors that were tested was taken as standard sensor characteristic. In order to get the characteristic of drop size channel versus drop diameter of an instrument (sensor and processor), the standard sensor characteristic is combined with the individual characteristic of the processor under consideration measured with pulses from a pulse generator.
The sensor is tested together with a calibrated processor using real drops of about 0.8 mm diameter, falling at near terminal fall velocity. The drops are distributed evenly over the whole surface of the styrofoam cone. If the error of the measured drop diameter is more than 3%, the gain of the amplifier module in the sensor is adjusted to correct for the error and the drop test is repeated. As long as the mechanical setup of the sensor is not disturbed, the response of the sensor is very stable and recalibration is usually not necessary.
The main purpose of the processor is to reduce the very large dynamic range of the sensor signal such that it can be handled by the analyser. In the RD-80 processor, the analyser is part of the processor, in the case of the RD-69 processor, the analyser is a separate unit. The circuit of the processor is essentially a pulse amplifier with a nonlinear characteristic. For calibrating the processor a special test instrument is used to adjust several potentiometers of the processor circuit, which control different parts of the slope of the characteristic. After this, the characteristic is checked at 31 points using electrical pulses with a pulse shape similar to the output pulses of the sensor.
A table of the characteristic is given in the calibration certificate, which is supplied with every instrument. The circuit of the processor is rather stable and recalibration is usually not necessary. A rough test of the characteristic can be made easily at one point by pressing the test signal button on the processor cabinet and by observing the LED display of the RD-80 processor or the LED display of the analyser ADA-90 in the case of the RD-69 processor. While the button is depressed, LED Nr.4 should go on. For this test the sensor must be connected with the processor.
If it is necessary to ship the disdrometer, the following precautions should be considered:
- The sensor should always be shipped with the styrofoam cone taken off.
- Use a large enough box and use enough padding between and around the instruments.
- To prevent the electromechanical unit from being torne out of the sensor housing, fix a tape over the top of the sensor.
Cable lengths as much as 80 m have been used successfully between the sensor and the processor of the disdrometer . When using 80 m length the sensitivity of the disdrometer will be reduced by about 0.4% in drop-diameter.
In order not to have to disassemble the sensor, you may want to use an extension cable with a pair of water-tight connectors.
The type of cable used is: 4 - conductor shielded cable; the crossection of the conductors is 0.34 mm2 each. A suitable type of connector, compatible with the standard connector is:
AMPHENOL-TUCHEL Typ C091 31H005 100 2 5-pole male cable connector
AMPHENOL-TUCHEL Typ C091 31D005 100 2 5-pole female cable connector
Distromet Ltd provides suitable extension cables from stock. The length of the cable is 30 or 60 meters.
The program acquires rain drop data continuously and writes new data every minute to “current data file“ and in addition (optional ) to a second data file “current file copy“. This helps prevent data loss in case of power failure as there will always be a second data file available.
After a power failure, when power starts up again, this DISDRODATA is not self starting and needs operating.
We recommend to use an uninterruptible power source (UPS) to enhance availability. Notebook computers can provide similar functionality through integrated batteries. In this case only the RD-80 needs to be connected to a UPS or battery.
Screensavers or power saving / sleeping modes can interrupt proper data acquisition as well. Make sure they don’t become active during the measurement.
Use DISDRODATA on a PC reserved for data acquisition. Using other software on the same PC during measurements (e.g. for printing or analysing data) is not recommended. The reliability of the data acquisition may be reduced by absorbtion of system resources by other activities or by affecting the stability of the system.
Data can be exported easily at any time and be treated on other computers.