Ultimeter Weather Station - Pro Rain Gauge
Rain Gauge Repair
The Pro Rain Gauge is an electronic water drop counter.
My Peet Brothers Pro Rain Gauge stopped working after 3 years so I took
it apart to see if I could fix it.
The first step in any repair is taking the device apart. This is
sometimes called a teardown. It must be done in a way that allows you to
put it back together again. Taking pictures and putting screws and other
removed parts in labelled bins will make correct reassembly easier.
Observations:
All components are through hole (no surface mount).
The PCB is single layer so none of the holes are plated.
The board has a sprayed on conformal coating that did not prevent the
screws from oxidizing.
The silk screen next to the TO-92 transistor looks like a date (940524).
The conformal coating makes reading the transistor part number
impossible.
The DIP IC is TI CD4017B a divide by 10 counter with hysteresis on the
clock input.
The 3 pin connector to the left has silk screen that encloses a pad on
the output trace.
This connector on my board had moved and caused the copper traces to
lift off the fiberglass PCB substrate. I held it in place and added some
glue around the base of the connector to hold it to the top of the PCB.
https://en.wikipedia.org/wiki/TO-92#Disadvantages
The TO-92 pin out is E-C-B.
Perhaps 2N3395 or 2N2714.
Drop formation and detection
The drop sensor and drop forming tube has 4 gold plated pins that are
positioned so a drop contacts at least 2 pins. The pins alternate with
the white wires connected to +5V via a 10K resistor and the black wires
connected directly to the base of a bipolar transistor. I confirmed it
was a bipolar transistor by measuring the diode drop of the PN
junctions.
The screen helps with drop formation by keeping the water surface
tension supported as the drop size grows to the diameter of the hole
below the screen. This abstract has a good description
https://www.sciencedirect.com/science/article/abs/pii/S0301932201000611
You can see the tips of the pins through the hole after removing the
screen and gasket.
Looking from the bottom of the drip tube.
Another problem I found during my teardown was one of the gold plated
pins had pulled part way out of its hole. Simply pushing it back in
wouldn’t keep it in place so I temporarily held the pins in place with
tape and put a large drop of gap filling cyanoacrylate on the outside.
After it cured, I made sure no adhesive had wicked onto the electrodes
inside the tube.
When a drop connects 2 pins current flows into the transistor base. My
tap water measures a few hundred kilo-ohms. Measuring flowing water
reduces the electrolysis during the measurement.
Why is the drop counter set to 0.001” per drop?
Because the tipping bucket rain gauge measures in 0.01” increments.
There is a practical limit to drop size and that volume will have much
less rain than the bucket volume. A simple circuit to count smaller
drops and then making the drop size followed by the counter produce
0.01” increments allows the same weather station to use either the
tipping bucket gauge or the drop counting gauge.
![](../files/PCB_bot.png)
PCB Bottom with signal names
![](../files/raindrop_counter.png)
Schematic w/o CD4017B
The RESET signal isn’t really useful. It just puts the CD4017B in a
known state at power up. It doesn’t really matter if we are off by a few
0.001” of rain.
V1 in the LTSpice schematic mimics a water drop. The drop duration is
determined by the speed of the drop as it falls and the drop size. My
numbers are made up. Let’s say the current gain is 100 and the drop
resistance is 500K. The base current is 10uA and the collector current
is 1mA except the collector resistor limits that to 5V/100K = 50uA. The
sensor side of C3 is pulled to ground and C2 is discharged at 1mA until
the transistor is saturated. When the drop clears the sensor, Q1 turns
off and C2 charges back up to 5V via R3. The CLK pin side of C3 follows
the transistor collector (the voltage across it was zero) and then
charges back up to 5V via R2.
What resistance do we expect for rain water?
https://sensorex.com/resistivity-of-water/ says 20Kohms.
Video of drop formation