Fixing Humidifier Red Light: Beurer LB37

Our dehumidifier stopped working recently: you'd fill it up with water, and it would not turn on, showing a red LED indicating "water level too low". Here's how I debugged the problem, though, sadly, I couldn't fix it.

White cone on coffee table, mist is coming out the top. The blue light is on.
Happier days. A promotional image of a (working!) Beurer LB37 Humidifier

Four years ago, we bought a Beurer LB37 humidifier. We'd fill it with water, and it had a ultrasonic speaker that turned the water into mist. It had a sensor that turned off the speaker when the water ran out.

With the red LED showing, I guessed something must be wrong with the water-level detection. There are a few ways you can detect water level: weight sensor, conductivity sensor. I wasn't sure what was happening. Opening it up, the only moving part I can see is this button that's pressed in by the top half (water tank).

I suppose you can use a button to detect weight – maybe put a spring inside and use Hooke's law to measure the force?

Lower half of humidifier; showing a button the water tank can depress, and the ultrasonic speaker membrane.
Button that gets depressed when the water tank is heavy enough.

Underneath each of the four corners of the humidifier, there are screws to open it up. Open it up, and I see this:

Pulled apart lower part of humidifier. Has AC coming in, chip boards at the top and left, wires connecting them.

On the right side is the transformer (AC in, 24V DC out x2). This supplies power to to the fan (middle, black), and to the ultrasonic speaker circuit (middle left, gray). It's all controlled by the push-button (left green board) and weight sensor (top, gray).

I pulled apart the weight sensor; it's just a switch: Jufond SW312.

Black box switch with tiny button at top; says JUFOND SW312, two wires coming out bottom
JUFOND SW312 Switch

I thought, maybe the switch is broken? That would be consistent with the error. To test this, I attached my multimeter in resistance mode to the terminals and depressed the switch: it still read infinity ohms. I guess the switch is the problem?

I doubt I can fix it, but let's try to see what's going on. I opened up the switch (it has a clip at the left hand side). It's surprisingly intricate inside, like a watch:

Switch internals.

The switch button (top) presses into the spring, which pulls down the top right-hand side contact, to (hopefully) press into the bottom contact, completing the circuit. The spring adds some hysteresis, ensuring the switch stays on or off and avoiding 'bouncing'.

I pressed in my multimeter (and recruited a second set of hands), and yeah, the pads weren't making contact enough to connect the circuit. The multimeter still read infinity ohms.

I thought, maybe the pads are slightly corroded? But I didn't see any rust on them when I pulled out the bottom probe.

If I applied a little more pressure on the top with the multimeter probe, then the pads connect to form a circuit. Maybe the spring just doesn't have enough tension to force the pads together any more?

Can I replace it? On a whim, I Googled [JUFOND SW312] and found some places selling replacements on eBay for USD 8.00, but they take a month to ship, and I'd have to solder it on.

Unfortunately, I think this is the end for my humidifier. It's in the e-waste now. At least we had some fun pulling it apart. Maybe this story can help you debug your humidifier?

Mark Hansen

Mark Hansen

I'm a Software Engineering Manager working on Google Maps in Sydney, Australia. I write about software {engineering, management, profiling}, data visualisation, and transport.
Sydney, Australia