Returning from holidays, my Dad often has surprises for me. This time, it was a few items that needed attention – the first of which was three Bauhn Remote Controlled Power Point units with various faults, and the other was an AKAI TV Remote Control that ceased to function after some liquid damage.
Bauhn Remote Controlled Power Point
These are nothing but your ordinary run-of-the-mill 433Mhz wirelessly controlled power points. These ones were purchased at Aldi in May 2014 judging from the label, with a model number of HE214039. It seems to have been imported by Winplus. They feature a pretty nice slim design that doesn’t obstruct neighbouring outlets, unlike the WattsClever units I personally use. As I haven’t got the transmitter, I didn’t have a chance to examine or test the RF-side signals.
Three units were received for repair. Of the three units, the relays of two of the units latched on as soon as the units were powered up, whereas the third unit had an entirely unresponsive relay. On all units, the LEDs lit up and turned off with each press of the manual power button. All units were faulty based on that quick diagnosis.
The units are made with Y-shape security screws, so getting into the unit is the first line of difficulty. Finding the right security bit was no big issue for me.
The internals don’t look particularly complex. This particular unit is single-pole switching which makes it illegal to sell in NSW due to a change of law. Anyhow, it seems that it’s got a fusible resistor as protection, a capacitive dropper and zener diode shunt power regulation for the ICs (to 5V at a guess), with an expected DC voltage of about 24V to match the relay drive requirements. The relay is an AFE BRD-SS-124LM with a rating of 12V at 250VAC. The relay is driven by Q3, and has a diode to protect against reverse EMF. The IC in the top corner is a regular LM358 Op-Amp.
Incoming dropper capacitor is an X2 rated 0.33uF polyester unit.
Smoothing capacitors are not any brand that I recognize, but has a rating of 220uF at 35V, so the incoming rail is not expected to be above 35V when smoothed and loaded.
The underside houses a few surface mount components and an unlabelled IC (probably a microcontroller) that does all of the work.
Troubleshooting and Repair
With the symptoms in mind, I decided to first try and check the resistances of the relays just to confirm the contacts are okay. Such cheap ice-cube style relays often fail especially when switching heavy loads. I measured the resistance across the input active pin and the active socket. All registered <0.05 ohms except one which registered 0.64 ohms and was slow to actuate.
Further examination showed that this was due to an obviously cold solder joint that was not properly made at the factory, resulting in a resistance which could (in future) cause heat to build up and possibly fire. A manufacturing quality problem. This joint can be seen to pose difficulties in another unit as well, where the insulation seems to have been partially melted during construction.
After that was repaired, the resistance fault was fixed, however slow pick-up and drop-out was still experienced. This suggested that maybe applying power direct to the relay coils (noting the reverse EMF protection diode) was allowing power to leak and charge the capacitors on the board.
I turned my attention to the two bad boards with latching relays that turned on immediately when powered up. Instead of powering from AC, I decided to apply 20V DC straight to the output of the rectifier after the capacitive dropper as a safer alternative. I first checked the 9014C transistors driving the relay on both units. It seemed both tested dead short after desoldering, and had also seemingly failed. Such failures might imply the back EMF protection diode may have opened, but testing it revealed it was still functioning. My next suspect was the drive to the transistor.
It was then that I noticed that the rail powering the IC had become unregulated, showing the full voltage across the IC. These two units were immediately declared beyond repair as the IC is very likely fried by having 24+ volts across it when plugged into the mains. This was verified by looking at the drive output to the transistor (a 9014C) driving the relay – it did not toggle with the state of the LED and manual override button.
The reason for this is due to the failure of the Zener diode in the shunt regulation scheme – the Zener is supposed to “waste” excess energy as heat to draw the voltage down as a crude method of regulation. It was probably a bad batch, or poor design resulting in overstress. Once it opens, the 5v rail “flies” up to the unregulated incoming smoothed capacitive dropper voltage.
Noting this, I approached the third unit which still had a functioning Zener, and was regulating correctly. Instead, its 9014C relay drive transistor was open. I scavenged a 9014C from the other two boards in the power section rather than the relay drive section and soldered it in. I also swapped the relay with one from the other unit as I suspected this unit may have a problem with the relay causing slow actuation. Once this was done, relay switching was verified when powered from the bench supply.
Re-sealing the unit, I plugged it into the mains only to find that it still did not function. I suspected a failure of the capacitive dropper, which was correct – the X2 capacitors used as droppers in all three units measured 91.48nF, 115.9nF and 89.85nF rather than the expected 330nF. As a result, insufficient current was being let through from the mains, so the voltage was too low to run the relay. Replacing it with a scavenged “dipped” capacitor from an old CFL was enough to make the unit run again. A one-out-of-three is better than nothing, however, I have a feeling that the Zener on this remaining unit may well fail soon enough.
AKAI TV Remote Control
While I’m at it, I might as well take a look at this. Nothing special – just a remote control.
As it turns out, it’s our worst enemy – it seems some leaking carbon zinc cell may have spilt its contents into the remote, corroding its terminals and causing it to stop working. This is why I’ve now instituted a policy of pre-emptive battery replacement on a schedule – just don’t leave it until its too late. Also, don’t leave your remotes where buttons could be stuck pressed, or you could come to see such nastiness.
The volume of liquid leaked was surprising – enough to coat the board with a small film.
It also permeated into the other side, where it was likely it shorted out at least one pair of sensing pads, resulting in a “stuck button” condition, which would cause some remote ICs to “shut down” or at least, fail to detect most buttons.
After a clean down with some methylated spirits, it was clean enough to use, although the terminals remained slightly corroded. The best option would have been to replace the terminals, but that’s easier said than done. Finding a perfect match is not easy. Besides, I know from experience, the heat from a soldering iron tends to make these reactions go quicker, so you get some very unpleasant smelling/looking gunky mess which may leave pads un-solderable. So for now, seeing as it works, it’ll do.