Need help for the repair of a Sony Bluetooth speaker (SRS-BTX500)

[Koray]

Hi all! New member speaking.

I acquired this Bluetooth speaker in faulty state without power adapter. It was described as failing to charge internal 7.4v battery and not turning on at all. I already have a smaller version (BTX300) using the same battery, so I plugged in a fully charged battery - yup no signs of life.

So I disassembled the unit and started probing the voltages using my cheap-ass meter. When I probed the transistor (?) Q713 (see the photo attached), I accidentally brought the unit to life. Apparently, when the battery is attached, the no1 leg voltage is around 3V and dropping, no2 voltage is 8V (same as battery), no3 voltage is 2.18V. Now the interesting thing is, when the negative probe is on the ground, and when I touch the leg no3 with the other probe, leg no1 goes up to 8V and the unit can be turned on. It works fine as long as the battery is connected. The resistor R730 just below the transistor is around 460K and feeds the leg no3 from 8V battery input.

My question is, when I probe leg no3 why the leg no1 goes up to 8V? How can I keep the unit doing this, I mean keeping leg no1 at 8V? Maybe I should add another resistor parallel to R730?

Sadly I don't have the correct charger, so I have no idea if the unit can keep the battery charged and get the required power from mains. This will be the second step.

No burnt components or broken solder joints.

Thanks in advance!

Koray

[max666]

Ok, I'm just guessing here, but if Q713 is a p-channel mosfet, pin 1 being drain, pin 2 being source and pin 3 being the gate. Then maybe connecting your multimeter to pin 3 is pulling the gate sufficiently low to turn on Q713.
So if you want to keep the unit doing this, you could connect a let's say ~500 kOhm resistor between pin 3 (gate) and ground. This should keep Q713 turned on.
Or bypass Q713 all together by shorting pin 1 & 2.

But like I said, I'm just guessing here, I'm not really sure what Q713 is actually doing, so no guarantee.

[Koray]

Thanks a lot max666 that did it for me! I used a ~3MOhm resistor to avoid battery drain since R730 is connected to 8v battery + side. So far it works fine, no battery drain. I will try it with a suitable power supply and see if there are any further problems.

By the way, the OEM power supply is an oddly specified 15V 3A. Would you use a 16V 4A power supply for this device?

Thanks again!

Koray

[max666]

Cool.
I said ~500 kOhm because I thought the chances of having something around that at home are better, but also because if it is a mosfet and it is carrying the whole supply current (which I don't know), then it's on-resistance, and consequently it's voltage drop, is dependant on how low you pull the gate.
So you have to weigh up a potentially partially turned on mosfet, carrying the whole supply current, against the additional 9 µA standby current you get with two 470 kOhms resistors.
But if the 10 MOhm input resistance of your multimeter were enough to do the job, then I guess 3 MOhm are fine too.

I wouldn't expect a 16 V power supply to cause any problems instead of a normally specified 15 V, but if you're paranoid you could always drop in a diode to get it down some 0.6 V.

[Koray]

Max666, Thanks again for all this information!

When I was fiddling with the speaker I noticed that volume up button was also playing funny tricks, i.e. reducing volume or changing sound mode, etc. I disassembled the button PCB and found out the reason, apparently each button was parrallel but each with a different resistor in series, identifying their function according to the resistance they create. The funny thing is, the volume button had water damage and corrosion, so it internally had extra resistance! Thus playing up funny tricks! This is the first time I saw something like this, so I wanted to share. :-) Photos attached.

Regards,

Koray

[max666]

Whaaaat! Resistive coded buttons?! 
How did that ingenious idea come to live? "Look, we don't have enough IO pins, but there's an ADC pin we're not using ... get creative."

Thanks for sharing, haven't seen that either.

[wblock]

That's kind of a neat idea.  Just one connection, and depending on the ADC and choice of resistors, could even read combinations of buttons pressed at the same time.

[Koray]

Apparently buttons on other PCB's are also connected on the same ADC port. Faulty momentary switch could toggle functions on other PCBs, too! The drawback of this approach, of course, is that it heavily relies on extreme well being (0 resistance) of momentary switches.

K.

[urbis]

The buttons on my car's steering wheel are wired the same way to the head unit.

Different resistor value for each button.

[samnmax]

I have been bitten in the past by multiple buttons connected to a single ADC input through a resistor divider. Recently, a Sanyo DC-DA1460M stereo system that would not turn on or act erratically. Apparently the tact switches were failing and generating noise, preventing the MCU from correctly reading the buttons.
Also a Benq LCD monitor that had spurious button presses, again tact switches and ADC input.

I have measured around 5Kohm in failing tact switches, this probably wouldn't be a problem for digital inputs but it totally messes up the resistor divider approach.

[Koray]

Exactly, as samnmax mentioned, the resistive switch identification approach requires very reliable switches.

I used a 16V power supply that outputs 15.75V without load, and  yay! the unit works perfectly! Charging and discharging nicely, with no problems.

Thanks to all that has contributed, particularly max666! This is a great community here.

Koray

[max666]

My pleasure 

[mikerj]

Whaaaat! Resistive coded buttons?! 
How did that ingenious idea come to live? "Look, we don't have enough IO pins, but there's an ADC pin we're not using ... get creative."

Thanks for sharing, haven't seen that either.

My wife's Smart Roadster uses this method for the controls on the steering wheel; gear change paddles and the horn button.  A common problem is a connector in the circuit going bad and increasing in resistance, the symptom is trying to change gear activates the horn instead

[Koray]

An update for this repair, the same problem may appear in SRS-BTX300 and the solution is the same. Sony SRS-BTX300 may stop responding to power button while charging properly and with a good battery. While the circuitboards are different between BTX300 and BTX500, the same symptoms and the solution are the same. I am including a photo of the solution, connecting the base of Q713 to the ground with a 1MOhm resistor.

K.