Harman/Kardon Onyx Bluetooth Speaker - No Power

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Also since I soldered in a new unknown value capacitor would that affect anything?

No that should not be an issue. That capacitor only determines the time it takes for the reset line to reach a logic high and bring the BT module out of reset. You said it turns on a little slower now? Likely the capacitor you put in is just a little larger than the old, so it takes a little longer to charge. Unless the longer delay bothers you, don't worry about it.

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

I'm wondering whether the charging problem is related to your earlier troubleshooting work (4V voltage source, thermistor, etc).

Think I damaged something with my earlier troubleshooting?

Possible. What exactly did you do when you were playing around with the battery and thermistor input?

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

So now I have a new charging issue? Lol

Also since I soldered in a new unknown value capacitor would that affect anything?

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Ok I removed the resistor I added and desoldered c304 and the unit powered up.

I measured C to E on Q301 and got .793 in one direction and infinity ohms in other under diode setting.

I replaced c304 with another smd capacitor off a junk board (it's a little bigger in size than original on there) can't tell what capacitance original was, and unit works I just noticed that it takes maybe fraction of a second longer to turn on now.

Reset voltage is now sitting at 3.25v
The capacitor I removed reads open/high ohms

So it all worked with your 2k resistor and C306 removed? With a new capacitor it measures 3.25v on the line? That's good, sounds like C306 was leaky then. If it was faulty before, your desoldering may have finished it off, and made it open circuit. You would have to test with a capacitance meter to be sure.

The flashing power button indicating low battery/charging is still going on and the battery reads 3.78v so it didn't charge at all over night.

I'm wondering if the excessive current draw from adding the 2k resistor kept it from charging?

When I say 'excessive' it's relative. It would be excessive for a reset circuit which would normally draw almost nothing. I doubt it would interfere with the battery, since it would only draw about 2mA even if C306 was dead shorted.

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Thanks guys. I'm curious, what's the purpose of the reset line? Is it important for operation? In other words if I removed c304 or r304 to disable it, would the unit work fine?

The reset line is to reset the BT module into a known state on power up. Pretty much every digital circuit aside from very basic ones need a reset signal, because logic gates, flip-flops, RAM etc don't automatically power up in a known state, and each chip will have slightly different behavior due to manufacturing differences. An external reset signal (and the corresponding internal reset circuitry of the module) deals with all these unknowns.
So no, disabling the reset line completely would not be a good idea.

Also suppose c304 is leaky, how do I determine the value to replace it?

Unfortunately it would be impossible without a schematic if the original capacitor was faulty and unmeasurable.. But given the pull-up resistor value and the capacitor charge formula, it would be easy to calculate which value would keep reset line low for at least 5ms as per the BT module spec.

And btw I measured the same 781 ohms from C to E of Q301 in one direction and open in the other using the diode function, if that helps.

Interesting, I wonder if Q301 has gone leaky C-E in one direction only.
EDIT: this one still puzzles me, since it now looks like C306 was the culprit for the low reset line voltage.
Which way around were the leads when you measure that 781 Ohms?

Which calculations are you using to determine something is a miss with the provided info?

Simple voltage divider formula, that's what I was using at least. But that is a static analysis from your static measurements. When normal power is applied in the circuit, the conditions of the fault may change, which probably explains the discrepancy.

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Reset voltage is now sitting at 3.25v

The capacitor I removed reads open/high ohms

Think I damaged something with my earlier troubleshooting?

[fzabkar]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

What is the reset voltage now? Is it sitting at 3.25V? Does the original capacitor test resistive out of circuit?

I'm wondering whether the charging problem is related to your earlier troubleshooting work (4V voltage source, thermistor, etc).

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Ok I removed the resistor I added and desoldered c304 and the unit powered up.

I measured C to E on Q301 and got .793 in one direction and infinity ohms in other under diode setting.

I replaced c304 with another smd capacitor off a junk board (it's a little bigger in size than original on there) can't tell what capacitance original was, and unit works I just noticed that it takes maybe fraction of a second longer to turn on now.

I'm going to try running it like this for a while and see what happens.

The flashing power button indicating low battery/charging is still going on and the battery reads 3.78v so it didn't charge at all over night.

I'm wondering if the excessive current draw from adding the 2k resistor kept it from charging?

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Thanks guys. I'm curious, what's the purpose of the reset line? Is it important for operation? In other words if I removed c304 or r304 to disable it, would the unit work fine?

Also suppose c304 is leaky, how do I determine the value to replace it?

And btw I measured the same 781 ohms from C to E of Q301 in one direction and open in the other using the diode function, if that helps.

Which calculations are you using to determine something is a miss with the provided info?

And one other thing, I had the unit plugged in over night to charge the battery, on this unit the power button flashes while it's charging and should go solid when charged but it didn't it continues to flash. Being plugged in overnight I'd imagine it should be charged. Could the resistor added be part of this behavior?

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Ok so I didn't have a 1k resistor but I had a 2k, soldered it in across R304... And it worked!!

That's good, it seems we're getting somewhere

Also it brought the voltage to 2.53 to that reset pin and 3.25 is on the opposite side of that same resistor. So was r304 just simply open? Because before I measured 3.25 on one end and only 1v on the other where now it's 2.53 with added 2k resistor.

Wait! I'm thinking of R308 connected between B and E oops. What was R304 connected to?

R304 provides pull-up for the Reset line, but at 14k, it's rather weak, so any significant current draw on the reset line will lower the voltage (and it did, to 1v)
Both sides would be very close to 3.25v during normal operation, since the BT module's CMOS input will draw very little (effectively no) current.
But in this case, something is drawing more current than it should, as evidenced by the fact that you still only get 2.53v with the much stronger 2k pull-up resistor.

Can I keep the 2k resistor soldered in and use the unit or will it cause damage somehow?

The "fix" is really only a temporary kludge. Leakage usually gets worse over time, so eventually the 2K resistor will no longer be able to compensate for the increased leakage.

Exactly. It might work for now but something is still wrong with the circuit. What you have proved though is that the BT Module is probably OK (Since it's now functioning), so it's likely just down to C306 or Q301 being faulty.

That said, smt rework is fiddly. I would try to narrow down the fault by heating the collector of Q301 with your iron, and measuring the reset voltage before and after. Do the same for C306, after allowing Q301 to cool. Hopefully the leakage will be thermally sensitive.

Before you try this, I would wait for Agent24 to come online. Perhaps there is a better approach.

Heating (or using freezer spray, or both) on Q301 could prove the theory if it's the fault, aside from poking it I can't think of anything else at the moment. I would be leery of heating C306 though as they can be easy to damage with thermal shock. You're *supposed* to use hot air on MLCCs.

The other thing that bothers me is the measured value of 781 ohms for C306. My calculations suggest that the leakage resistance required to produce a reset voltage of 2.53V is around 6Kohm. I'm wondering if this discrepancy could be the result of deforming the capacitor during the measurement process.

And the calculations for a 781 ohm and 14k don't make sense either, the resulting voltage should be around 0.2v, not 1v.
C306 could be pressure sensitive (would make sense for a cracked MLCC I guess), or perhaps the voltage\current supplied by the meter is different enough to show a different result compared to the in-circuit operation.

[fzabkar]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

I would also perform a B-E and B-C diode test on Q301, in both directions.

Something else which occurred to me is that R304 and C306 determine the RC time constant for the reset power-on delay. The miminum delay is specified as 5ms so we would need to ensure that lowering the value of R304 does not violate this spec.

The other thing that bothers me is the measured value of 781 ohms for C306. My calculations suggest that the leakage resistance required to produce a reset voltage of 2.53V is around 6Kohm. I'm wondering if this discrepancy could be the result of deforming the capacitor during the measurement process.

[fzabkar]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

The "fix" is really only a temporary kludge. Leakage usually gets worse over time, so eventually the 2K resistor will no longer be able to compensate for the increased leakage.

That said, smt rework is fiddly. I would try to narrow down the fault by heating the collector of Q301 with your iron, and measuring the reset voltage before and after. Do the same for C306, after allowing Q301 to cool. Hopefully the leakage will be thermally sensitive.

Before you try this, I would wait for Agent24 to come online. Perhaps there is a better approach.

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Wait! I'm thinking of R308 connected between B and E oops. What was R304 connected to?

And can I keep the 2k resistor soldered in and use the unit or will it cause damage somehow?

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Ok so I didn't have a 1k resistor but I had a 2k, soldered it in across R304... And it worked!! This thing is alive and working. Wow thanks so much guys I honestly appreciate it a lot.

I'm curious, how did the addition of that resistor work if the resistor is a B-E resistor? I thought those were just used for internal current of the transistor. Confused how it brought it to life

Also it brought the voltage to 2.53 to that reset pin and 3.25 is on the opposite side of that same resistor. So was r304 just simply open? Because before I measured 3.25 on one end and only 1v on the other where now it's 2.53 with added 2k resistor.

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Add a resistor to r304 or r306? R306 is the one going to the reset pin

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

What if I provide 3.3v to the reset pin to keep it high? Thinking of ways to avoid having to remove smt components lol but I could try removing them and testing as well, just curious if that would be a good test as well.

If you give 3.3v to the reset pin directly from the 3.3v rail you risk a potentially shorted part going up in smoke or just shutting down the supply. I wouldn't want to blow C306 if it is bad, burning ceramics can make nice holes in PCBs.

fzabkar's idea of lowering the value of R304 is probably a better compromise. You should be able to tack a standard resistor's leads across R304 fairly easily too.

"BUT... If you are measuring only 781 Ohms across C306 (and incidentally, probably also across C-E of Q301) that could indicate a problem." I'm curious what the problem could be, shorted capacitor?

Yes, the only things that make sense as I understand it, are that Q301 or C306 could be shorted\leaky, or the same could apply to the BM150 module itself, since those are the only 3 parts that seem to potentially provide any path to ground from the reset line.

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Maybe C306 is leaky. I've seen leaky disc ceramic caps, but the only MLCC faults that I've encountered have been dead shorts. Perhaps that's because they have all been supply bypass caps rather than "signal" caps. A leaky cap across a supply would go undetected until it goes completely short circuit.

Or perhaps with the relatively high resistance of R304, the current through C306 is being kept low enough to stop it completely breaking down. Shorted capacitor biased in the leaky region?

Another possibility is that Q301 may have a leak between C and E.

Seems maybe more likely, would explain why the meter doesn't show 0 ohms across C306, which I would expect if it were bad.

Would it make sense to place a 1K resistor in parallel with R304? This would pull the reset pin higher without risking damage to Q301 or U300.

It could work, assuming the potentially bad part is only leaky. If it's properly shorted I would assume that it's unlikely to make a difference. Worth a try though, I guess.

Disconnecting\removing parts would be a more reliable test though.

[fzabkar]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Maybe C306 is leaky. I've seen leaky disc ceramic caps, but the only MLCC faults that I've encountered have been dead shorts. Perhaps that's because they have all been supply bypass caps rather than "signal" caps. A leaky cap across a supply would go undetected until it goes completely short circuit.

Another possibility is that Q301 may have a leak between C and E.

Would it make sense to place a 1K resistor in parallel with R304? This would pull the reset pin higher without risking damage to Q301 or U300.

[caphair]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

What if I provide 3.3v to the reset pin to keep it high? Thinking of ways to avoid having to remove smt components lol but I could try removing them and testing as well, just curious if that would be a good test as well.

"BUT... If you are measuring only 781 Ohms across C306 (and incidentally, probably also across C-E of Q301) that could indicate a problem." I'm curious what the problem could be, shorted capacitor?

[Agent24]

Re: Harman/Kardon Onyx Bluetooth Speaker - No Power

Yes, I would personally try disconnecting the BT module, C306 and Q301, and check if the conditions change after disconnecting each one. However, SMT can be tricky and ideally you'd want hot air for this.

MLCC can definitely fail shorted, I don't know about leaky, but I suppose it's possible.