I have a cheap 50W stereo amplifier PCB that receives its audio signal via Bluetooth. It is currently powered via a wall wart. (I could also power it with batteries.) I have one of the the board's outputs connected to a speaker and it works as expected. However, when I touch either the oscilloscope probe ground or probe tip to one of the 2 speaker leads, the sound stops. I haven't been able to imagine why. If I power cycle the board, then it will work again.

I've tried doing this with the oscilloscope turned off, and also with the oscilloscope not plugged into the wall (no earth ground). Same behavior. It doesn't happen if I touch one of the leads with my figure or use the probe while it is disconnected from the oscilloscope. Anyone have an idea of what is happening and why it would cause the sound to stop? Protection circuit on the PCB? Protection from what? (I may next be asking how I am supposed to see the audio signal on the scope.)

Also, I am seeing about 2 mV of signal noise on my (new, and new to me) oscilloscope when the connected probe is not probing anything. This seems like an awful lot. Is this normal?

 

Does it use a class D amplifier ? If do it could be harmonics from switching frequency getting into the Bluetooth receiver when a length of wire is connected to the amplifier output.

Les.

 

Thanks for the idea. It is class D. I wonder if the DAC, and perhaps anything else, would prevent any effect on the Bluetooth circuitry. Anyway, if I terminate the Bluetooth link then reestablish it (without power cycling) I still do not get sound. It seems that something needs to be reset other than the Bluetooth connection.

I can turn the amplifier's stand-by mode on then off, then upon my 2nd attempt to reestablish the Bluetooth link, it starts working again.
Also, if I put a resistor of high enough value between the probe and the speaker, it doesn't happen.

 

Thanks for the idea. It is class D. I wonder if the DAC, and perhaps anything else, would prevent any effect on the Bluetooth circuitry. Anyway, if I terminate the Bluetooth link then reestablish it (without power cycling) I still do not get sound. It seems that something needs to be reset other than the Bluetooth connection.

I can turn the amplifier's stand-by mode on then off, then upon my 2nd attempt to reestablish the Bluetooth link, it starts working again.
Also, if I put a resistor of high enough value between the probe and the speaker, it doesn't happen.

The output of wall-wart power supplies is usually not grounded. Your oscilloscope is grounded when it is plugged in. Connect the circuit common of the amplifier to the same ground as the scope (mains ground). This will probably solve your problem.

 

I was afraid to try that since I didn't understand how it might affect the amplifier circuit. But since you suggested it, I tried it. Unfortunately, it did not fix it. (And my board did not fry.)

 

I was afraid to try that since I didn't understand how it might affect the amplifier circuit. But since you suggested it, I tried it. Unfortunately, it did not fix it. (And my board did not fry.)

At least the amplifier will be much safer now it is grounded. It sounds as though the amplifier is shutting down. It's time to suspect your scope input. What scope do you have? With the scope turned off, try measuring the resistance between the scope input and ground. It should be very high.
You mention that you get 2mV of noise on the scope. That is reasonable if the input is open circuit. It should be negligible when the input is shorted.

 

My Siglent SDS1104X-E 100Mhz just arrived two days ago. The potential for operator error is large. I pressed the auto-calibrate button without probes attached and waited quite a while for that to finish. I calibrated the probe I am using. I liberally take advantage of the buttons marked Auto.

The resistance between probe tip and ground when the scope is turned off is 10.2 MΩ.

Btw, the amplifier board is the Wondom JAB2-50W (spec sheet). They have a slow web server.

 

I have a CHEAP 50W stereo amplifier PCB

The word CHEAP means poor quality and probably missing important filter parts that would prevent your problems.

 

In general, it is not a good idea to capacitively load a class D amplifier's output. The filterless class D amplifiers are (to my limited knowledge) full bridge outputs so you are probably loading down the amolifier's output.

If you don't need to connect anything to those outputs it is best to leave it alone. A loose corollary to "If it isn't broken,..."

 

Thanks. I tried touching 55 feet of 16 gauge wire (open at the other end) to the speaker lead and it didn't cause the problem, but since I haven't broken the board testing, I will see what happens when I touch one end of a capacitor to it. If I don't learn anything from that, I will give up trying to understand it, which kinda makes sense since I don't have a schematic of the amplifier.

However, I do wonder how a person is supposed to examine the variations from a straight line frequency spectrum the amplifier puts out. Perhaps measuring with a resistor between the speaker lead and the scope will provide me with a accurate visualization, if not an accurate voltage vs. frequency.

 

If it is class D you will see rectangular waveforms on a scope and a forrest if switching harmonics on a spectrum analyzer.

If you insert a low pass filter between the amplifier output and the scope
you should be able to see familliar audio waveforms if your amp doesn't shut down.

 

I have a cheap 50W stereo amplifier PCB that receives its audio signal via Bluetooth. It is currently powered via a wall wart. (I could also power it with batteries.) I have one of the the board's outputs connected to a speaker and it works as expected. However, when I touch either the oscilloscope probe ground or probe tip to one of the 2 speaker leads, the sound stops. I haven't been able to imagine why. If I power cycle the board, then it will work again.

I've tried doing this with the oscilloscope turned off, and also with the oscilloscope not plugged into the wall (no earth ground). Same behavior. It doesn't happen if I touch one of the leads with my figure or use the probe while it is disconnected from the oscilloscope. Anyone have an idea of what is happening and why it would cause the sound to stop? Protection circuit on the PCB? Protection from what? (I may next be asking how I am supposed to see the audio signal on the scope.)

Also, I am seeing about 2 mV of signal noise on my (new, and new to me) oscilloscope when the connected probe is not probing anything. This seems like an awful lot. Is this normal?

The high-frequency components of Class D amplifier outputs can generate large amounts of EMI and disrupt operation of other equipment. If you don't want yours to shut down, don't connect anything to the speaker leads that can act as an antenna .

 

The SURE amplifier uses a Texas Instruments TPA3116 IC. The IC detects input DC voltage, shorted output and/or over temperature, causing it to shut down and give a warning on its "Faultz" pin. The IC is filterless but shows ferrite beads or LC output filters needed sometimes.

Maybe your wiring is picking up its 1MHz switching output from a speaker wire and feeding it into the input. Your input wiring might have poor shielding.
Maybe your amplifier IC is defective.
Maybe the input coupling capacitors pass DC.

 

Ah, ah, super good info! Thanks all for responding. I will leave the PCB alone now. Your comments have directed me to a number of things to study.

 

The output of wall-wart power supplies is usually not grounded. Your oscilloscope is grounded when it is plugged in. Connect the circuit common of the amplifier to the same ground as the scope (mains ground). This will probably solve your problem.

That’s the first thing I thought of...well either that or he was shorting something?

 

I made a wooden case for this amplifier and gave it away as a gift, so I'm done with it. I did notice beforehand, though, that if I ran my 18V DeWalt electric drill near the PCB that it would cause the output to stop like it did when touching the speaker wire.

 

Thanks for the idea. It is class D. I wonder if the DAC, and perhaps anything else, would prevent any effect on the Bluetooth circuitry. Anyway, if I terminate the Bluetooth link then reestablish it (without power cycling) I still do not get sound. It seems that something needs to be reset other than the Bluetooth connection.

I can turn the amplifier's stand-by mode on then off, then upon my 2nd attempt to reestablish the Bluetooth link, it starts working again.
Also, if I put a resistor of high enough value between the probe and the speaker, it doesn't happen.

Petkan:
It is the probe input capacitance causing the problem. Try setting it 10:1 division mode. It will reduce input capacitance tenfold