Pin A7 used for Buzzer, makes high frequency low volulme tone with active charging

E-series LTE
OS: 1.0.1

I have a buzzer hooked up to pin A7 (WKP) and a high-frequency low volume tone comes from it only when the device is plugged in and charging on the Eseries. Even with the notone function called and also analogwrite(pin,0).

I have the same buzzer and firmware hooked up to a P1 and this noise doesn’t come from the Buzzer.

I’ve tried many different things, but what I’ve concluded is that this ONLY happens when the charging circuit on the E-series is active. As soon as I unplug external power the high-frequency noise stops immediately.

I am thinking that there is either EMI from the charger, causing the buzzer to energize or the WKP pin is doing something weird when the charger is active.

Any thoughts on this would be appreciated.

Not confirming since I don’t have an E-series. But for a fix, or just to test the theory, you could try adding some bypass capacitors to A7 or an LC low-pass filter.

Right now this is my current circuit. So put the RC low pass on the A7 signal?
image

Hard to say without a bit more research on my side but I have limited information about your components. For my 2 cents… I would put a .1uF and maybe a 1 or 10 uF cap on the A7 line to start. And I might do the exact same on the buzzer’s POS or NEG line (or between the two). I would assume the EMI would be affecting the NEG line since it has the connection back to the E-series but I haven’t done a low-side modulation where I’ve done the research. Assuming this isn’t a PCB yet, try the A7 then add to the NEG line and then finally the POS line as needed in your testing.

Thanks, this is on a 4 layer pcb the eseries is on top and the buzzer is on the bottom. It’s connected to A7 and GND. Not sure if this helps, but definitely seems to have some sort of EMI

This is before the USB is plugged in

This is after the USB is plugged in:

Does that scope show a small DC offset the after USB waveform? Maybe the 1M Ohm pull-down is too high allowing a bit of positive voltage on the A7. Since you can get a scope on it, I guess you should be able to add a smaller resistor in parallel to see if a lower resistance is better… and add those bypass caps.

Maybe ping @ParticleD, @mstanley, @rickkas7… they might have some insight or best design practices with the E-series.

It does show a small DC offset when the USB is plugged in.

I’ll try those suggestions with the scope on it.

@ninjatill The high tone actually goes away when I connect the o-scope ground to the PCB ground. If I connect the groud from the PCB to the USB ground on the outside of the connector, the tone also goes away.

Do you have anythoughts why the connecting the ground to a differnet ground stops it?

I think I found the issue, I have a Ferrite bead (600 OHM @ 100MHz, 500MA
[BLM18AG601SN1D]) connected between USB ground and the PCB ground. When I bypass this bead the sound goes away.

Below is an image from a spectrum analyzer (audio). The red is when no short (high pictch sound), the green is when there is a short no ferrite bead (no sound).

I also only notice this high pitch sound when the charge cycle is almost complete or when the charger reinitiates a charge cycle by unplugging the usb then plugging it back in. The sound lasts for a bit (maybe a minute) then it stops. Directly correlating to the charging circuit.

Do you have anythoughts why the ferrite bead would cause this issue?

Might be a good question for @bko or @peekay123.

@wesner0019, the inductor separates the USB ground from your board ground which is never good. A direct connection is preferred. The inductance may be reacting with a real or stray capacitance in the circuit causing unwanted oscillation. Or, it may cause a ground potential difference leading to potential ground loop issues.

4 Likes

@peekay123 thanks, do you think I should connect it to ground directly or would a 0 ohm jumper work so the PCB doesnt need to be redone?

@wesner0019, a zero ohm jumper would help with your existing board but your next board version should have a direct trace connection.

2 Likes

Hi @wesner0019

I agree 100% with @peekay123. It looks like you have multi-layer board, so tying to the ground plane is best for the next board run. There are only a handful of cases where an isolated ground is a benefit (extreme low noise A to D, for instance) and then it requires pretty careful engineering to avoid problems.

2 Likes

Thanks, @peekay123 and @bko, I’ve updated my PCB.

2 Likes