BQ25792: Emit noise when charging is in progress

HI Burhanuddin,

The most common components that make noise are the ceramic capacitors due to their piezo electric properties.  Can you provide a scope shot of SW1, SYS and BAT voltages when you hear this beep?  Also, what is the voltage rating of the 47nF BTST capacitors?  They need to be 35V or higher rated capacitors.

Regards,

Jeff

Hi Jeff,

Thanks for your quick response.

Sure, I will share the Snaps of SW, SYS and BAT voltages.

Unfortunately the 47nF is rated at 25V. I will try to use 35V and above and test again.

Hello,

Due to the U.S. holiday our responses will be delayed, sorry for the inconvenience.

Sincerely,

Wyatt Keller

Here are the snaps for your reference, this is with the 47nF/25V caps

Changing the 47nF with 50V, has no effect on it. The noise is still there.

Could it be the LC combination at the battery terminal due to which the noise is there?

Hi Burhanuddin,

I am confused by the pink trace scale.  Are BAT and SYS really down to 1.x V?  Ignoring the scale, it appears that the SYS output is hitting OVP and shutting down the converter.  Do the status and flags from 0x1B through 0x27 report OVP?  

Regards,

Jeff

Hi Jeff,

I checked and found that the Vsys voltage goes upto 13.1V (With ringing) when Battery is charging.

In addition to that I found that there is a spike in Vsys when the battery is not connected, it last for around 14.6mS

In addition to this, when the battery charges at lower currents, there is no noise and no ripple at Vsys.

Hi Burhanuddin,

When no battery is attached, the VSYS spike is expected.  It can lowered by disabling PFM and/or adding more capacitance on SYS.

When charging with battery attached, I do not expect the spikes.  Can you please provide a register dump of REG0x1B through (meaning all in between) REG0x27.

Regards,

Jeff

Hi Jeff,

Here it is, Register dump from 0x1B to 0x27:

1. Without Charging (Initially):  2F, 2A, 0, 50, 0, 0, 0, AF, 92, 50, 10, 20, 40

2. Without Charging (after 2-3sec):  2F, 2A, 0, 50, 0, 0, 0, 0, 82, 10, 10, 20, 40

3. With Charging (Vin=20@0.59A): 2F, 6A, 1, 40, 0, 0, 0, 0, 0, 0, 0, 0, 0

4. Only Battery:  20, 0, 1, 40, 0, 0, 0, 0, 0, 0, 0, 0, 0

Hi Burhanuddin,

With WD Timer expired, all registers are eventually (after 40s default)  in their default states, some of which are set by PROG pin resistor.  The WD Timer can be disabled or the host can periodically write to it.

Regarding 1, the only way for charge to report BAT and SYS OVP with charging not enabled is for a battery with voltage > VREG register setting to be attached.  But then how does the charger also report trickle charge which is V(BAT)<2.5V?  Is the battery pack protector opening and closing?

Regarding 2, see 1 above

Regarding 3, looks normal.

Regarding 4, looks normal.

Regards,

Jeff

Hi Jeff,

I ordered the EVM BQ25792EVM, and tested it with the same battery, there is no noise.

Register dump

1. Without Charging:  2B, 2A, 0, 50, 0, 0, 0, 0, 82, 10, 10, 20, 40

2. With Charging (Vin=20@0.67A): 2B, 6A, 1, 40, 0, 0, 0, 0, 0, 0, 0, 0, 0

3. Only Battery:  20, 0, 1, 40, 0, 0, 0, 0, 0, 0, 0, 0, 0

Let me know what else can be the issue?

a. VSW1 waveform

a. VSYS waveform

Hi Burhanuddin,

Can you compare the EVM BOM and layout for the VBUS, PMID, SWx, SYS and BAT components to the same on your board?

Regards,

Jeff

Hi Jeff,

I compared the two and found that in terms of BOM we are good. In layout, our PMID, VBAT caps and Inductor are on other side of the IC.

Sharing the Schematic for your reference. This design is for 2S but we are using 3S battery and its configuration is ok. R1254 is 10K5
ChargerSCH.pdf

Just an update,

Reducing the charging current to 730mA the noise vanishes, I sent 0x00 on REG 0x03 and 0x49 on REG 0x04.

Another update:

I replaced our board inductor with EVM inductor, still noise is there.

I tested with 750Khz 2.2uH inductor still the issue persists.

Hi Burhanuddin,

On the layout, the 0.1uF and at least 1 of the 10uF caps on SYS and PMID need to be on top of the IC on same side without vias as shown below:

The inductor and additional bulk caps can be on the other side.  Can you manually modify your board by scraping solder mask to add those components and see if that helps?

Regards,

Jeff

We already have one 0.1uF cap on top for PMID, I added extra 10uF on same layer as IC, still the same result.

Vsys Caps are on same layer as the IC.

Hi Burhanuddin,

Can you remove C807, the 47nF on BATP.  It is only recommend for 4S applications and is not installed on the EVM.

Regards,

Jeff

Hi Jeff,

I removed the 47nF (C807) cap on BATP, the noise still persists.

Hi Burhanuddin,

I am stumped.  Have you tried different battery or a simulated battery = >1000uF || 4-quad power supply (sourcemeter) OR >1000uF || e-load in CV mode || with power supply?  In some cases, I have seen the battery pack protector opening because it thought the IC was being over-voltaged.  This would cause the IC to trip OVP fault and stop switching like you are seeing.  What is the minimum OVP threshold for the protector?

Regards,

Jeff

Hi Jeff,

The same battery works well with the EVK.

and in our system if I reduce the charging current then there is no noise.

Hi Burhanuddin,

Oh. Ok.  What is different between the BQ2579XEVM and your PCB?  What is your PCB inductor saturation current rating?  How is the your PCB layout different from the BQ25798EVM?

Regards,

Jeff

I compared the two and found that in terms of BOM we are good. In layout, our PMID, VBAT caps and Inductor are on other side of the IC.

Sharing the Schematic for your reference. This design is for 2S but we are using 3S battery and its configuration is ok. R1254 is 10K5
2480.ChargerSCH.pdf

Reducing the charging current to 730mA the noise vanishes, I sent 0x00 on REG 0x03 and 0x49 on REG 0x04.

I replaced our board inductor with EVM inductor, still noise is there.

I tested with 750Khz 2.2uH inductor still the issue persists.

1uH inductor saturation current is 18A, DCR=6.05mOhm PN: EXLA1V0603-1R0-R

2.2uH inductor saturation current is 13A, DCR=8.5mOhm PN: EXL1V0605-2R2-R

HI Burhanuddin,

If you can use a BQ25792EVM to charge and it does not have this noise, then the problem is the layout.  I have not seen this exact issue before but I know that not having the PMID and SYS 0.1uF and at least one 10uF cap on top of the IC results in the IC not regulating properly.  

Regards,

Jeff

Hi Jeff,

We already have a 0.1uF on the same side as the IC for PMID and SYS.

Following your previous comments, there is already one 10uF cap on same side as the IC for PMID and VSYS.

We have made some changes in the layout, once I receive the new board I will test it and share the result.

Hi Burhanuddin,

Ok.  In meantime, I think it would be helpful to test using one of EVM boards and your battery.

Regards,

Jeff