bq51013A output noise

Is this similar to what you are experiencing?

http://e2e.ti.com/support/power_management/wireless_power/f/693/t/278092.aspx

If so, I've seen this with the bq51050B receiver as well. It has to do with the piezoelectric effects of MLCC capacitors. The linked comment thread has some suggestions on how to minimize the audible noise.

Hi,

 It looks like the issue is the tuning of the coil, please tune the coil according to the final configuration.

 Measure Ls and Ls’ and use the equation in the datasheet in order to calculate Cs and Cd. Use a ferrite material as in the WPC spec and measure inductance at 1V RMS and 100 kHz.

 The Cs Cd provided in the EVM will work only for this EVM, because the final configuration of your system is much different from the EVM. The EVM is tuned and according to its final configuration.

 Thanks!

Tahar

Thanks both for your replies!

Ashley:

That is most likely the same issue, but I'm not confident that my design is good enough and simply changing capacitors to get rid of the sound is the best solution. It may be though.. 

Tahar:

This is where I'm confused. I've used the equations in the datasheet to calculate Cs and Cd, for both the 10 uH specified, and the 10.2 uH I'm experiencing when I have the coil in our system.

These two values are rather similar, and I notice no practical difference when running the receiver inside our system or on its own.

I'm using the Wurth coil (760308201) which is specified as an alternative in the BOM for the EVM. Therefore I would guess that running both the EVM and my system in the same environment should work with the same Cs and Cd values? 

I feel as I'm missing out on something here..

Hi Peter,

Even when using the same coil, the behavior of the coil will be and should be different from board to board with different configuration. The EVM and your board are different, so the behavior should be different. That the purpose of the tuning of the caps withnthe final conf.

The other thing you may need to consider is increasing the COMM caps and see if that helps.

Sometimes, the final configuration of the board has a lot of effect on the coil inductance, consider increasing or reducing the number of turns to have it around 9uH to 12uH for Ls. Also consider adding more shielding helps too.  

Of course, if you are using wpc1.1 Tx and the FOD is enabled, you need to calibrate you board for FOD, otherwise you will have FOD issues. if you are not calibrated yet, consider using wpc1.0 Tx or disable FOD function on Tx.   

Thanks

Tahar

Hi,

But Equation for C2 is wrong in bq51051B datsheet.

Ls - 1/C1 is negative value as Ls is small and 1/C1 is very big number.

Please help,

All the best

Hi, The equation should be the same as in the bq51013B datasheet (page 30) Thanks! Tahar

Hi,

I had the same problem and I solved it.

In the Analog Applications Journal for designing a Qi-compliant receiver coil
for wireless power systems, Part 1, (Documentation from TI) I saw the following equation. But i found out, that this equation is wrong. Perhaps, there are more wrong equations for calculating C2/Cd  in the Web. 

Wrong equation:

This following equation is the correct one from WPC-Standard-Sepcifactions:

Correct equation:

All the best

Jonas

Hello,

Peter Faltpihl, I have the same problem with my wireless receiver, would please tell me if you have succeded to resolve it by changing the values of the matching capacitors ?  Did you test another solution?

Thank you very much

sorry,

I can not send you a picture because I do not have the circuit for the moment, I've worked with a friend on this project and we had these pulses on the current but as the charger worked well we didn't think to fix it, and here I found this post and I really wanted to know the cause and solution? so I can't send any picture of pcb or oscillo :(

Do you think It was a problem self made coil? or the matching wasn't good? or the nature of the capacitors (x7R, X5R)

e2e.ti.com/.../278092
Thanks

There are a lot of responses to this post so it is difficult to answer each one.  I'll start with an overview. 

The original post showed a plot of OUT with periodic "noise" on the trace.  This corresponds to the normal communication pulses from the RX to the TX.  This is expected based on the method of communication.  The communication pulses are a 2kbps signal superimposed on the magnetic field created by the TX.  Details of the communication can be found in the datasheet.

The post also mentions a high pitch noise that seems to correspond to the communication pulses.  In most cases, this noise is caused by ceramic capacitors.  The vibration of the ceramic capacitors on the PCB generates this noise.  There have been several solutions for this.  One popular solution is "slots" in the PCB beneath the capacitors.  Other solutions include "low noise" capacitors.  Murata produces "Acoustic Noise Reduction Capacitors" that mitigate the issue.

Regards,

Dick

Dear Dick and Miroslav

Sorry I'm not an english speaker, I tried to understand Dick's  response, would you please tell me what 's the correct conclusion because I m confused now:

1) These pulses in the charging current are normal and this is due to the communication protocol

2) The pulses in the charging current are not normal and caused by a not good matching, solution recalculate the values of the matching capacitors

3) Pulses are not normal and caused by the types of capacitors used in the pcb (X7R, X5R...) and the solution is to add slots or use a low noise capacitors; please in  this case Capacitors that cause these pulses are capacitors used for matching or others capactors : connected to the pin Bat or rect?? I used a BQ51050b for my pcb

Waiting for your help

Best regards

Star,

Sorry for the confusion.  

Number 1 is the correct answer.  The pulses are normal.  Portions of number 3 are also important.

The amplitude of the pulses are mainly controlled by the capacitor on the COMM1 and COMM2 pins.  Higher value capacitors creates larger pulses.  Lower value capacitors create smaller pulses.  The concern in a system is that communication is critical.  If the pulses are too small, then the communication becomes less robust.  

In general, the acoustic noise comes from the transmitter side and not the receiver side.  In most cases, changing the COMM capacitors on the receiver side will not impact the acoustic noise on the transmitter side.  Tracking down acoustic noise can be very tricky.  You will have to identify the source on a case by case basis.  Techniques to detect noise include removing the suspected capacitors or adding wires to the capacitors to move them physically away from the board.  Adding slots can be a solution, but it is not the only solutions.

Let me know if you have more questions.

Regards,

Dick