BQ51003: No power transfert when strong coupling just high voltage large pulses for VRECT.

Hi Alain

The L value of the RX coil looks high, what is the coil dimensions?

The communication capacitors are typically 10% to 20% of the Cs capacitors.  Standard value 22nF but that may be too large for your coil.

Hi,

The dimensions of the coil is the ones of the reference coil in  Qi V1.3 specification RX reference design page 14. Outer diameter is 32 mm and inner diameter is 21 mm. The Amplitude modulation looks pretty strong. The picture above is VRECT during normal transfert. When we face pulses of 50 ms every 500 ms I do not see back modulation from RX to TX. My communication capacitors are 22 nF.

BR/

A.

Here above, the back modulation from the RX with our 22 nF communication capacitors.

BR/

A.

The modulation looks a little deep and smaller communication capacitors may help.

Hi,

I have replaced the 22 nF communication capacitors by 10 nF. The modulation depth has decreased from about 30% to 20%...it is not better.

It looks like an illegal packet and it may cause (according to the standard) a power drop.

The measure I did is at the receiver coil level.

What is the optimal ratio between the carrier and the modulation?

Should i reduce again the communication capacitors?

BR/

A.

Hi

"It looks like an illegal packet and it may cause (according to the standard) a power drop." -- Bill J - This looks like the Signal Strength Packet (Header 0x01) the first packet sent at start up (SS, ID, Config).  If the TX has an issue with parity or check sum bit it will stop power transfer and restart.

Looks like this is what is happening with this one.   

Alain RHELIMI said:

What is the optimal ratio between the carrier and the modulation?

This modulation still looks high, Your Cs is 40nF and comm is 10nF about 25%.  May want to try 5nF to 7nF.

Can you send a scope capture of the RECT pin?

Below is typical start up 

3125.bqTESLA FAQ--Start up.pdf

Hi,

Please find above the startup at the RX coil level

Above the full waveform

The message sent before the power drop.

I have already sent the waveform about VRECT pin, there is 6 days ago.

I will replace the communication capacitors from 10 nF to 6 nF.

BR/

A.

Thanks for the scope capture, this does look like a communications problem, the TX is not decoding the first packet at start up and stops power transfer.

Thank you very much. That was my conclusion as well, but by reducing the communication capacitors, I risk further reducing the back-modulation to the TX, which will result in a lower signal-to-noise ratio (SNR) and could exacerbate communication problems. it seems counter-intuitive, but let's give it a try.

I have decoded the message sent to the TX. It is a SIG message containing 112 (0x70). No parity error, no checksum error. The message informs the TX that 43% of expected Umax is received. The message is not illegal (Qi specification v1.3, communication protocol page 118). TX  seems having a decoding issue.