NXQ1TXH5 init before charging

[tomaszbadyla]

Hi NXP,

We are in assembly process, and during testing we have found following issues.

Sometimes docking start charging after couple second 3-20 seconds. I dont know whats causes this problem. 20 s is to long for us. On different docking it works fine but some of them have this issue.Can component tolerances influance init process ?

Best regards,

Tomek

[tomaszbadyla]

Hi Tomas,

Last week i've done final tests with grounded crystal (soldered wire between crystal casing and gnd).

20 faulty docking now starts in less than 3s.

There was a lot of problems connected with this issue . Until now we have 300 fault docking station. We have to fix them all. For the next batch we gona use shielded crystal.

I hope you will add information to DS that using shielded crystal is obligatory, hence nobody else get into trouble.. In small volume problem didnt exist, but at production stage it was big problem. 

Regards

Tomek

[TomasVaverka]

Hi Tomasz,

Here is an additional comment from our application engineer:

The grounding of the crystal is already advised in the application note, but it is up to customers to decide what to do. Further we did not have problems with it in our demoboard.

Regarding your comment on the distance, 2mm between transmitter and receiver coil is not much: it is the minimum distance that is given in the Qi specification. When the distance becomes smaller, then the maximum transferred power goes down, but the efficiency increases. So smaller is not always better.

Please let us know in case of any other questions.

Best regards,

Tomas

[tomaszbadyla]

Hello,

Ok i've tested 80 docking station and i have found 10 with misalignment issue (TX coil isnt properly centered on ferrite plate) another 10 docking still has some problem with initialization. In this case i've tried with shielding foil covering all

PCB and it works fine starts immediately. I would like to add some shielding tape but i dont know what part of the circuit is responsible for initialization process. Picture below show how coil and pcb are sets together in our product. In different situation when PCB and Coil lays next to each other everything works fine. So PCB and coil setup can influence initialization (delay init process). One more thing i've designed PCB according to all your rules from AN. 

[TomasVaverka]

Hi Tomasz,

There is not one specific part of the circuit responsible for the initialization process, the important thing is that the communication is not disturbed.

Distortion from the crystal oscillator by the magnetic field of the coil is a good explanation of the startup problems as we have seen it before. I should have mentioned this in the list.

The problem occurs when the crystal is directly facing the magnetic material of the coil. A simple solution to prevent the problem is to flip the PCB so that the (copper) bottom side of the PCB is facing the coil as in the picture below:

Now the voltage on the Xtal pin is a perfect sine wave with a frequency of 32.768kHz:

Another possibility to reduce the susceptibility of the oscillator circuit to the magnetic field of the coil is to use a crystal with a metal casing that is connected to the ground.

In the DB1401 layout then G2 is mounted instead of G1.

G2 Crystal; 32.786kHz; SMT Abracon AB26TRQ-32.768kHz-T Mouser 815-AB26TRQ-32.7-T

The wires from the crystal to the NXQ1TXH5 should be kept as short as possible. I do not have further details about the oscillator circuit.

Best regards,

Tomas

[tomaszbadyla]

Hi Tomas,

Thank You for response Tomas. Currently im testing by my own fault docking station. 

-distance is quite big 2mm

-i've thought smaller is better

-I will check alignment. I was assured that coil are align. 

-coils are differnet TX coil - A11 RX coil- 28mm x 15mm 7.5uH. I know that it matters, but in 80 percent of our production it works.

-i dont think that something near coils can influence charging process. I will take some picture for ref.

I havent tried with your demoboard. 

In attachement you will find video. Receiver is on the docking station and initialization time takes one minute. You can see there analog pings(500ms) and digital pings every 4s. When i place receiver again on docking it start with time less than 5s.

This issue happens sometimes, so it is hard to find solution.

I will be working on today and next week. I've partialy fixed this issue by adding capacitor between gnd and the end of coil winding. It helps, failure rate is lower in faulty docking.

Best Regards,

Tomek

[TomasVaverka]

Hi Tomasz,

The NXQ1TXH5 is doing an analog ping every 500milliseconds. During the analog ping is only detected if threre is an object on the charger.

If an object is detected, then the analog ping is immediately followed by a digital ping during which the charger is trying to get a message from the receiver and if the communication is established, the charging starts.

If no object is detected, then the charger goes to sleep again. So during 10 seconds there have been 20 attempts to establish communication, so it seems that there is something wrong there. Things that deteriorate the communication are:

- too large distance between charger and receiver coil

- too small distance between charger and coil

- misalignment of the charger and receiver coil

- different size of coils

- foreign objects (coins, wiring, metal casing near the coil)

Does the charging start within a second when the receiver is placed on our NXQ1TXH5DB1401 demoboard?

Best regards,

Tomas

[tomaszbadyla]

Hi Tomas,

Could you tell me a little bit more about crystal circuit in NXQ1TXH5. At XTAL_IN input it has 2.2pF capacitor it is huge impedance for 32kHz something about 2MOhm. So electromagnetic interference  can easily sneak into IC because for higher freq this capacitors has  less impedance than for 32kHz. Coil emits 100-200kHz high intensity field this frequency can easily influence crystal circuits. Please tell me something more about it. If im wrong correct me.

Regards

Tomek

[tomaszbadyla]

Probably the main reason for this issue is long delay between digital ping. On different Qi charger sollution delay is something about 500ms. I know that it is somehow compromise between standby current and detection rate. For us it is quite big problem because 20% of production volume has receiver detection time over 10s   (in 5 out of 10 attempts  was over 10s)

Do you know how to change this delay or what can enhance detection rate.  e.g. how to increase digital ping strength. 

Our receiver has 2.8V detection voltage threshold . At receiver site just after rectifier i have 3.3V it is 0.5V over required voltage threshold.

Best Regards

Tomek