TPS43060: Internal VCC loading

VCC Waveform

Hi

Thank you for posting. If you just read 21 ohms from LDRV to GND, that means your IC is damaged.  Sorry the link to your circuit cannot be opened.  Could you repost your schematic?

Thanks,

Youhao Xi, Applications Engineering

Hi,

Please find the snap below. But I think its not legible, let me know your email ID, I will send it across.

Thanks and I can read the schematic.  I will study this and get back to you by COB tomorrow my time zone.

Thanks,

Youhao

Hi,

I studied your schematic and there is no obvious problem to explain the issue, except that the boot diode does not appear to be an appropriate selection.  Firstly you don't need 3A diode in SMB size for a boot diode, and secondly this diode has too much leakage current which may cause overstress to the VCC pin. 

Could you check your boot diode on the same problematic board, and check the diode's leakage current at your highest ambient temperature?

Thanks,

Youhao

Hi Xi,

Sure I will measure the drop across VCC series resistor to give an idea as to what could be reverse leakage current of the diode at highest ambient. My question would be that the failure happened at room temperature and also the leakage current of selected diode could be around 1mA at junction temperature 125degree, would it still load the VCC?

Hi Ashwini,

What I worried is the overvotlage at the VCC pin that may overstress VCC.  Loading does not seems to be a big concern.  Since VCC itself is generated by an internal LDO, it will not overstress the VCC pin.  However, the external circuit can.  The boot diode is one factor that we need to check.  Please check the diode on the damaged board.

Thanks,

Youhao

Hi,

The damaged board doesn't show anything at all The high side MOSFET is not being driven so no bootstrapping occurs.

Also we found that adding Rcharge resistor (in series with boot diode) helps and VCC on healthy board no more appears loaded while drawing full 2A current at the output.

However the datasheet, application note or EVM doesnt elaborate on the value, size, package selection of the resistor!

Do you have any literature on how the value and package is selected?

The resistor selection depends on the high side MOSFET that you choose. It is basically a RC charging circuit repetitively charging the boost capacitor to replace the gate charge of the high side FET each cycle. Let me check with colleagues if there is any literature or app note existing to share.  I will get  back to you by next Wed.

I also checked your MOSFET selection and it has a too much gate charge and not suitable for your 24V@2A load application. It consumes too much driver power.  You should choose a 30V or 40V rated MOSFET having the similar RdsON and much smaller gate charge to get better performance.  

Thanks,

Youhao

Hi Youhao,

Thanks for the input. I rechecked the MOSFET parameters and looks like 73nC is too high for the 24V,2A application.

What would be a recommended range? Filtering on the basis of balance between Rds-On and Qg(tot) we looked at following options:

BSZ070N08LS5ATMA1

NVD5C464NT4G

RD3G400GNTL

We understand that we will have to change boot capacitor as per new MOSFET selection Qg and ripple expected.

Let me know about literature for Rcharge value and package selection

Regards

Ashwini

Hi Youhao,

Thank you! I will have a look.

Any thoughts on MOSFET selection?

Regards

Ashwini Singh

Hi Youhao,

Even on a healthy board, it is observed that the internal VCC decreases from 7.5V to 6.3V when circuit is loaded with 2A load. We suspect it is due to MOSFET only but could not test until we receive the MOSFETs. The circuit still is regulating voltage correctly works but can it be one of the reason why the IC might eventually fail?

Regards

Ashwini

Hi Ashwini,

So far we cannot conclude this is the cause for the damaging.  Let's wait for your results with new MOSFETs and see if we can get some hint about the problem.

Thanks,

Youhao