TPS62A01: Startup / Stability / Hiccup Problem

Hi Cedric,

I am not able to clearly read the schematic you shared above. Can you please share the schematic as an image attachment? I want to confirm the part numbers of the passive components you used. Also, Is there a ferrite bead at the output? 

In addition, Could you also share the waveforms of Vin, Vout, Vsw and the Inductor current when you observe the above mentioned issue?

Also, please let us know what is your end application? 

Thanks a lot!

Best regards

Sneha 

Hello Sneha,

you can save yourself the image or open it in a new tab.

Its definition is good enough to zoom in and clearly see the different part references.

Kind Regards,

Cédric

Hi Cedric,

Oh yeah. I missed it. My apologies. Thanks for letting me know.

The critical issue I observe with your schematic is that the overall capacitance (200uF) at the output is way higher than the recommended output capacitor for the device. Please refer below table for the recommended LC combination with which the device has been optimised to perform with. 

In my opinion, this would be causing loop stability issues in the device. I would recommend to check the device behaviour again after reducing the output capacitance to recommeded limits.

Can you also share the schematic for 1V and 1.8V Vout rails? 

Is there any specific reason of using large amount of capacitance at the output? What is your end application?

Thanks!

Best regards

Sneha 

Vin:

Vsw with 500us period:

Vsw(zoom in) 1/2:

Vsw(zoom in - startup) 2/2:

Vout:

Hi Sneha,

for 1V and 1,8V as output we use the same schematics apart from R31 in the feedback voltage divider.

With 1V and 1,8V there is absolutely no problem even with much larger capacitors like 400uF and more! The system is starts and is stable with 1A load.

The TPS62A01 has a built in soft start that doesn't overcome a load of 50mA and 200uF for a 3,3V output when using a 120pF feedback capacitor (C35).

This is a big problem!

When this feedback capacitor C35 is increased to 1nF or 2,2nF, the system starts and is stable but with a total load on the 3,3V of less than 200mA.

We repeated this with a TPS62A02 instead of the TPS62A01 and the system starts and is stable until 750mA load.

This is far away from the expected 1A with the TPS62A01 and 2A with the TPS62A02!

Thank you in advance for taking care of our problem and making efforts accordingly!

Kind Regards,

Cédric

Hi Cedric,

Thanks for sharing additional information. Please be informed that I am working on your case and will give you an feedback by end of this week. 

Thanks for your understanding.

Best regards

Sneha 

Hi Cedric,

I had a discussion with my team here regarding your issue. It looks like there are chances of the following behaviour in the device.

- With the large amount of capacitance at the output, there will be a huge current requirement to charge the capacitors during startup. This is causing the inductor current to hit the current limit of the converter and enter the HICCUP mode. 

- Also, this behaviour is observed with 3.3 Vout because of the longer ON time (higher duty cycle) and it hits the current limit even with a lower load current (~ 50mA). However, with 1.8V and 1V output, the on time is shorter(less duty cycle) and the inductor current does not hit the current limit. When you increase the feedback capacitors, it is making the loop response faster  and the system remains stable with slightly larger load current. But it cannot support the full load capability of the device.

- As TPS62A01 has been tested and validated for its optimal performance with the recommended LC combinations mentioned in the datasheet only. We cannot guarantee the device performace outside of datasheet specs. Therefore, I would suggest to check the device behaviour with the recommended output capacitors. In case your system requires such a large amount of capacitance, you can choose TPS62A02 (as this has higher current limit capability). However, I would recommend to check the loop stability of your system with large output caps for TPS62A01/02 both. 

Thanks!

Best regards

Sneha 

Hello Sneha,

thank you for your discussion. You are confirming what we experimented.

I communicate our test results with different combinations for other customers that might be interested in:

- 5V input / 2.5V output with TPS62A02 /  Cfb=2.2nF : startup and behaviour ok with a load current less than 1.5A (output capacitors ~ 150uF)

- 5V input / 3.3V output with TPS62A02 /  Cfb=2.2nF : startup and behaviour ok with a load current less than 1.1A (output capacitors ~ 400uF)

My impression regarding this step-down regulator is that the overcurrent protection is a bit pessimistic and limits its capabilities.

Anyway, we agree on the point that we needed to test our setup in order to validate it :-).

Kind Regards,

Cédric

Hi Cedric,

Thanks for sharing your test results. Do you move forward to use this device in your application?

Could you also share more details on your application? We just want to understand what kind of application has a large output capacitance requirement. Customer's feedback is valuable to us during product definition phase and help us defining features accordingly.

If you could share your desired specs for the device in your application, I could also check for other possible alternatives here. Please let me know your thoughts.

Thanks!

Best regards

Sneha 

Hi Cedric,

Thanks for the confirmation. 

Regarding the supply, I would need to check with my Marketing team here. Let me check if I get information about these two devices and will update you by early next week.

Thanks!

Best regards

Sneha