TPS63000: TPS54160A and TPS63000 instable output after some time.

Hi  Fabio,

Since this phenomenon change with input voltage, i think the root cause should be buck and i will forward this thread to right tem.

Regards

Tao

Hi Tao,

thanks for your anwer.
We did a test over the weekend wit 32V Input Voltage and the behavior also occurs. But it took more time.
So my initial post was not correct.

Regards
Fabio

Hi  Fabio,

Got it,  from tps63000 side, could you replace inductor with recommended 2.2uH?  What is the status difference between stable and instable, just different Iout?

Regards

Tao

Hi Tao,
have to check if we have a smaller inductor, otherwise I'll order some.

There is no difference in the load current.
The setup for the Manufacturer is as following.
Connect Power 41V and check the Current consumption of the whole board, and measure the Voltages on 3V3 and 5V.
In production they observed this behaviour, but measuring with a Mulitmeter the output Voltage is in spec. But leds are flashing and the communication with the Board was not possible.
So I got some Boards for investigation and I've measured with an Oscilloscope. 

After a while (30min to 90min) the output behaviour changes.
on the 3.3V converter ther is an LED which is blinking with 0.5Hz (1mA)
on the 5V there is a LED constant on (3mA)

If you need the full Schematics, I can provide it in a privat message. 

Hi  Fabio,

Is it possible your can cut off the 5V output of TPS54160, Then power the BUBO with a stable power supply?  

Meanwhile, use a resister with similar power consumption with BUBO part as the load of TPS54160 output.

Regards

Tao

Hi Tao,

We will test with a smaller coil as suggested and also test with an external power supply and artificial load.
I was not able to test your suggestions today.
I appreciate the suggestions and will get back to you once I have tested them.

Kind regards
Fabio

Hi  Fabio,

Got it, will keep the status of this thread as "waiting for customer" till you next update. 

Regards

Tao

Hello Tao,

We have tested as you suggested.
It seems that the behaviour is gone with a smaller inductor. Currently we have used a 3.3uH type. But 2.2uH and 1.5uH are on order.

In our test setup, every 5 seconds the NFC reader is activated and at startup this part of the circuit consumes 140mA @3.3V.

The error slipped in because we wanted to use a TPS560200DBV for the 3.3V in the initial design. However, there were supply shortages during the development period and we had to switch to the TPS63000. We also simulated the TPS63000, but with the originally proposed 10uH inductor.

We were unable to generate a bode plot for the BUBO in webench.
We should be able to explain to our customer why we did not detect any errors with the 10uH in the first batch.

Can you simulate the Bode plots or the stability conditions for this system with 10uH and 2.2uH?

Thanks for your hints and support.

BR Fabio

Hi  Fabio,

Yes, i can do this. Test condition is 5V input and 3.3V/0.4A output, right?

Please expect this test will have a big delay because i have two urgent case on hand, sorry for that, late this week is ok?

Regards

Tao

Hi Tao, 

yes, 5V input and 3.3V/0.4A output are correct.
end of this week or early next week would be great.
Thank you very much for your time and your willingness, I really appreciate it.

Regards
Fabio

Hello Tao,
I hope you are well. I wanted to ask if you will be able to simulate the stability conditions for us by the end of the week.

Best regards
Fabio

Hi Fabio,

Really sorry for that, i try to do bench test to confirm the difference of Bode Plot with 10uH and 2.2uH this Monday. i forgot the detailed conditions and i want  go back to this thread to find out. Then, i can not find this thread any more, do not know what happened, may be i made a mistake or not carefully enough, i thought you already have answer about this and delete the thread.

Unfortunatly, i will on my holiday and will not back to office till next week. i am afraid i can do nothing to make up my fault.

But as soon as i back to office, will do this test firstly.

Regards

Tao

Thank you for your feedback.
Enjoy your holiday.
We look forward to receiving a result when you are back.

Regards
Fabio

Hy Tao,
welcome back.

Many thanks for your support.

This is probably not your responsibility, but a note in the data sheet or in the webench regarding the inductor value would be useful.
We have found that as long as you select components from the catalogue, no warning will appear; if you create your own, you will be warned.

Best regards
Fabio

Hi  Fabio,

Thanks and apologize again for my mistake.

I do not know is it a good idea to have such remind for each one. But one way which will 100% work is follow the user guider of EVM if no special requirment because the parameters of external components are verified in our side.

Regards

Tao