TPS53355: Input variation characteristic

Hi T Kishi-san,

Let me look into it further.  I have not experience this behavior before.  In the mean time, would you please send the schematic?  Would you also send some switch node waveforms during before, during, and after this VOUT change happened? Does this behavior occurs at room temperature?

Thank you.

Amnat

Hi Newey

Thank you for reply.
This is the circuit diagram. (Basically, the constant of EVM is changed)

TP53355.pdf
The voltage fluctuation fluctuates slightly at about 8.5 V at 25 °C.
When the voltage fluctuation is 25 ° C.
At -40 ° C, it varies greatly with Vin = 5 V to 11 V

Regard

T Kishi

Hi Newey

Thank you for responding.
I and the customer feel that there seems to be a cause for　ripple Injection circuit.
So I have a question.
1. Datasheet P24, "N" in equation (11) is a consideration of component variability?
2. [Ripple Injection Voltage <50 mV] is recommended, but is it necessary to keep this? If you need to protect, you have to set a very severe constant.

Please continue to cooperate in solving problems.

Best Regard
T Kishi

Hi Newey

Thank you for reply.
The red frame is a snubber circuit.
Even without this circuit, the phenomenon does not change.
Because it occurs particularly conspicuously at low temperature, can you check it?
As for equation (11), is it not necessary to consider errors of L other than DC Bias?

Considering the error of the worst case of each part in equation (11), when trying to keep the ripple injection voltage at 50 mV,
the ripple injection voltage becomes about 100 mV in the normal case.
How should we determine constants?

Regard
T Kishi

Hi T Kishi-san,

1. Yes, the N value is higher because the effective output capacitance is reduced with higher DC bias.

2. We usually inject about 15mV ripple to the FB with all ceramics output capacitors.

One question, what is the purpose of this circuit (red circle)?

Thanks,

Amnat

Hi T Kishi-san,

Thanks you for the clarification.  We are in the process of testing your condition now.  We'll let you know as soon as the data is available.  As for equation 11, the N coefficient of 4 has already taken the L and Cout variation into account.  The higher N value gives you enough margin for stability.

Thank you.

Amnat

Hi Newey

Thank you for your response.

If we assume that the ripple injection voltage is about 100 mV,

we think that it relates to the accuracy of the output voltage, but what about other problems?

Regard

 T Kishi