TPS5430: Output current at turn on

Hi Andy,

Is the OV circuit of the block diagram average measurements?

Regards,
PAN-M

Hi Andy,

Is the OV circuit of the block diagram average measurements?

Regards,
PAN-M

Hi Pan

No, the device will stop switching if OV is triggered, the current is limited by high-side current peak.

B R

Andy

Hi Andy,

Pls. teach me a little more.
Are an inductor current and High-side MOSFET current to say that is different?
Though an inductor current reach at 5A in the waveform, the switching does not stop.
Is this to say that load is constant and is not an OC state?

Regards,
PAN-M

Hi Pan,

Inductor current peak is same as HS switch current peak, so PCM control circuit sense HS current to monitor load current.

When current peak is trip, converter will regulate the inductor current like a constant current source, output voltage is determined by the load resistance,

B R

Andy

Hi Andy,

I am confused.
You said that it did not have any problem as for you because a mean current was 2A first.
However, I think that it is just trip before as the waveform of the customer is 5A if it trip by peak current.
Is there the fear to repeat Hiccup mode when load was beyond over current trip point?
Is this correct?

Over current.pdf

Regards,
PAN-M

Hi Pan

TPS5430 data sheet had no detail explain or spec. on how to enter Hiccup mode. for example, how may cycles to trig the peak current limit? or FB drop to what level of Vref?  

Of course there is risk to repeat Hiccup OCP if current limit is triggered. customer should increase the soft time of Q5.

B R

Andy

Hi Andy,

I heard that TPS5430 was in Hiccup at 2 cycles from our co-worker.
Some other device are 7 cycles.
There was a reply from customer that a state did not have a change even if I change a capacitor of Q5.
I think that 5A is load of the system when I see waveform as both the switching and the inductor current are normal, not inrush current.
In this case, may it be necessary to change the device to TPS5450?

Best Regards,
PAN-M

Hi Andy,

Is the waveform at boot up of the customer an OCP state?
I think this OCP to become the Hiccup state.
Does it not become the Hiccup state during SS state?
As the maximum internal SS time is 10ms, I think that it becomes Hiccup after this time.
Is Hiccup not effective until it arrives at the setting output voltage?
In that case, as this device boot up while hanging a current limit, I think that even this waveform does not have any problem.
Please teach me whether you do not have any problem even if they mass-produce this.

Best Regards,
PAN-M

Hi Pan

On the waveform, the inductor current peak reach 5A peak limit, this should be caused by too big output capacitor or bias load current when Q5 turn on, there is potential to trig the Hiccup OC protection, so it's better to use TPS5450.

another solution is to place gigger cap on C35 side, smaller cap on C32 side.

B R

Andy

Hi Andy,

What means "gigger"? "bigger"? or "jigger"?
Specifically, what should we do?
C32 has a small capacitor C36, is there a problem in this?

Regards,
PAN-M

Hi Pan,

sorry for typo, it's "bigger", not jigger.

Small C36 parallel with C32 is OK,  but C34/35 are much smaller than C32 will cause inrush current(inductor current trig peak current limit) during Q5 turn-on, 

Suggest solution:

1. Increase Q5 soft turn-on time

2. Use TPS5450, or newer LMR14050

3. redesign circuit, increase C34/35, reduce C32

B R

Andy

Thank you for support, Andy,

Does the condition not to enter Hiccpu during boot up know what it is?
I think that close is done if we understand this cause.

Best Regards,
PAN-M

Hi Pan,

During TPS5430 soft start time, the inductor current is far below current limit as the Cout is quite small capacitor of C34/35,  after some delay after the soft start is finished, the Q5 will turn on, TPS5430 will see heavy load as the voltage on C32 is zero, it has potential to trig Hiccup OC protection during this stage. I recommended above solutions to overcome this risk.

Hope this make things clear.

B R

Andy   

Hi Andy,

Thank you for your support for solutions to the problem.
Our customer repeated trial and error and found out the best change.

TPS5430_current_0616.xlsx

As Hiccup does not occur in SS state, I think it seems to be no problem at the first overcurrent state.
What do you think about this waveform?

Best Regards,
PAN-M

Hi Pan,

Yes, this is more better solution to reduce inrush current by enlarging TPS5430 SS time with C31, it also can be improved by adding resistor divider on EN pin to startup at Vin=8V~10V.

B R

Andy

Hi Andy,

I transferred your comment to our customer.
I wait from the correspondence of our customer.
In addition, I will inform you if I have any question from customer.

I appreciate your support.

Regards,
PAN-M