LM3409: About Minimum ON-time

Hello Yusuke,

The LM3409 does not pulse skip.  This is the leading edge blanking.  It gets shorter  as the device operates at colder temperatures.  If the current during the switch turning on goes to or above the peak current trip threshold after the leading edge blanking expires the LM3409 will turn off the MOSFET.  Current during MOSFET turn-on can be monitored on the  current sense resistor to see if it is an issue.

Best Regards,

Hello Irwin,

I appreciate your kind response.
>If the current during the switch turning on goes to or above the peak current trip threshold after the leading edge blanking expires the LM3409 will turn off the >MOSFET.  Current during MOSFET turn-on can be monitored on the  current sense resistor to see if it is an issue.

I report to the customer to check ”current during MOSFET turn-on”.
However, customers are concerned about the following in the datasheet:
Therefore, Ton min minimum value information is needed as a reference.

Could you give me some information on "Ton min minimum specs or individual dispersion information" to convince the customer?
My apologies for asking so much of you.

Best Regards,
Yusuke

Hello Yusuke,

I have the question out to a coworker.  Generally the number used is the maximum to determine what the maximum switching frequency for a design can be.  External leading edge blanking can be added via a capacitor.

Best Regards,

Hello Irwin,

Thank you for your support.
I'm sorry, but I do not understand what you said.
In the first place, we have to reduce the Rsns noise.
Enlightened the customer about the root cause.

However, there is a phenomenon that the pulse width becomes smaller when the IC cools and Ton min becomes smaller.
Therefore, I would like to investigate the relationship between the problem and Ton min.

I will send the customer tested waveform by direct message.
Could you please give me your advice on the relationship between the sent waveform check and Ton min?

Best Regards,
Yusuke

Hello Yusuke,

What is the MOSFET being used?

What is the freewheel diode being used?

What is the inductor being used?

What is Viadj set to?

Is there a gate resistor?

The turn-on time seems long and at high current.  There is a lot of noise and ringing present where the peak current trip occurred right after the leading edge blanking time.  What is being used for Rcs (part number and value)?  It also appears there is another switcher running causing other noise on the current sense signal.  How is the layout with respect to the on-time loop and off-time loop?

This is a common leading edge current sense issue.  The first time it tries it reached peak current and shuts off, the inductor current drops more and the following turn-on can happen since the inductor current is lower.

I don't know if the information you are asking for is available, the person that knows this should be back from vacation today.

This can be improved by adding additional leading edge blanking though there is an issue with this design, either part choice and/or layout.

Best Regards,

Hello Irwin,

I am grateful for your response.

>What is the MOSFET being used?
>What is the freewheel diode being used?
>What is the inductor being used?
>What is Viadj set to?

I report all your advice to the customer.
However, the parts used in the current evaluation are unknown.
I will get the latest circuit information and report it to you.
Gate resistance is 0Ω. I requested the customer to test the 1Ω installation. However, the noise did not change.

>I don't know if the information you are asking for is available, the person that knows this should be back from vacation today.
Please contact me again once you have some updates.

Separately, I will send you the additional waveform obtained by email.

>This can be improved by adding additional leading edge blanking though there is an issue with this design, either part choice and/or layout.
I have the same opinion as you.
However, the noise of Rsns voltage had the same waveform at room temperature and low temperature.
Customers are investigating why LM3409 are more susceptible to noise at low temperatures.
When the LM3409 is cooled with a point cooler, the changing specification is Ton min.
I would like some advice on this matter.

Best Regards,
Yusuke

Hello Irwin,

Thank you so much for your sincere correspondence to my requests.
>What is the MOSFET being used?
>What is the freewheel diode being used?
>What is the inductor being used?
>What is Viadj set to?

I have sent you the latest schematc and additional waveform information via email.
This schematic is the schematic of the board where the most problematic phenomenon is confirmed.
Customer studied the countermeasure parts centering on the board where the phenomenon occurs most frequently.
I have contacted the customer for all the advice that layout measures are necessary for the root solution and your advice.
However, customers are only able to consider adding parts and changing measures.

Customer is going to proceed with the development with the measure in ③.
Furthermore, although the diode change addition measure in ③ is noise exceeding VCST, noise is suppressed within 80 nsec(Minimum Ton min spec).
This measure eliminated the phenomenon that the pulse width became smaller. 

As a final check, I would like to hear your opinion on the validity of this measure.


Best Regards,
Yusuke

Hello Irwin,

Thank you for your response.
>I do not see that test in the data you sent.
Yes. Customer did not get the waveform for that test.
However, “test 3 with only one of the capacitors” could not suppress the noise.
Problem phenomenon occurs depending on the variation of Ton min.
The capacity of one capacitor was gradually increased(0.1～1uF), and the problem could be solved by combining Vin-Vcc:0.47uF  and Vcc-CSN:1uF.

>It should be okay regardless but it if works with only one capacitor it is one less part.
Is there a problem with a waveform that uses two capacitors?

Customer wants to get approval to install capacitors on Vin-Vcc and Vcc-CSN.
Is there anything else we should check other than the issue of reduced pulse width?


Also, Could you please update any of the following items if you have additional information.
">I don't know if the information you are asking for is available, the person that knows this should be back from vacation today.
Please contact me again once you have some updates."

Best Regards,
Yusuke