This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM3429: LM3429 current spikes

Part Number: LM3429


this chip when one uses the appnote, delivers Heavy but short current spikes to the load.

When running at 12Vin that is not an issue since the pulses only come at abt 40kHz.

But when running it at 6.5V, which one appnote says should be possible, then the current

spikes arrive at a frequency of abt 500kHz. 

This means that the heating of the LEDs becomes significantly higher and is a hazard.

My question is this:

- is it possible to Place an cap of 22uF 25V X5R after the sense resistor and to GND without¨getting any

problem with the constant current Control? This function uses the voltage over the sense resistor

to Control the current.

This is urgent, thanks!

  • I do appreciate ypur fast and accurate response!

    Also the phone line in Sweden is blocked, no one returns your calls.

    So, HOW on Earth do you think we designers can have any faith at all in TI?

    For me, it is over..

  • Hi Zacharias,

    I am sorry but I’ve been answering all your questions as best as I can from the E2E site.

    I do understand some of your frustrations but from my answers to your previous posts. We need schematic along with load requirements and scope captures in order to help you. There is also some fundamental understanding in designing power circuits that you also need to understand.

    Thanks Tuan
  • HI, it is clear that I know more about these things than people on the E2E- Scoping the signals I can see that those transients are delivered

    to the LED load and that causes a heating. When running at 7V the heating becomes excessive and when running at 12V there is

    virtually no heating at all. Diffrence is the the spikes that are delivered at abt 40kHz at 12V and at about 500Khz at 7V.

    The design in the appnote for LM3429 was used. Copied 100%. The use of an electrolytic cap to remove 500kHz ripple can

    most benevolently be said to be ignorant.

  • I'm also most surprised that you did not know that PRF has since times immemorial been used as

    an acronym for Pulse Repetition Frequency.

  • I have requested exact schematic, load voltage current requirements along with scope captures explaining your exact issue and only will be able to help you if you provide those to me first.   Now, I am also asking for the layout along with part list.   If we do not get these required things then I cannot not help you.

    I’ve been designing power circuits and regulators for over 30 years and have not heard the term PRF being used in regulators designs and still at lost here for your questions.  PRF is as you explained is pulse repetition frequency which is generally used in radar.  For our LED controller, there are 2 things that gets PWM.  One is the switching frequency of the power FET/inductor and the other is the PWM’ng of the output current to the LED.  Which one are you talking about here related to our part definition?

    Thanks Tuan

  • as I have stated the design is a copy of the appnote for Boost. The very same component value are used.

    What can be seen is that the oscillations frequency goes up when the Vin lowers. At12V it is around 40kHz and

    around 7V is abt 500kHz. The latter casues a heating of the leds that is unacceptable. Since I cannot simulated the design

    I measu the finished voard. The phenomen is clearly design related. At 12V there is no heating of the leds.

    Using same board, same design, same all.

  • Once again, we will not be able to help you troubleshoot your circuit if you do not provide us with what we need and requested above.

    I will close this post until you provide us the the requested information.

    Thanks Tuan

  • Hi

    I take it that your appnote is flawed. We have used it Verbatim. Every Component every Little twist is exactly as per

    ypur appnote in the data sheet of the LM3429.It shows that when running at 12V it is just fine, but when going down to 6.5V

    the switch frequency goes up to abt 500kHz. 

    Probably this is quite normal with this chip BUT it damages the LEDs by heating them up too much.

    Shall I copy your appnote from the data sheet here? I could do that if mandatory.

    The very idea of putting elecrolytic caps as decoupling around 500kHz does seem to be a  kind of university joke.

    An experienced designer would never do that.

    BTW, how can I go about to simulate the Buck.Boost design? The boost design has error in your schematics

    but is there a working schematic for the B-B that follows your schermatics? I would try it out if so.


  • You are assuming a lot of things here and we will not comment on anything until we receive what we’ve been requesting from you.  Troubleshooting with words and assumptions will not resolve the issue.

    Thanks Tuan

  • So, it is flawed. Thanks for that confirmation. It goes hand in hand with the test results.

    And it IS the very same design that is in your data sheet. Same compoments and values.

    So, I will recommend my customer to swtich to LT3477 since I don't get any help, just

    troublemaking and insinuations.

    Also, the simulation does not work which makes the situation even more difficult.

    But I have resoved the immediate needs and after that it is the LT3477 that will be

    the component of choice.