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.

LM2673S-5.0

Other Parts Discussed in Thread: LM2673

I have a DC/DC converter using a LM2673S-5.0. Input is 24Vdc +/- 5% and it works in continuos mode for output currents above 380mA. The design was based on a webrench design. My application requests current from the 5V rail at diferent levels depending on the function in progress. At start-up is requests only 25 mA.

For low current levels, below 380mA the device enters a non continuos mode and this generates some unwanted EMI.

Can I change my design to get countinuos mode al lower currents ?

For info, the LM2673S-5.0 is using a 22uH inductor, a 120uF elco and a 4,7uF X7R input capacitor along with a 100n near the circuit pins.

Best regards,

Pedro Oliveira

Mezão Telecomunicações e Electrónica, Lda.

  • Hi Pedro,

    Above what current do you want the device to stay in continuous conduction mode?

    Marc

  • Hi Marc,

    Thanks for your reply.

    In my application, the 5V supply will have a minimum current consumption of 30 mA, so 25 mA will be fine.

    Best regards

    Pedro Oliveira

  • Hi Pedro,

    I took a quick look and to get to a DCM threshold of 25mA I would need a 220uH inductor.  This moves the LC double pole to a very low frequency.  I might be able to get the adjustable version stable by adding a Cff network .However this will take me a couple days to order boards and some time to check the design.

    I can do this but it might be easier to solve the EMI issue directly by looking at board layout.

    Can you tell me how the part is failing EMI, and send me the part of the board layout with the LM2673?

    Alternatively I can suggest some parts that use a synchronous low side switch and do not enter DCM but always stay in CCM. 

    Regards,

    Marc

  • Hi Marc,

    I agree that looking at the layout is the best next step. The board has a 5V and a 12V converters and gets 24V input. The 5V converter, U2, is on the left.

    Let me explain to you in more detail the problem I found:

    I am working on a lab with a FM radio on the back wall. If I take my osciloscope probe and measure the PWM output the radio on my back starts emiting noise. The module has not been checked yet on a pre-compliance test, but I am afraid that this will cause a problem.

    Maybe I am concerned with no problem at all. After all, in normal operation, no one is going to put a probe on the PWM output. Please check the layout attached. I made a rather long connection between the 5V output and the feedback input, but the track goes along a ground pad. please let me know if the layout is clear to you.

    Regards,

    Pedro Oliveira

  • HI Pedro,

    Looking at your design.  The 12V rail is layed out better to avoid EMI than the 5V rail.

    Here are the changes that I would recommend.

    If we look at the 12V rail.  The loop made by the C1/C2 input caps and diode D2 is nice an tight.  The inductance for this loop directly affects high frequency ringing on the switch node.  On the 5V rail notice that the diode D3 and input cap C3 is large. I would copy this part of the 12V layout onto the 5V layout and shorten the loop.

    On both of the Switch nodes should be thick until it reaches the inductor but you do not need the extra copper that is going past the pad.  This copper can capacitively couple to near by objects.

    In terms of the noise you induced on the radio.  Adding the scope prope probably made the loop length long enough to make bad antenna.  The higher ringing frequencies on the 5 v will be reduced by improving the input capacitor bypassing.

    Regards,

    Marc

  • Hi Marc,

    Sorry for the delay on my response. Your comments are great. I will change the layout and send it to you as soon as it is done, so you can comment again.

    Regards,

    Pedro Oliveira