• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 65 subscribers
  • Views 1012 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Power Supply Problem - LM43602

Jon Lister
Prodigy 30 points

I'm using an LM43602PWPT in a circuit created by the WEBench app to take in 24VDC and regulate it down to 3.3VDC. There are 4 prototypes, none of them working. One puts out 0V, one 0.3V, one 1.2V, and the last one oscillates around 0.7-0.9V.

The documentation shows a pad is required underneath the chip for thermal relief, but my company doesn't have the facilities for under-chip pads. I decided to put through holes under the part to flow solder through and connect to ground. Unfortunately the board house we use solder masked the vias so the solder won't flow through and contact the bottom of the chip. I thought it wouldn't be an issue since it's just for thermal relief, but I'm thinking it may be used for a ground connection due to the wrong outputs and the fact that the chip never warms up when the board is plugged in.

Is the "Thermal Relief Pad" actually used for a ground connection, or is there something else at play here?

  • 0
    TI__Mastermind 41310 points

    Jon,

    I am moving your post to the Simple Switcher group. They should be able to help you with your questions.

  • 0 in reply to Britt Brooks - Online Design Tools
    TI__Guru 71985 points

    The bottom pad is used for both thermal and electrical connections to ground.  So you will need to get some kind of ground connection to the pad.

    However, I do not think that this is the complete cause of your trouble.  Be sure that the EN input is correctly connected and that the output is not shorted

    by mistake.  Please send your schematic and layout and we will review.

  • 0 in reply to Frank De Stasi
    Prodigy 30 points
    Thanks for the input! I just edited the post adding the layout and schematic. Some of the parts are on the bottom layer underneath others, but I tried to get everything close to the chip.
  • 0 in reply to Jon Lister
    TI__Guru 71985 points

    The layout seems OK; can we see the bottom.  It seems to me the input and output caps would need to be larger; are those the ones that Webench choose?

    I will pass this along to the engineers that worked on this device.

  • 0 in reply to Frank De Stasi
    Prodigy 30 points
    Yes sir, this was from the Webench. The current requirements are pretty low, I assumed that's why the caps were so low.
  • 0 in reply to Jon Lister
    TI__Genius 11085 points

    Jon,

    The first thing to fix is to tie SYNC pin to GND. SYNC pin shouldn't be floating when unused.

    For 24V to 3.3V 1MHz operation, the datasheet recommended 100uF output capacitor. I only see a 22uF on the schematic. I'd recommend follow the datasheet recommendation here. If you are using all ceramic capacitors, Cff should be 33pF.

    To make the design more immune to noise, you can reduce the value of the resistor divider. R14-->100kOhm and R10-->43.2k, then Cff should be 10 times larger: C26--> 100pF. The location of the R10 is good. I cannot see the R14 and C26 on the top layer. Please make sure R14 and C26 are close to R10, so that FB node doesn't have a long trace. The trace connecting to Vout should be routed away from noisy nodes.

    If the issue is still not fixed, please carefully measure the FB pin voltage. FB pin voltage should be 1V, if not, the part is not working right. If FB = 1V but Vout is not in regulation, please check the connection from Vout to FB.

    You can also send waveforms of the SW node and Vout to us if issue still exists.  

    Best, Yang