LM5176: LM5176 as a supercap charger

Hi Ilhan,

Sorry to hear that you are having issues with the LM5176.

Regarding the schematic I have a few questions/suggestions
1. Please try reducing the slope capacitor to 820pF and reduce the current sense resistor to 40mOhm.
2. Add 22uF to the output of the converter this should be on the same node as the ISNS- pin
3. Can you please provide the waveforms of the switch nodes when the device is operating.

Please let me know if you have any questions.

Thanks,

Garrett

Dear Mr. Roecker;

First of all thank you for your reply.

During bring up the board was completely failed, i think addition to LM5176 some extra components has already gone.

I forgot to say, while trying to get it worked the temperature of the switcher was more than 100 degree, i dont think it is normal. The other heating component was D5 boost diode.

As regards to your recommendations, i think changing sense resistor with 40mR instead of 50mR and decreasing slope cap would not be useful.
By decreasing Rsns the signal will be much less immune to noise. By decreasing slope compensation switcher get closer to voltage mode. If you insist on i can repair the board and measure the gate waveform?

What do you think?

Regards

Hi Ilhan,

Please repair the board so that you can provide the waveforms of operation. My advice to test the LM5176 by itself using an power supply and an e-load to test for stability. If something else is damaged on the board it is possible that the LM5176 was damaged as well.. This is one reason that the LM5176 is running hot.

Thanks,

Garrett

Dear Mr. Roecker;

Finally we have assembled a new board and still have the same problem. I did your changes as well.

I will summarize what i saw like this:

*I put 1R instead of bead and when i turned power on everytime this resistor blown-up. Every time i changed it and restarted to test.

**I looked at PCB with thermal camera and everytime LM5176, D5 and of cource the mentioned 1R resistor were like hell.

***I know it makes no sense but still i took  the scope snapshots of gate waveform as attached. Yellow-buck top gate, Blue-buck bottom gate, Purple-boost top gate, Green-boost bottom gate. The strange thing i think is that when initially startup it should work in buck mode so Boost top Fet should be fully on and bottom Fet should be fully off, am i right?

It seems it is impossible to test and make measurement without burning something out. What is your recommendations?

Shall i pass one another Buck&Boost chip, i do not have too much time.

Hi Ihan,

D5 should have a higher current rating, please try a diode with a 1A current rating. This diode provides a high peak currents to charge the boot capacitor with current. Please try this first.
What is the power rating of the 1ohm resistor on the output?
When start-up is happening is the output voltage pre-biased with a voltage?
Another IC suggestion is the LM5175 that is DCM capable. It is pin to pin compatible with the LM5176, however the RT/SYNC pin resistor will need to be modified per the datasheet to match the desired switching frequency.

Thanks,

Garrett