LM5176: The LM5176 has no power output

HI Zhang,

thank your for using the E2E forum,

Note sure if I understood all point above right - this is what I understood:

So if you replace the LM5176 all works as expected.

You checked SW1 but this was not as expected. Was this with the failing unit or the working one?

How many boards are failing? How many boards are working.

The naming is very confusing - clarify:

VOUT is the input voltage of the power stage

VIN is the output voltage of the power stage

Right ?

Can you provide scope plots: ( time scale to 1 few ms and to a few switching cycles)

- VIn and VOUT

- SW1 and SW2

- COMP  and FB and SS

Could you please attach the filled out LM5176 Design Calculator to this thread? This would be the most efficient workflow to do a schematic review.
 

Best regards,

 Stefan

Hi Stefan，

Answer as followed.

So if you replace the LM5176 all works as expected.

-->Yes。

You checked SW1 but this was not as expected. Was this with the failing unit or the working one?

-->The SW1 does not work as expected. This is with the falling unit.

How many boards are failing? How many boards are working.

-->We used about 5k PCS mounted on board. Total 15pcs LM5176 failed. 

The naming is very confusing - clarify:

VOUT is the input voltage of the power stage-->Yes

VIN is the output voltage of the power stage-->Yes.

Right ?

Can you provide scope plots: ( time scale to 1 few ms and to a few switching cycles)

- VIn and VOUT

- SW1 and SW2

- COMP  and FB and SS

-->I will upload the failed LM5176' wave forms about VIN/VOUT, SW1 and SW2.

Could you please attach the filled out LM5176 Design Calculator to this thread? This would be the most efficient workflow to do a schematic review.

-->i will upload the Lm5176 Design Calculator.

Thanks a lot!

Hi,

please also check if you put a failing unit an working board - does this the fail again.

Thanks,

 Stefan

Stefan，

please also check if you put a failing unit an working board - does this the fail again.

-->Yes, We change the failed and worked LM5176 between two MB. The failess is followed as failed LM5176. 

The waveforms as attached.

副本LM5176 Buck-Boost Quickstart Tool.xls

Hi Hongbin,

this signals looks so far OK:

- SS ramps up (would be good to see the plot with a few ms to see if it stays high as well)

- VIN seems to be stable (at least for the short time shown in the plot)

- FB is low according to the output voltage

- there are some spikes on the SW - so it looks like the device tries to start but can not switch the MOSFETs

As SW1 is not going high it would be good to check next:

- BOOT1

- HO1

- SW1

- LO1

Can you also probe on a working board:

- SW1

- SW2

- the MOSFET Gate signals to check of over or undershoots which might damage the device.

Best regards, 

 Stefan

The working board wave forms as followed

As SW1 is not going high it would be good to check next:

- BOOT1- HO1- SW1-LO1

-->I will test the signals tomorrow.

Thanks!

.

Hi Hongbin,

i think you have enabled the 20MHz filter in the scope, please do not enable this as this might hide spikes esp. on the Switch node where under or overshoots can lead to component damage. Esp. for that we are just now looking for.

Best regards,

 Stefan

Stefan,

- BOOT1

- HO1

- SW1

- LO1

I will disable the 20MHZ filter in the scope。Test the failed MB and worked MB。

The above wave forms is continued. It is stable.

Hi Stefan,

The worked MB waveforms as followed.

Hi, Stefan,

The faied waveforms as followed.

Hi Hongbin,

still just shows the HO1 signal does not get enabled but I do no idea what is going wrong on this boards.

You may can try to add a  100nF Capacitor in parallel to Resistor PR668. 
Just to see if this has an impact e.g. if the input supply comes up very bouncy.

Best regards,

 Stefan

Stefan，

I'll do this rework today.

Thanks a lot!

Hi Hongbin,

Thanks for the update and doing the rework.
Stefan is currently out of office and will return on Wednesday.

Once you finished the rework and made new measurements, please allow us some additional time to get back to you.

Best regards,
Niklas

Stefan, Niklas,

After adding this 100nf capacitor, the board's LM5176 can work. The input=19V or 9V, It could work well.

While, I mounted another two failed LM5176  on another 2 worked boards. The tow LM5176 failed still.

I will test the failed 2 MBs' waveforms and verify the difference between the failed 2 LM5176 to another worked one which could work after reworked.

Thanks

Hi Hongbin,

Thanks for the update.
I am glad to hear the board is working with the 100nF cap.

Please let us know when you completed the tests with the two failed ICs.

Best regards,
Niklas

Hi, Niklas,

The first failed chip's wave form as followed. The failed LM5176 has the same wave forms after adding 100nf cap.

The second failed LM5176's wave forms as followed.

Hi Hongbin,

as LO1 only goes up ot ~3V, it looks like something with the supply of VCC is not OK. Can you check the level of VCC as well and also check of anything additionally is loading the VCC supply and if the MOSFETs are OK.

Best regards,

 Stefan

Hi Stefan，

The failed LM5176's VCC is 3V. The worked LM5176's VCC is 7V.

VCC is just for LM5176. 

Thanks! 

Hi Hongbin,

VCC supplies some internal logic and the Gate driver for the MSOFETs.

So you might try to remove the MOSFETs and check if VCC then recovers to ~ 7V.

If it is still just going to 3V I would assume that the internal gate drive got damaged.

Best regards,

 Stefan

Stefan，

I'll remove the MOSFET to verify the VCC next week.

Thanks!

Stefan，

Afer removing the mosfets， the VCC still is 3.3V.

Hi Hongbin,

then it looks like the LM5176 is (also) damaged. You should replace this component as well.

Best regards,

 Stefan

Stefan，

We got it.

Is there something to be changed about our schematic and PCB layout except the 100nf MLCC？

Thanks！ 

Hello HONGBIN HU,

Today is a holiday in Germany, so please expect a delayed response. For the schematic review, I recommend using the quickstart calculator: SNVC208 Calculation tool | TI.com

Best regards,
Brigitte

Hi Hongbin,

just reviewed your layout again.

Just know there is only one item i have seen.

The cap for the VCC pin on the LM5176 is connected through the GND plane to PGND of the LM5176.

Would be better to have a direct connection to decrease the inductance. If not possible I would recommend to have to Vias on the LM5176 PGND.

Best regards,

 Stefan

Hi Stefan,

Thanks a lot!

Could you upload a picture based our brd file? The picture  shows how to change the conections between VCC cap's GND pin and PGND.

Hi Hongbin,

please have a look at this application report series:

(1) Four-switch buck-boost layout tip No. 1: identifying the critical parts for layout

(2) Four-switch buck-boost layout tip No. 2: optimizing hot loops in the power stage

(3) Four-switch buck-boost layout tip No. 3: separating differential sense lines from power planes

(4) Four-switch buck-boost layout tip No. 4: routing gate-drive and return paths

and also the the layout of  the EVM - they give good references for the layout.

https://www.ti.com/tool/LM5176EVM-HP

Best regards,

 Stefan

Stefan，

Thanks a lot！We‘ll review the layout tips.

There is no other questions. 

Should i close this question？