LM5116 buck converter not working when output connected to 12 volt source

[smijesh]

Hi all
I have noticed an issue in my Buck converter, converter not turning on when output of the buck converter connected to a 12V DC source.
Vin 50V,
Vout 13V
I Out 10A
IC LM5116
By trail and error methode i fixed this issue by changing series gate resistance to 0E, it was 33E.
What could be the reason
similar schematic attached

 

[barry]

why would you expect it to work when you connect the output to a 12V source??? Your converter is trying to produce 13V, your source is trying to hold it at 12V. why would you even do this?a

 

[andre_luis]

At a glance, your above schematic has few to do with the typical circuit shown on datasheet.
For instance, the current sense input pairs (CS/CSG) are not connected to a shunt resistor.

--- Updated May 19, 2023 ---

By trail and error methode

In Power electronics this method in general result in smoke.

 

[Easy peasy]

there are likely reasons for your predicament that are related to your build that have nothing to do with the circuit you posted

--- Updated May 20, 2023 ---

also - your schematic shows TWO " Rg's " which one are you talking about .... ?

 

[smijesh]

why would you expect it to work when you connect the output to a 12V source??? Your converter is trying to produce 13V, your source is trying to hold it at 12V. why would you even do this?a

Hi this 130W buck converter is an auxiliry supply and always connected to the DC source, buck converter input will active wheniever its requied.
The buck converter working well when not connected to 12V DC source

--- Updated May 20, 2023 ---

At a glance, your above schematic has few to do with the typical circuit shown on datasheet.
For instance, the current sense input pairs (CS/CSG) are not connected to a shunt resistor.

Hi current sense through Low side MOSFET RDson

--- Updated May 20, 2023 ---

also - your schematic shows TWO " Rg's " which one are you talking about .... ?

Hi attached schematic is the latest one, before i was connecting LO and HO pins to MOSFET through 20E each. due to this problem i connected LO and HO directly

The buck converter working well when not connected to 12V DC source

 

[Easy peasy]

no attachment - ( or still two Rg's )

--- Updated May 20, 2023 ---

if your feedback resistors are set for lower than the battery voltage ( which we don't know exactly ) then the converter won't go - this is why we need the exact schematic with values to determine the fault - AND photo's

as previously requested

 

[KlausST]

Hi,

i fixed this issue by changing series gate resistance to 0E, it was 33E.

this tells me that the shown schematic isn´t your schematic at all. There are no gate resistors involved.

If you want someone else to find issues/mistakes I strongly recommend to post the very your schematic. Not the one of a datasheet, really your own. So often we see that posters "think" they did it like in the datasheet, but they made mistakes. We can not find those mistakes when we don´t have the according schematic.

Klaus

 

[smijesh]

if your feedback resistors are set for lower than the battery voltage ( which we don't know exactly ) then the converter won't go - this is why we need the exact schematic with values to determine the fault - AND photo's

Hi
Buck converter converter output is 13VDC and xeternal source voltage is 12VDC

 

[KlausST]

Buck converter converter output is 13VDC and xeternal source voltage is 12VDC

Why do you make it more difficult than necessary?
Why don´t show your actual circuit? ..so we can see all your connections and part values.
Is it a "top secret design"?

Klaus

 

[smijesh]

Hi please find the schematic

 

[betwixt]

Looks like a recipe for smoke to me.

Brian.

 

[KlausST]

Hi,

a rather different design now. I asked for your schematic. And you told us it is working well when not connected to 12V.
But this schematic never can work. So it simply can´t be your schematic.

The schematic of post#10 has several issues. It is not according datasheet recommendations..
And worst of all: with the output connected to GND it never can work properly as you told.

I´m not here to waste my time.

Good luck.

 

[barry]

I STILL don't understand how you expect a 13V regulator to 'work' when you connect it to a "12V source"(which, by definition, will produce as much current as necessary to maintain 12V).

 

[smijesh]

I STILL don't understand how you expect a 13V regulator to 'work' when you connect it to a "12V source"(which, by definition, will produce as much current as necessary to maintain 12V).

Buck converter always connect to 12V battary line , when more power required buck converter will switch on and battery line voltage will become 13V. it will happen only when gate resistor become 33E, why

--- Updated May 20, 2023 ---

The schematic of post#10 has several issues. It is not according datasheet recommendations..
And worst of all: with the output connected to GND it never can work properly as you told.

you are right, schematic is wrong can you explain what happening when gate resistor change from 33E to 0E

 

[Easy peasy]

you need to fix this bit:

--- Updated May 20, 2023 ---

when the load is a 12V battery ( we assume a good one with very low internal resistance ) you are now at the mercy of the current limit within the converter - since the converter will try to run at max current until the battery volts come up to the regulation voltage.

There may be some odd effect at start up due to the current being quite high that is overcome by changing from 33 ohm to zero ohm gate drive resistors . . .

--- Updated May 20, 2023 ---

also your feedback is set for 12.73V, if the batt was higher then no start ( or a very brief start then stop as the batt voltage holds up more than 12.73V )

if later, after you changed the resistors the battery was flatter then it would start, so it may be unrelated to the resistors

--- Updated May 20, 2023 ---

I note that the data sheets etc do not call for any gate resistors, this may be due to dead time limitations

in any event if you were to use 33 ohm for the top fet it would be a good idea to use a reverse shottky diode across the 33 ohm to speed up up the turn off so as to keep dead time provided by the IC

I note the bottom fet does not need a gate resistor ( often put in for emc reasons ) , as, it is turned on after the internal diode is conducting, and turned off under ZVS. Under diode emulation mode it is turned off before the top fet turns on.

There is a further issue not discussed in the IC literature - namely if you apply a 12V batt to the output the voltage will reach the input via the top fet diode, as you have used a 1 Meg pull up on the EN pin, this may violate the pull up time needed for this pin ( < one sw cycle ) and may cause a funny latch out.

--- Updated May 20, 2023 ---

It may also be that by using 33 ohm gate drive resistors, the bottom fet is not fully turned off before the top fet becomes ON - this would cause a shoot thru that will cause an hiccough on the IC

similarly if the top fet is not fully off ( due to 33 ohm ) when the bottom fet turns ON, another shoot thru may occur and cause an hiccough on the IC

again - this may be why no resistors are shown in the documentation - to preserve dead time

- you 're welcome.

 

[barry]

Buck converter always connect to 12V battary line , when more power required buck converter will switch on and battery line voltage will become 13V. it will happen only when gate resistor become 33E, why

--- Updated May 20, 2023 ---

you are right, schematic is wrong can you explain what happening when gate resistor change from 33E to 0E

Then why didn’t you say BATTERY instead of VOLTAGE SOURCE.

 

[Easy peasy]

@barry - it seemed fairly clear to all other readers that the batt was going on the output .... no need to remonstrate really - I suspect even you managed to work out what the poster really meant ( like the rest of us did ).

--- Updated May 21, 2023 ---

Further - you may have issues with your feedback loop since R160 is so low ( 1k2 )

this puts the roll off at F = 1 / ( 2.pi R C ) = 186kHz which is way too high ( wayyyyyy too high )

better to aim for 1kHz rollover say, which makes C115 = 150nF ( the output filter has a resonance at 2.5kHz )

then R154 can be used for a phase lift @ 10kHz by making it 100 ohms.

The sw freq is a litlle high - making R155 = 18k ( ~ 180kHz ) will give better efficiency

at 48Vin and 13.5 Vout the ripple will be +/- 1.85A, so you could easily go as low as 150kHz without issue.

 

[KlausST]

Hi,

I can´t blame Barry for this. It´s clearly the OP who gives just piece by piece, inclomplete and erroneous informations.
We need to guess - And in my eyes "guessing" is not the right way to design reliable electronics.

but this is not the end of the schematic issues:
* There is R163 - which in my eyes - is completely against the manufacturer recommendation to keep HB signal low impedance. It´s likely to cause the high side undervoltage protection to be activated.
* R161 is not according the datasheet recommendation - and I can´t see the benefit now. It may be used to optimize efficiency after the circuit is running fine.
* Battery charging usually needs to apply a suitable charging current. I´m not sure whether this is made correctly - but honestly I´m too lazy to do the calculations.
* "constant voltage" isn´t the best idea at all to charge a battery. The charging state never comes near 100%

* and when the schematic is that messy ... how does the PCB layout look like?

Usually one can explain why one modifies the recommended schematic of the datasheet. The OP gives no idea why they did the modifications.

In the end we may provide a lot of effort ... but still the application will never work properly.

Klaus

 

[Easy peasy]

R163 may be un-neccessary but at least it won't upset things too much @ 2.2 ohm

R161 selects the level for diode emulation of the bottom fet - it may well be OK @ 1k, it could be zero ohms without issue.

 

[KlausST]

Hi,

R163 may be un-neccessary but at least it won't upset things too much @ 2.2 ohm

I personally see no benefit in placing a series resistor to a power supply pin of any IC.
It just destabilizes the power supply voltage. Counter-productive.

The datasheet says: "and (the bootstrap capacitor) should be placed as close to the controller as possible". So for me this means every extra milliOhm should be avoided. The voltage should be as stable as possible.
The resistor adds not only it´s resistance but also extra inductance due to the part and the traces.

At the pin the peak current is 1A, thus - according Ohm´s law - it immediately drops 2.2V by the resistance .. and even more by the additional impedance.

Klaus