LM5118 EVM not working problem

Hi

Would you please double check your power supply's current limit ? It should be enough to hadle full load current at low input voltage (>9A~10A)

Regards,

My PSU wouldn't be able to cope with full loads, its only rated at about 2amps. Its why I wasn't loading the output much.  Though the output does not work with no load either. I would assume with no load, its not going to need more than 2A input anyway ?

Hi

That's right. Please turn it on at high Vin (make sure 12V is estabilished) and then decrease the input voltage slowly while monitoring the input current.

Regards,

starting with 25V input, on turn on it pulls about 300mA

20V at turn on, it pulls about 400mA.

15V at turn on it pulls about 450mA for about 5 seconds then drops to 200mA and the output voltage drops to 7V.

If I vary the voltage from 25V downwards while it is turned on, the current starts to go up, but it does not want to pull more than 400mA. Once it hits that current the voltage on the output just drops.

Hi

If you can, please check the Q4 drain to source voltage or its gate voltgae at low Vin.

This is to check Q4 switch runs well in boost mode. If Q4 switch is not switching well, LM5118EVM can do 'Buck' operation at high Vin, cannot do 'Boost' operation at low Vin. 

Regards,

Q4 ? I can only see Q1 & Q2, I will check to see if they are both switching though.

Sorry. Please check Q2 drain to source voltage and its gate signals at low Vin.

Q2 gate in ref to GND  , above 15V input it pulses about 88khz, looks like very bad ringing on the rise of the cycle, 800mV on the gate.  Below 15V the waveform pulse frequency increases, above 15V it bascailly flat lines. I can take scope images if needs be.

Q1 gate taking TP5 as the ground ref,  cant get a stable reading below 20V input, above 20V, 350khz bad ringing ont he rise, about 600mV on the gate.

I don't think either of those gate voltages seem correct.

Q1

Q2

It looks like something is wrong. At low Vin, Q2 should switch at the same frequency as Q1's.

they both do switch at about 350khz, but the ringing and distortions in the waveform is making the scope readings incorrect. Its very hard to get a frequency lock, it centers around 350khz, but its also going down to a few hz and up to 5mhz.

I am just thinking the gate voltages do not seem correct, but if the fets were undervolted that much I would assume they would have blown by now. Can only assume the IC itself is not giving enough drive to the fets. So will just have to order some replacement chips and see if changing it solves the problem.