LM5036: The REF voltage timing is incorrect when starting.

Hong Huang said:

Please label each waveform and also indicate where in your circuit the waveforms are measured. Please provide your circuit/schematics.

the waveforms are measured as shown in the test points:

the schematic file:

功率板PLP101.0a.pdf

the test points is the red circles:

Hi,

The datasheet from ti.com shows VREFSec instead of yours VREF which does not match the current datasheet on ti.com. You need to download the current datasheet from ti.com for your design work.

LM5036 Half-Bridge PWM Controller with Integrated Auxiliary Bias Supply datasheet (Rev. C) (ti.com)

Please read the question again,my proplem is "the voltage REF（green）is not 0 V when VAUX2-OFF is above the threshold VTHSec",I need your help for the reason of this abnormal phenomena .

I have readed the lastest datasheet before,all show "The off-state auxiliary output voltage level present on the secondary side VAUX2-OFF is above the threshold VTHSec, which activates a reset circuit that discharges the output voltage reference VREFSec. This ensures that the opto-coupler is producing a 0% duty-cycle command."

I think the pre-biased start-up process is handled automatically by LM5036. I don't know how to solve the proplem.

Hi,

Your circled in blue is not VREF but VREFSec. VREFSec = 0 not VREF = 0.

Note the datasheet shows the below. 

VREF changes to 5V when 

Hi,

I have remeasured the VREFSec waveform:

the VREFSec is simlar with VREF,the starting sequence is still incorrect.

Please help me to solve the proplem  "the voltage REFSec（blue）is not 0 V when VAUX2-OFF is above the threshold VTHSec"

Hi,

 Q7A and Q7B has no problem.

The VAUX2 is not monotone increasing. this maybe the reason for the  wrong starting sequence.

Now，would you tell me how to change the circuit to make the VAUX2 be monotone increasing?