Part Number: LM74502-Q1
Other Parts Discussed in Thread: LM74502, LM74930
Tool/software:
Good morning everyone.
I've some problem passing Load dump ISO16750-2 with VsMAX =160V (other pulse for 24V systems like pulse 1, pulse 3a, 3b and so on are correctly passed without damage )
At the moment LM74502Q seems to withstand load dump pulse until 100V max but when i try to reach 160V as required by customer test list , LM74502Q broke up.
The circuit used for the last run of test is depicted below
The circuit is configured as active limiter (OV clamping) with maximum output voltage of about 34V.
Max current flowing through mosfet at 28V is around 200mA.
MOSFET used are Vishay SQSA70CENW (VDS max 150V, Ciss max each 540pF, VGS MAX 15V))
The circuit try to adapt the one present on the datasheet used to withstand 200V unsuppressed load dump (see reference below).
As a possible cause for this behaviour i've guessed it could be the LC filter placed after LM74502.
During bench test i've noted that in case of output voltage clamping i have periodical negative spike (very quick around 200ns with maximum value of -15V) as indicated on the picture below.
Measure is taken on drain of Q49. Green channel.
Below a zoom of negative pulse.
These negative peaking could be a possible cause for the malfunction of my circuit during overvoltage clamping?
Another thing i've noted during normal functioning is a periodic drop on VS caused by R224 (4V with 28V battery IN) (probably related to the current required by the internal charge pump and the high value of zener current limit resistor). (see below on red channel)
I've seen on datasheet of LM74930 that you have placed a diode connected between common source of mosfet and VS to reduce this drop during normal functioning.
This solution can be used also for LM74502? (see below for reference )
Another thing coming to my mind is related to the value of capacitor connected beetween VS an VCAP.
Used mosfet has Ciss max of 540pF so total capacitance should be around 1nF. DS states that VCAP should be at least 10x total Ciss. I've used 1uF... Could be too much in your opinion?
If you have any other suggestion related to component values used on the circuit is really appreciated !
Thank you vey much and have a nice day!
