Burning FETs on new layout with 6EDL04I06PT

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
User17346
Level 2
Level 2
10 replies posted 5 replies posted 5 questions asked
Hi,

I made a motor drive using 6EDL04I06PT, STD16NF06T4 FETs (16A TO252), 0.033ohm triple shunt, 2.2uF bootstrap caps and 24ohm gate drives on all gates.

I laid out one PCB with this, and it works pretty near perfectly... but there was some switch ringing, and ringing on the shunt resistors which was causing occasional current trips, so when I respan the board (needed to change a few unrelated things) I moved the shunts around to reduce the loop areas and minimise the inductance.

The other change I made, was to use the NC pins to feed through the LO1,2,3 signals, which made for a much tidier routing with fewer ground plane breaks. Since the board is being used for low voltage (up to 48V) no concern about creepage distance.
Is this OK to do? The datasheet says n/c - not connected, which in my mind meant they were probably intended for this purpose, but now I'm less sure.
Despite being as close to identical schematically as conceivable apart from those layout changes, the board has destroyed two sets of FETs after only a few seconds of operation. Symptom has been that the low side FETs have had their gates burnt through - only tens of ohms between gate and source. High side FETs have possibly also died, but it's less clear cut; the gates still have high impedance and ~3Mohm drain source resistance.

Board has been run with 15V gate drive voltage, 24V supply.

Firmware being run on it is identical. I can still upload the firmware to the old version and have it spinning away happily.

Any advice? I have tried prising up the N/C legs on the gate drive, and it ran again, but breifly... I only changed the FETs that seemed to have died, perhaps they were all partially knackered.

3989.attach

3990.attach
0 Likes
1 Solution
Srivatsa
Moderator
Moderator
Moderator
First comment on blog 10 sign-ins First like received
We checked on the 6EDL internal connections and our feedback is as below:

1) There is no connection to those NC pins, however we do not recommend to use those NC pins since there could be soldering residues or shorting to nearby pins issues with PCB assembly
2) You can also check whether there is cross-talk issue when you utilize NC pins. Monitor the input and output voltages of the driver. Are layout parasitics causing cross-talk here?
3) Further, you should check whether there is parasitic turn-on of the MOSFETs. Monitor the gate to source voltage to see if the device turns-on in high dv/dt or di/dt conditions such as start-up or load jumps

View solution in original post

0 Likes
2 Replies
User17346
Level 2
Level 2
10 replies posted 5 replies posted 5 questions asked
So I replaced everything power on the board, bending the n/c pins up on the 6EDL chip this time. It now seems to work fine.

Can someone from Infineon check and confirm whether NC pins truly are NC? Perhaps if they are actually connected to something, the datasheet could represent this?
0 Likes
Srivatsa
Moderator
Moderator
Moderator
First comment on blog 10 sign-ins First like received
We checked on the 6EDL internal connections and our feedback is as below:

1) There is no connection to those NC pins, however we do not recommend to use those NC pins since there could be soldering residues or shorting to nearby pins issues with PCB assembly
2) You can also check whether there is cross-talk issue when you utilize NC pins. Monitor the input and output voltages of the driver. Are layout parasitics causing cross-talk here?
3) Further, you should check whether there is parasitic turn-on of the MOSFETs. Monitor the gate to source voltage to see if the device turns-on in high dv/dt or di/dt conditions such as start-up or load jumps
0 Likes