IFX007 parallelizing

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

cross mob
User16261
Level 1
Level 1
First like received 5 replies posted First question asked
Hi all,

I tested the IFX007 (no heatsink) with up to 20A current and it got very hot: 4W * 19K/W = 76K above ambient temperature.

Is it possible to parallelize IFX007, also in an PWM use case?

Thanks!
Wilhelm
0 Likes
7 Replies
JingN
Moderator
Moderator
Moderator
First like received First solution authored 25 replies posted
Hi Wilheim,

It is possible to use to parallelize IFX007T in PWM mode. Of course the power dissipation highly depends on your mission profile: Vs, I, frequency, switching time(slew rate resistance)...
If you can provide more detailed information, I could perform some thermal simulation to figure out if the device is able to be used for applications.

Regards,

Jing
0 Likes
User16261
Level 1
Level 1
First like received 5 replies posted First question asked
Thanks for your reply.

Vs = 24V
I > 20A
PWM-Freq=16KHz
slew-rate: as fast as possible

But isn't it a problem due to delay variations, because either to high-sie or the low-side is conducting?
0 Likes
JingN
Moderator
Moderator
Moderator
First like received First solution authored 25 replies posted
Since we have protection feature integrated into IFX007T, we won't have any cross current caused by the delay variations. But due to the variation the output duty cycle might be a bit different from the input duty cycle.

I did some thermal simulation based on your mission profile and some other assumptions:

Load to Vs
Vs = 24V
I = 20A
PWM-Freq=16KHz
Duty cycle: 50%
slew-rate: as fast as possible
PCB: 2s2p

The temperature increase is roughly 90°C after 10s. The switching loss is the main contributor to it. If the application keeps running afterwards, the temperature will keep increase and the device might be turned off by over temperature protection feature.
If would be quite helpful if lower frequency is acceptable for your application.
0 Likes
User16261
Level 1
Level 1
First like received 5 replies posted First question asked
If I use two IFX007T for one half-bridge, I have to connect Vs, Gnd and Ouput. If the switching delay in one IFX007T is slightly different (say 0.9us) from the other IFX007T (say 1us), we have a shoot-through from one FET from one IFX to the other FET from the other IFX. This is a matter of how large is the variation in switching delay in the production process of the different IFX.

Well, the PWM-frequency can be dynamically lowered down to 1KHz. Would you please simulated the switching losses in this case too?.

Thanx!
0 Likes
JingN
Moderator
Moderator
Moderator
First like received First solution authored 25 replies posted
I did thermal simulation for the new mission profile below:

Load to Vs
Vs = 24V
I = 20A
PWM-Freq=1KHz
Duty cycle: 50%
slew-rate: as fast as possible
ambient temperature: 25°C
PCB: 2s2p

In this case, the temperature increase is only 59°C after 500 seconds, and the temperature curve is going flat afterwards. Therefore it should be ok for your application.
0 Likes
User16261
Level 1
Level 1
First like received 5 replies posted First question asked
Did you consider my quote belonging the parallelizing in general?

From above:

If I use two IFX007T for one half-bridge, I have to connect Vs, Gnd and Ouput. If the switching delay in one IFX007T is slightly different (say 0.9us) from the other IFX007T (say 1us), we have a shoot-through from one FET from one IFX to the other FET from the other IFX. This is a matter of how large is the variation in switching delay in the production process of the different IFX.

Do you still see no problems?

Thanx!
0 Likes
JingN
Moderator
Moderator
Moderator
First like received First solution authored 25 replies posted
I thought you might use more than one IFX007T on one PCB. Sorry for my misunderstanding. Now I understood your point.

We don't recommend our customers to use two devices for one half bridge. Due to the variation of the delay time, we might have some current flow through HS of one device to the LS of the other.

For the simulation, I only take one device into account.
0 Likes