Comparison of Bootstrap Driver ICs for SMPS?

[cupoftea]

Hi,
For a 150W Full bridge off 400V, (35Vout) which driver would be more robust to the ills of hi side bootstrap driving for use in this to drive the bridge FETs?
The 2ED2106, or the 2EDF7175?
They are both same price, both have bootstrap hi side drive supply, and both could do the job.

2ED2106

https://www.infineon.com/dgdl/Infineon-2ED2106-4-S06F-J-DataSheet-v02_32-EN.pdf?fileId=5546d4626cb27db2016cb8d7402029e7

2EDF7175

https://www.infineon.com/dgdl/Infineon-2EDF7175F-DataSheet-v02_08-EN.pdf?fileId=5546d462636cc8fb0163b09026be3060

 

[tepalia02]

Looks like the operating voltage rating is higher in 2ED2106.

 

[D.A.(Tony)Stewart]

Neither have dead-time control, others do.

Only the 2EDF7175 has 8k impulse isolation.

 

[Tahmid]

Don't forget to consider the negative max voltage rating for the pin that connects to the high-side FET's source. What you need that to be will depend on your application and layout. The 2ED2106 datasheet clearly specifies -11V rating, and section 5.9 calls this out.

 

[cupoftea]

Don't forget to consider the negative max voltage rating for the pin that connects to the high-side FET's source. What you need that to be will depend on your application and layout. The 2ED2106 datasheet clearly specifies -11V rating, and section 5.9 calls this out.

Thanks, you have a superb point....and many would say that the 2EDF7175 does not need to worry about the negative voltage pulse on the switching node relative to "primary" ground...because it uses an isolation transformer.....but the datasheet does not help us here, so i am sightly guessing this.

I actually think that bootstrap high side drive for a smps with 400V input is suicide..for the very reason that you highlight.....will lead to many failures after 2 years or so....but there is no evidence to back this up...just a lot of www docs from experts who point to this.

--- Updated Nov 5, 2023 ---

Don't forget to consider the negative max voltage rating for the pin that connects to the high-side FET's source. What you need that to be will depend on your application and layout. The 2ED2106 datasheet clearly specifies -11V rating, and section 5.9 calls this out.

Thanks, you have a superb point....and many would say that the 2EDF7175 does not need to worry about the negative voltage pulse on the switching node relative to "primary" ground...because it uses an isolation transformer.....but the datasheet does not help us here, so i am sightly guessing this.

I actually think that bootstrap high side drive for a smps with 400V input is suicide..for the very reason that you highlight.....will lead to many failures after 2 years or so....but there is no evidence to back this up...just a lot of www docs from experts who point to this.

 

[Tahmid]

Thanks, you have a superb point....and many would say that the 2EDF7175 does not need to worry about the negative voltage pulse on the switching node relative to "primary" ground...because it uses an isolation transformer.....but the datasheet does not help us here, so i am sightly guessing this.

The switching loop is relative to PGND so you can still see a negative transient on GNDA relative to PGND/GNDB. You're right it won't be relative to the logic GND. Does it matter if GNDA shoots very negative relative to GNDB? Maybe - perhaps the datasheet has more details on it and I missed that from a quick glean.

I actually think that bootstrap high side drive for a smps with 400V input is suicide..for the very reason that you highlight.....will lead to many failures after 2 years or so....but there is no evidence to back this up...just a lot of www docs from experts who point to this.

That's a bold claim to make. You can definitely design and characterize your driver to fall within the safe design operating limits under a variety of operating and corner conditions and have the design operate reliably.

 

[cupoftea]

That's a bold claim to make. You can definitely design and characterize your driver to fall within the safe design operating limits under a variety of operating and corner conditions and have the design operate reliably.

..Thanks, but because of common mlode noise its virtually impossible to scope how far the pulse goes below ground.....the common mode pulse associated with the spike will couple into the scope probe lead and make the spike look higher (more negative) than it really is.

--- Updated Nov 5, 2023 ---

That's a bold claim to make. You can definitely design and characterize your driver to fall within the safe design operating limits under a variety of operating and corner conditions and have the design operate reliably.

..Thanks, but because of common mlode noise its virtually impossible to scope how far the pulse goes below ground.....the common mode pulse associated with the spike will couple into the scope probe lead and make the spike look higher (more negative) than it really is.

--- Updated Nov 5, 2023 ---

That's a bold claim to make. You can definitely design and characterize your driver to fall within the safe design operating limits under a variety of operating and corner conditions and have the design operate reliably.

..Thanks, but because of common mlode noise its virtually impossible to scope how far the pulse goes below ground.....the common mode pulse associated with the spike will couple into the scope probe lead and make the spike look higher (more negative) than it really is.

 

[Tahmid]

The leading contributor to the spike looking more negative than it really is would be probing inductance. If you use good probing techniques to to reduce the parasitic inductance, you should be able to get a good accurate reading.

You can use a diff probe with pretty good common mode rejection. Even a low-end diff probe like the DP10007 has >45dB CMRR at 1MHz and >60dB CMRR at 20kHz. https://www.micsig.com/Differential Probe04/