Bootstrap high side drivers are not good?

[eem2am]

In the following article, Dr Ray Ridley, the world's pre-eminent SMPS designer, dislikes the use of Bootstrap High Side Driver chips such as NCP5181. (IR2110 etc etc)

Dr Ridley states that with mains voltage SMPS's, its always preferable to use gate drive transformers.

This statement can be found in the first page of the following article........the text is at the top right hand corner, just above Figure 1.

Gate Drive Design Tips - Ridley


How can Dr Ridley make such a sweeping statement?
-and is he right?

NCP5181 DATASHEET
http://www.onsemi.com/pub_link/Collateral/NCP5181-D.PDF

[mag]

I don't like those bootstrapped drivers that much for high voltage I prefer to use GDT where possible. On top of that GDT are often cheaper than the ir2110 and similar ic.
Anyway it is only a personal opinion.

If they were so bad and unreliable why IR produces the 2110 since decades?

What shocks me is how a guru like Ray Ridley can write something like that on a paper.... For sure IR (who has bootstrapped drivers up to 1200V), onsemi and other will be very happy....

Ciao

-marco

[gmarsh]

They've got some issues - maximum dv/dt handling, frequency limitations, delay uncertainty, etc. but there's nothing fundamentally wrong with high side drivers provided you take all this into account. The wider duty cycle capability is nice, and they seem good for space constrained applications.

I've just never used one. Don't need them for a flyback, and for larger power supplies I've always found myself building a resonant converter of some sort where high side MOSFET drive is 50% duty cycle.

Curiously, the power supply standard for the "open compute project" forbids the use of high side gate drivers, so there must be something I'm missing.

[johansen]

I did not find any such statement. however, i did find
"High-side integrated drivers remove too much control from the designer, and do not provide the samelevel of protection, isolation, immunity from transients, or common-mode noiserejection as a well-designed and imple-mented gate drive transformer."


"Curiously, the power supply standard for the "open compute project" forbids the use of high side gate drivers, so there must be something I'm missing."

I think you're missing the heading that says avoid the use of " " to increase reliability. Note they also mention tantalum capacitors

[eem2am]

Thanks.... which type of high-side drivers is forbidden?

Also, the statement that you kindly quoted is damning enough for me.....from that i would steer well clear of bootstrap drivers...period