Reverse overvoltage on diode

[cupoftea]

Hi,
Do you know why the ringed diode would ever need to be 400V rated? Its a general purpose, SMA, 400V diode.
Why ever do you think they used such a high voltage part?
LTspice and PNG attached.

 

[cupoftea]

Thanks, in the schem of #14 it should not , but you never know with stray inductance, and release-of-short-cct etc etc...we are most worried about cathode of the D5 diode going way below ground, due to stray reactance ringing...and thus we are thinking of adding the shown 27V zener.

 

[D.A.(Tony)Stewart]

I see no clamping action for Q1 Vbe < Vbeo max .

A Vbe reverse diode is necessary for Q1be in #14

 

[cupoftea]

Thanks, i noticed the following on page 11 doesn't bother with this diode (reverse BE) either..

https://ww1.microchip.com/downloads/aemDocuments/documents/OTH/ApplicationNotes/ApplicationNotes/01465A.pdf

There is a Dave Jones Video showing how reverse biased BE acts like a 6v Zener, and in some applications,
its used as that.

Fundamental EEVblog #1157 - Transistor Zener Clamp Circuit from Dec. 2018.

Also, there are many PNP turn-off circuits where a resistor is placed from Base to emitter, and the voltage across it, during
the fet drive pulse, is deliberately greater than Vebo

..As such, i thought the lack of Veb protection would be ok....i appreciate the 27V zener could conduct far too much for the BJT_eb to clamp, but i believe that any reverse voltage on the zener will be very short lived...just a super fast transient.
Though to be honest, being the conservative type, i usually do put your kindly mentioned protection diode in there, but in this case, i dont have room.

 

[D.A.(Tony)Stewart]

I find Mr. Jones overlooks details like the max. current for Vbeo.
Like this part he went ecstatic about.

Don't assume if there is no current limit on the datasheet for Icbo that it doesn't matter. I assume there is no safe limit until otherwise specified.

It certainly will be far less the OA limit which is another reason to add Rs to the base and diode across EB.

Devices with Vebo >> 6V will likely be superbeta types with lower current.

Bottom line. You can use Vceo for low current Zeners but Vebo power limit is extremely low.
Don't take design advice from Dave Jones. He didn't master that skill but is ok at basic stuff.

Most of the high Vebo parts > 25V on Octopart are data entry errors but there may be one that meets your needs.

 

[cupoftea]

Thanks, with the PNP turn off circuit, with say 15R added across the BE of the PNP, this is the same situation, its possible to thave the protection diode there, but few people use it. I remember calling this out many years ago, but its a bona fide way, if not the favoured way of doing PNP turn off.....ie, subjecting the PNP BE to more than its Vebo.

Its done deliberately, since that series resistor (ie in series with the fet gate ultimately) could have gone elsewhere.

We discussed it here some time back...and there were thoughts that the resistor gave a high voltage to the VBE and so helped very quickly ensure that it (the PNP) is deffo off during the FET_ON drive pulse.

Attached is one PSU schem with pnp turn off.....this one has lifetime well over 10 years on oil rigs...as can be seen, the veb is going to be high

 

[cupoftea]

Don't take design advice from Dave Jones. He didn't master that skill but is ok at basic stuff.

Thanks, but in the above video on Vebo, Dave Jones actually links in the article by that design engineer from Fluke, who is the guy that designed that "Veb zener" and put it into all the fluke meters.