Swapping diode for P-MOSFET: reverse spike protection

[kellogs]

How do I replace D2 in this answer with a P-MOS FET ? I am thinking about this diagram:



The two D1 symbols are just one bidirectional TVS diode (missing symbol in circuitlab) with Vcl_max = 24.4V and M1 has Vgs_th = -4V Vgs_max = -20V. Vds_max - should I worry about it ? Will this protect fine against -220V reverse voltage spikes (and +150V spikes) of the http://www.sft-lab.com.cn/uploads/ISO 7637-2-2011.pdf kind ?

Someone suggested this diagram instead:

Mine is simpler and I still do not know if the second diagram is really warranted. A few questions:

1. Is that ES3J really needed or is the -24.4V from TVS not enough to cause any damage ? Why is there no limiting resistor on that branch, does the MOV work as one ?
2. Is the GS junction protection (Zener + 10k resistor) really needed, will the source pin ever get so much higher above gate pin ?
3. If GS protection is really needed, then can the ES1J be eliminated if the Zener limiting resistor was made 100 - 1k ohm ?

Thank you

 

[crutschow]

Why do you want to use a more expensive MOSFET to replace a diode?
With a 3.3V output from the regulator, the 0.7V drop for the 13.8V input will just reduce it to about 13.1V, which will have no effect on proper circuit operation.

And why did you get rid of the series resistor?
That helps minimize the spike current through the Zener.

 

[kellogs]

Not sure whether it will be more expensive, definitely cheaper when taken together with heatsinks and fan.

Which series resistor is that, from what diagram ?

 

[crutschow]

when taken together with heatsinks and fan.

What do you think needs a heatsink and fan?
You don't need either if the load is only 50mA.

Which series resistor is that, from what diagram ?

The one in the referenced schematic.

 

[kellogs]

The load is not 50 mA but up to 5A and perhaps even larger. Diagram shown is just a zoom in of the full thing.

The series resistor - I recon its use is to limit current through diode(s). Would not want its voltage drop otherwise, hence I have moved it on the TVS diode branch in first diagram. Second diagram is not mine - also asked about missing resistor at point 1 in my question.

Thanks

 

[crutschow]

The series resistor - I recon its use is to limit current through diode(s). Would not want its voltage drop otherwise, hence I have moved it on the TVS diode branch in first diagram.

That will protect your diode but not the circuit, since now there is noting to keep circuit voltage from rising way above the TVS voltage.
You need a small resistor in series so that your circuit only sees the TVS voltage, while also limiting the TVS current.

 

[kellogs]

Oh, snap! how about swapping places D1 with R1 instead?

 

[crutschow]

Electrically there's no change when you reverse the order of two series elements.
Why would you think there is?

You circuit is still not protected from high energy spikes.

 

[kellogs]

Err... please excuse my noobness. I surely would not want any resistor in series with it all.

Is the second diagram going to do the job ? I did lay out 3 points where i am having uncertainty.
And a 4th one - that MOV, I really don't like. It will likely not withstand 10 - 15 years of service.

 

[Alec_t]

Is the second diagram going to do the job ?

No. The P-FET drain and source are the wrong way round.

 

[crutschow]

The P-FET drain and source are the wrong way round.

No, that is the correct orientation to block a reverse power connection, which is its purpose.

 

[Alec_t]

You're right. I mis-read the intention.

 

[crutschow]

I surely would not want any resistor in series with it all.

Then you need to remove the resistor in series with the MOV and hope the circuit power circuit impedance is enough to protect the MOV from being zapped by any transients.

 

[kellogs]

Electrically there's no change when you reverse the order of two series elements.
Why would you think there is?

You circuit is still not protected from high energy spikes.

Hey,

could you, or anyone else, comment on this diagram ?


How will R1 affect the TVS clamping voltage here, if at all ?

And a side question, is there any difference between D2 above and two identical unidirectional TVS arranged like below (disregard the text around them and consider D1 and D2 identical to the TVS above in all regards except directionality)?

 

[crutschow]

How will R1 affect the TVS clamping voltage here, if at all ?

It will keep the TVS from clamping properly, as noted in the other posts here.
What part of "you can't put a resistor in series with a clamping diode" is hard to understand?

And a side question, is there any difference between D2 above and two identical unidirectional TVS arranged like below (disregard the text around them and consider D1 and D2 identical to the TVS above in all regards except directionality)?

The forward drop of one diode would add to the reverse clamp voltage of the other.
Why do you want to put two in anti-series?