Automotive loaddump protection ?

[elcielo]

I want automotive loaddump protection circuit.
How I protect it ?

40V 400ms input

[Tupperware]

I'd go for something like this http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00000725.pdf or http://www.datasheets.org.uk/indexdl/Datasheet-072/DSA00361196.pdf

[FTL]

400ms is a long time to protect against.

If the device is to absorb the pulse, it has to turn it into heat. Unless the source is high impedance, there will be significant current (maybe more than 10A) and to absorb that for 400ms is a lot of energy.

To protect against voltage spikes of that duration (not really a spike), I would look at a device to absorb the first 10us of the pulse while a series FET is turned off to limit the current.

[solutions]

Use a 60V input capable regulator?

What's the rise/fall on that 400ms?

[miserable]

Hi elcielo,

I have developed power supplies compliant with ISO 7637 Pulse 5 (Load Dump) test.
Protection methode differs according to power comsumption.
Lower the power, easier to protect.
For high power there are bottlenecks. For example addes series protection device adds extra impedance and voltage drop. And also series protection devices like polyfuses opened in over current event and it requires for time for recovery. But ISO 7637 does not let power interruption long time.
Thera are a lot of consideration changing with your power reqiurement.
Please give your current consumption, voltage level, test level. So I can share my little experience with you.

Best regards.

[alexisnik]

Just a thought here (so don't bite if I'm wrong  ), but I think that the use of a varistor is this (or similar). Again, not sure, but I found it installed for protection within an A/C unit I opened up to fix. It seems that it absorbs the extra current after a voltage * current value. Maybe it's only for AC currents, I don't know. Just a thought...

[bigtoy]

I've dealt with it in the past simply by making my input power supply capable of handling a 40V input. As Miserable says, for low power devices (which mine was) its not a big deal - any extra heat dissipation due to the higher input voltage only lasts for a few hundred milliseconds - pretty easy to handle.

[miserable]

Hi alexisnik,

The load dump test pulse energy it too high to handle with a single varistor.
Because pulse is too long. So if you estimate the required absorbtion energy by varistor, you will see that there is no feasible-practical varistor solution possible unless you use a secondary protection with polyfuse.
You could use polyfuse+varistor to limit varistor absorbtion, but 2 problem arises. You need fast action polyfuse to protect varistor. So its acting current should not be so high, that comes with high on resistance and long recovery time. Once Load Dump Test pulse applied, circuit shouldn't cease his functionallty at least given time by ISO 7637 standard. Maybe you can manage this for low power but when power goes up, problems and considerations fallows it.

Here is a proven solution for high power. As you know when we design linear regulators and always consider the power loss and main switch SOA for continious operation. Please think a kind of regulator that pass input voltage without doing nothing up to given limit, say 36V. But if an over voltage event (over 36V) it starts regulation to 36V up to 400V input for properly selected mosfet for up to 1 or 2 second most beyond to load dump test pulse.
Your pseudo regulator can handle up to 400V, so your input protection device (TVS-recommended, MOV..) can be higher rated than any load dump pulse, say 150V. So your protection device will never struggle with pulse, instead lower powered  bust, surge, EFT pulses.
You will ask how I am design this regulator, do not worry, Surge Stoppers, Hot Swap Switches designed by Linear, TI or any other vendor can handle this for you with additional reverse voltage proction bonus without using input diode. (ISO7637 requires working with 6V during crankind drop)
Please consider LT435 or similars.
You will face any other problems when power goes up. That is a long story and but has a simple solutions. I can write a little bit more if you asked.
Best regards.

[alexisnik]

Great info, thank you!

[bobcat1]

Hi 

Maxim has special IC for handling automobile load damp - I don't remember the part number , but search maxim site for load damp

All the best

Bobi

[Codeman]

If you want a regulator (Buck or Boost) look at this:

http://cds.linear.com/docs/en/lt-journal/LTJournal-V22N4-02-df-LTC3115-1-Canfield.pdf

[reksbg]

Here is another simple circuit that complies with the standard (just don't forget to protect the mos transistor with zenner ):

[havok1919]

This is a lighter weight version of the technique mentioned above:



(resistor values aren't critical, and the bulk bypass cap would be sized and rated per your application, etc.)

A more expensive (but much better) solution is possible with something like the Linear Tech LT4356.  Very handy for both over and under voltage protection (it can activate much faster than the discrete solution above):

http://www.linear.com/product/LT4356-1_-2

[analognoise]

Here is another simple circuit that complies with the standard (just don't forget to protect the mos transistor with zenner ):

Very nice!

This is a lighter weight version of the technique mentioned above:



(resistor values aren't critical, and the bulk bypass cap would be sized and rated per your application, etc.)

A more expensive (but much better) solution is possible with something like the Linear Tech LT4356.  Very handy for both over and under voltage protection (it can activate much faster than the discrete solution above):

http://www.linear.com/product/LT4356-1_-2

Does anyone have any more documents about this technique for further study, or a good reference?