Load Dump/Transient Protection.

[AJ]

Hi all,

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a PolyFuse/TVS
arrangement on the input but if the system is to work on 24V I would have to
use a 30V TVS which clamps at around 48V, a little high for the input of my
regulator! Any ideas?

Regards

AJ

 

[John - KD5YI]

Hi all,

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a PolyFuse/TVS
arrangement on the input but if the system is to work on 24V I would have to
use a 30V TVS which clamps at around 48V, a little high for the input of my
regulator! Any ideas?

Regards

AJ

A shunt clamp using a transistor and zener following the PolyFuse?

Just a quick thought.

Good luck.

John

 

[John Popelish]

Hi all,

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a PolyFuse/TVS
arrangement on the input but if the system is to work on 24V I would have to
use a 30V TVS which clamps at around 48V, a little high for the input of my
regulator! Any ideas?

Regards

AJ

If you really want to clamp the peaks to less than 40 volts, you might
use a 36 volt, 3 watt zener. But with a +-5% tolerance, this means
that it might clamp as low as 34.2 or as high as 37.8 volts. The only
energy in a load dump is that stored in the inductance of the wiring.
There is also a possibility that the alternator will produce its
full output into the battery till the regulator recovers, and the
worst case for this would be when the vehicle is operating with the
battery disconnected.

How much current must your regulator deliver?

 

[Fred Bloggs]

If you really want to clamp the peaks to less than 40 volts, you might
use a 36 volt, 3 watt zener. But with a +-5% tolerance, this means that
it might clamp as low as 34.2 or as high as 37.8 volts. The only energy
in a load dump is that stored in the inductance of the wiring. There is
also a possibility that the alternator will produce its full output into
the battery till the regulator recovers, and the worst case for this
would be when the vehicle is operating with the battery disconnected.

How much current must your regulator deliver?

What???!!! The load dump scenario is a catastrophic loss of battery
connection, the alternator pumps the whole car load up to whatever 60V
with allowable 500ms rundown.

 

[Rich Grise]

Hi all,

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a PolyFuse/TVS
arrangement on the input but if the system is to work on 24V I would have to
use a 30V TVS which clamps at around 48V, a little high for the input of my
regulator! Any ideas?

I don't know what TVS you're using but a 1N6281A/1.5KE27A has a nominal
clamp voltage of 25.7 min - 28.4 max, and a "Maximum Clamping Voltage at
Ippm", which is spec'd at 40.0A, of 37.5V. It's called a "Transzorb":
http://www.vishay.com/docs/88301/15ke.pdf
Digikey and Mouser and Newark should have them.

Good Luck!
Rich

 

[John Popelish]

What???!!! The load dump scenario is a catastrophic loss of battery
connection, the alternator pumps the whole car load up to whatever 60V
with allowable 500ms rundown.

Is that the official definition of a load dump in vehicle
applications? Is this documented somewhere?

 

[John - KD5YI]

Is that the official definition of a load dump in vehicle
applications? Is this documented somewhere?

According to SAE J1211 (1978), pages 19 and 20, the load dump transient has
this characteristic:

V = 106*exp(-t/.188) + 12

It is caused by the load on the alternator being suddenly reduced. Worst
case is caused by disconnecting a discharged battery when the alternator is
operated at full load (often initiated by a defective battery terminal
connection).

It is not the highest voltage mentioned in the document. There are inductive
switching (-286V) and mutual coupling (214V) transients but their time
constants are about 1ms.

Note that SAE J1211 concerns itself with 12V systems. There is also good
information on other conditions within the automotive environment
(temperatures encountered in various areas, shock, vibration, dampness, etc.)

For anyone contemplating designing equipment for automotive equipment, I
would recommend visiting the SAE site for documents. I think they have more
up to date information on 12V systems than does J1211, and documents for 24
and 48V systems. The documents are not free.

Good luck.

John

 

[Spehro Pefhany]

Is that the official definition of a load dump in vehicle
applications? Is this documented somewhere?

That's what a load dump is. For an official version of a load dump
transient, see: ISO 7637-2, Test Pulses 5a and 5b. There is a lot of
energy in a load dump event.


Best regards,
Spehro Pefhany

 

[Dave]

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a PolyFuse/TVS
arrangement on the input but if the system is to work on 24V I would have to
use a 30V TVS which clamps at around 48V, a little high for the input of my
regulator! Any ideas?

I believe a decent electrical system should have a load dump clamp
somewhere in the system. There is quite a bit of energy in a load dump
so it takes a sizeable device to handle it. I believe alternators can
be purchased with such a clamp. If you can determine that there is a
clamp in your system then you only have to find out what the maximum
clamp voltage will be, and deal with that.

 

[[email protected]]

That's what a load dump is. For an official version of a load dump
transient, see: ISO 7637-2, Test Pulses 5a and 5b. There is a lot of
energy in a load dump event.

Well, that is an icy bucket of water!

However, the clamp does not need to protect the transformer and fet
with the same protection that the chip needs. The 3 watt zener is good
for something like 10 watts for a half second starting from cool, so
the peak current needs to be limited to something like 1/4 amp. If the
chip runs okay with a 270 ohm resistor between the battery and its
positive supply line, a 100 volt, 1/2 second pulse will not destroy the
zener. The resistor must withstand the 20 watt pulse for that time,
also, but I don't think it is so hard to cover that base. The bypass
capacitor across the chip's supply also helps with big, brief pulses.

 

[Jim Thompson]

I believe a decent electrical system should have a load dump clamp
somewhere in the system. There is quite a bit of energy in a load dump
so it takes a sizeable device to handle it. I believe alternators can
be purchased with such a clamp. If you can determine that there is a
clamp in your system then you only have to find out what the maximum
clamp voltage will be, and deal with that.

Try 400V at 60A ;-) Clamping is not so wonderful an idea. Various
isolation/disconnection schemes work better.

...Jim Thompson

 

[Dave]

Try 400V at 60A ;-) Clamping is not so wonderful an idea. Various
isolation/disconnection schemes work better.

...Jim Thompson

Yeah, a load dump is a nasty thing. In either case; clamp or
disconnect, the OP should determine whether a load dump handler (of
some sort) is already present in the system.

 

[Jim Thompson]

Yeah, a load dump is a nasty thing. In either case; clamp or
disconnect, the OP should determine whether a load dump handler (of
some sort) is already present in the system.

Imagine 1963-1964, my alternator regulators were blowing apart like
sticks of dynamite for no "apparent" reason. Then I bought a storage
'scope... a _very_ expensive item in those days ;-)

...Jim Thompson

 

[AJ]

Hi all,

I am just looking for some idea's on 24V automotive systems. I have a
regulator that I would like to use (LM3488) which has a max input of 40V.
To protect against load dump and transients I was thinking of a
PolyFuse/TVS arrangement on the input but if the system is to work on 24V
I would have to use a 30V TVS which clamps at around 48V, a little high
for the input of my regulator! Any ideas?

Regards

AJ

Hi guys

Thanks for the feedback thus far. I was thinking of using a 5KP30A
originally and the regulator will be running at around 800mA. Im not sure
how much energy transzorb may need to dissipate, could I get away with using
a lower power one to get a lower max clamping voltage? I like John's
idea, "A shunt clamp using a transistor and zener following the PolyFuse"
and have also found this site
http://www.maxim-ic.com/appnotes.cfm/appnote_number/760 which show a circuit
I may be able to use to disconnect the load during transients and spikes.
This seems like a good way to go to me, has anyone tried any of these
circuits or have any feedback about them?

Best Regards


AJ

 

[John Popelish]

Hi guys

Thanks for the feedback thus far. I was thinking of using a 5KP30A
originally and the regulator will be running at around 800mA. Im not sure
how much energy transzorb may need to dissipate, could I get away with using
a lower power one to get a lower max clamping voltage? I like John's
idea, "A shunt clamp using a transistor and zener following the PolyFuse"
and have also found this site
http://www.maxim-ic.com/appnotes.cfm/appnote_number/760 which show a circuit
I may be able to use to disconnect the load during transients and spikes.
This seems like a good way to go to me, has anyone tried any of these
circuits or have any feedback about them?

Do you need to protect the power switch or just the control chip?
Isn't the power switch rated for a lot more than 40 volts?

It seems to me that they do not need the same level of protection.

I would also consider a lead lag network between the power input and
the sense node to feed forward input voltage compensation.

 

[AJ]

Do you need to protect the power switch or just the control chip? Isn't
the power switch rated for a lot more than 40 volts?

It seems to me that they do not need the same level of protection.

I would also consider a lead lag network between the power input and the
sense node to feed forward input voltage compensation.

Hi John

I was trying to protect the LM3488 controller which has a max input of 40V.

 

[John Popelish]

Hi John

I was trying to protect the LM3488 controller which has a max input of 40V.

But it doesn't need any where near 800 mA, does it? My point is that
you can put quite a bit of resistance between the chip and its
protection to limit the current while connecting the fly back
inductor/transformer much more directly to the battery line.

 

[AJ]

But it doesn't need any where near 800 mA, does it? My point is that you
can put quite a bit of resistance between the chip and its protection to
limit the current while connecting the fly back inductor/transformer much
more directly to the battery line.

I didn't think of it like that.... It only draws 2.6mA so I could get away
with a 1K resistor is series, that voltage drop is ok for my application!

Thanks heaps mate!

 

[Dave]

Imagine 1963-1964, my alternator regulators were blowing apart like
sticks of dynamite for no "apparent" reason. Then I bought a storage
'scope... a _very_ expensive item in those days ;-)

...Jim Thompson

What sort of regulator could you build in 1964? Germanium maybe?

 

[John Popelish]

I didn't think of it like that.... It only draws 2.6mA so I could get away
with a 1K resistor is series, that voltage drop is ok for my application!

Thanks heaps mate!

You're welcome, but don't get that excited. You still have to add the
current consumed to drive the gate capacitance. But I think excluding
the power section makes a low cost zener (1 to 3 watt surface mount)
look like a practical clamp. About $0.25 in quantity.