When is 12V so much more?

[Ubergeek63]

when it is an automotive power rail!

there is a reason that "automotive grade" parts have 60V ratings people...

 

[tcmtech]

And that's supposed to mean what?

 

[Ubergeek63]

And that's supposed to mean what?

that means that unless your circuit can protect the load from 80V for a half a second you risk blowing out your $500 iPhone. Those are official automotive industry standards, not mine.

Now if whoever it is does not mine risking that much, it is on them. I know I sure can not afford to be replacing critical stuff, never mind overpriced trinkets from Any Person Purchasing Lacks Experience.

 

[Diver300]

voltage drop across Q1 can be reduced to
10s or 100s of milliamps

That mix of units reduces my respect for the author of the paper.

 

[Ubergeek63]

That mix of units reduces my respect for the author of the paper.

respect for the author does not enter into the fact that the industry recognizes the very real possibility of seeing 87V on the DC rail for 400 mS, 200mS of which are over 40V

 

[Diver300]

I would agree that large voltage transients like that come from load dump. I've seen an engine management computer fried when the earth lead from the engine to the body was intermittent. There was a transient voltage suppressor, and that eventually failed short circuit and took out the fuse. The TVS had protected everything else, and once the earth lead, the fuse and the TVS were replaced it was back up and running.

I've designed a protection circuit that disconnected the supply, much like the P MOSFET one in the paper. I don't know why people would be worried about the distinction between a Class A pass (where the item works perfectly) and the Class C pass (where the item stops during the test and recovers) for a load dump test, as the load dump can only happen when the battery is disconnected. In that case, the power is likely to fail completely any moment.

 

[tcmtech]

So if they expect regular spikes of 87 volts for nearly half a second or so why are most electronics in automotive systems only using 16 -24 volt rated capacitors and similar rated components many with no surge or spike dampening protection of any sort?

I have dissected and worked on many electronic modules, components and electronic controls over the years and rarely have I ever seen any form of surge or spike suppression on the 12 volt feeds.


I have several old control modules from a Ford pickup I scrapped out now sitting on my computer desk today and all of them have nothing more that a 24 volt rated capacitors directly on the 12 volt supply lines where they attach to the circuit boards.

 

[MikeMl]

You will get away with it as long as you never disconnect the battery with the engine running fast, right after starting the engine.

A load dump happens if the alternator is sourcing current for charging the battery and the battery lead is suddenly disconnected. The current that was formerly going into the battery suddenly has no place to go, and this causes the positive voltage excursion. If you never disconnect the battery with the engine running, you will never have a load-dump.

 

[Diver300]

Load dump is very rare. You don't expect 80 ish volt spikes for 1/2 second, it's just that they can happen in certain circumstances.

To get a load dump you need an 80% or more reduction of alternator load over less than 1/4 second, and the situation where none of the remaining loads will take the current. Any car battery that was capable of starting the engine will prevent load dump damage, because the battery will absorb the alternator output and keep the voltage down until the alternator regulator has reduced the output.

The only way to get it happen is for the battery to be disconnected while it is being charged by the alternator with quite a large current, and when there isn't much other load on the alternator. That's why car handbooks say that you shouldn't disconnect the battery with a running engine.

A very large proportion of car electronics will survive years of service with 25 V capacitors on the 12 V line, as they will never experience a big load dump. Also, I don't know how long a 25 V capacitor would last on 80 V, but it might be longer than 1/2 second, so your modules might well survive one or two load dump events.

(edit) I think that MikeMI and I said the same things, at the same time!(/edit)

 

[Ubergeek63]

So if they expect regular spikes of 87 volts for nearly half a second or so why are most electronics in automotive systems only using 16 -24 volt rated capacitors and similar rated components many with no surge or spike dampening protection of any sort?

I have dissected and worked on many electronic modules, components and electronic controls over the years and rarely have I ever seen any form of surge or spike suppression on the 12 volt feeds.


I have several old control modules from a Ford pickup I scrapped out now sitting on my computer desk today and all of them have nothing more that a 24 volt rated capacitors directly on the 12 volt supply lines where they attach to the circuit boards.

simple... tranzorbs blow the line fuse. they do not have to be very big since the electronics modules are generally on their on fuses rated for just the module. a random load dump blows the radio fuse, it gets replaced and forgotten. but that is a far cry from the 10A available without blowing the lighter fuse!

 

[tcmtech]

Cant say I have ever replaced a fuse from anything other than a short circuit or some new added component drawing more than what the original circuit was fused for.

As far as the battery coming disconnected during a high drain event I jump start stuff fairly often and have a number of time had a weak battery connection come undone or lose their connection even when running retrofit high output alternators running at full output while jump starting big diesel engines on tractors and other heavy equipment.

Still I never had to replace fuses or do anything else for electrical repairs afterwords other than clean and fix the battery connection.

Basically I am still not seeing what all the over voltage fuss is about. Modern solid state regulated alternators have very good response times unlike the old mechanical points regulated DC generators of long ago.

 

[Diver300]

There is far less problem with DC generators, also known as dynamos. On those the regulator is mainly needed so that the voltage stays the same when the engine speed varies. The voltage won't jump up nearly as much if the load reduces as it would with an alternator. On an alternator, there is a large winding inductance, so the EMF has to be 80 V or more to achieve the rated current. When the load reduces, there is far less voltage drop due to the inductance, so most of the EMF appears on the output terminals.

I've seen damage from load dump. It was caused by repeated disconnections from a loose connection. Also, I've seen that the slow response and jumpy voltage of systems without a battery are real. When I've been testing car electrical systems without the battery, the slow response of the alternator is really noticeable. If a big load is turned off, the remaining lights get really bright for a second or so. When a big load is turned on, the voltage dips sharply, and recovers after a couple of seconds, or the engine stops.

 

[ronsimpson]

It is too common for people to install a new battery backwards.
For trucks, I have seen '24 volt' jump starts.
People use electric welders to jump start cars & trucks.
If the battery connection is bad or a cell fails it is the same as no battery.
Alternators can produce some serious voltage with no battery, or bad battery.
On a series of school buses, they had a bell to make a noise when backing up. It was a solenoid and hammer. That made +/- 100 volt spikes.
The automotive designs I did were testes at 100 volts for u-Seconds and 48 volts for seconds and reverse voltage (no failure, fuse OK). The last test was to plug it into 110 VAC. It must only pop the fuse.
I know the power in a new-good running VW looks clean. When new is long gone, good running is not happening, some fool has tools or the battery dies you can expect very large voltages for short periods of time. There is a reason why we use high voltage load dump.

 

[OutToLunch]

i would suspect that if there is a load that requires a good amount of current, it is located relatively far from the battery, and it is cutoff quickly then there could be a localized spike on the rail.

 

[Noggin]

My brother in law spent almost $50k building a twin turbo Mustang over the last 5 years. He finally got it finished about a month ago and it dyno'ed at over 1200 HP. Last week, something happened and every electronic module in the car smoked. Wonder if it was a load dump... He said it is going to take about $7k to fix. Sucks that all it would take to protect against that would be a 100-200 amp, 100v fuse from the alternator and a 24v TVS diode. Doesn't look like a 200amp 100vDC fuse is very common though

 

[MikeMl]

i would suspect that if there is a load that requires a good amount of current, it is located relatively far from the battery, and it is cutoff quickly then there could be a localized spike on the rail.

But that is not a "load dump" in the same context as being discussed here. What you are describing is due to the inductance of the long wiring to the load. It creates a very short spike.

What is being discussed is the long-duration voltage surge (spike) created when a spooled-up alternator is suddenly disconnected from a battery which is acting as a sink for the alternator current.

 

[Diver300]

My brother in law spent almost $50k building a twin turbo Mustang over the last 5 years. He finally got it finished about a month ago and it dyno'ed at over 1200 HP. Last week, something happened and every electronic module in the car smoked. Wonder if it was a load dump... He said it is going to take about $7k to fix. Sucks that all it would take to protect against that would be a 100-200 amp, 100v fuse from the alternator and a 24v TVS diode. Doesn't look like a 200amp 100vDC fuse is very common though

If you are going for one TVS for the car, then it has to take the whole alternator load for 1/4 second or so. The alternator can't produce more current than it's rating, because it is limited by the inductance of the windings, so the fuse won't ever blow. The TVS has to be very large to take that power. You might need the fuse to protect the wiring if the TVS fails short circuit and the battery is reconnected.

If you are going to protect something small, like a radio or one module, then a small TVS can hold on long enough for a small fuse to blow, but calculating and testing for those circumstances isn't easy.

 

[audioguru]

Most of the "low dropout" voltage regulators made by National Semi have car load dump protection. When the voltage becomes 60V or more then the regulator simply shuts down.