Vehicle Transients

[Strube09]

Hello everyone,

Anyone have any insight to a vehicles electrical systems and transients seen during operation, startup, and shutdown?

I have some a few products in the field that are blowing up while operation in a car on a 12V Cigarette lighter adapter. Some of these products have been in the field for several years but it seems as though we have been getting more and more failures of this nature. I am not sure if the issue is seen on newer or older cars. And most of these cars are in the police/fire field. The only thing I can seem to trace it back to is a high input voltage / surge.

Thanks for any help,
Strube

[Simon]

maybe something has been added to the vehicles ? what sort of fail is it ? if you suspect spikdes try a spike protected version ? and see what happens ?

how old are trhe failed units ?

[Fraser]

My experience has been with mobile car GPS units....

Two failures are common... during cranking the 12V supply rail can have some pretty ugly transients present on it. My car and many others disconnect the cigarette lighter during engine cranking to protect connected devices. When running, the '12V' rail is actually nearer 14V so equipment has to be able to tolerate that continuously. Whilst transients can occur as relays open and close, it is not normally as serious as during engine cranking and automotive appliances should be adequately protected against such occurrences.

GPS units are built down to a price and that includes the charger/power supply unit that plugs into the Cigarette lighter socket. Two failures are common. Failure of the DC-DC converter due to supply transients and DC-DC output over voltage due to fracturing of the voltage sense wire in the cable. This normally occurs at the GPS unit power plug due to regular flexing. Explanation... many GPS power supplies are a three wire system, .... 0V, +5V and sense. The sense wire is a feedback to the DC-DC converter so that any voltage drop across the power lead can be compensated for by increasing the DC-DC output slightly. If the sense wire fractures, the DC-DC converter applies the full 12V or nearly 12V to the GPS and lets all the magic smoke out  

Some GPS manufacturers advise that the GPS should not be connected during the cranking of the engine for start-up. This is because they know their DC-DC converters can be damaged by transients.

I have had people confused by transient related failure, as they wonder why a modern car can have loads of sensitive electronic modules in it's design yet it can still kill accessories with transients. My answer.... automotive electronics manufacturers understand the environment in which their product will operate and take appropriate precautions. Some accessory manufacturers build down to a price and fail to take said precautions.

If you are seeing a lot of failures of a particular brand of appliance it would be worth your while having someone do a post-mortem on it to establish the cause of failure. It may be possible to take some simple action to protect the appliance from future failures and so save both money and your equipments reputation.

Be carefull if you decide to connect an oscilloscope to a cars 12V rail... some of the engine cranking transients can be high peak voltages and potentially bad for your scopes health !

[Neilm]

I did work with military equipment working at 24V. The input filters had to be rated at a very high voltage. I don't remember it exactly but I think it was about 60V.

Neil

[jimmc]

I don't think the full specs are available on line (for free) but this may give you some idea as to what you're up against
 http://www.tracglobal.com/transient-specification-iso7637-22004.html

Jim

[Fraser]

I tend to shy away from auto electronics...nasty things to work on...dirty cramped and these days, bl**dy complicated 

I have a diagnostic PC based OBDII tool called VAG-COM for diagnosing faults and configuring modules on my Audi. It's very useful for tracking down faults by interogation of the various control modules that reside on the CAN Bus. It will even diagnose faults in the supply rails, alternator etc. It impresses me that a CAN Bus and associated microprocessor modules work so reliably inside such a hostile environment that is the average modern motor car   

My other car is a Mini Moke (remember those Dave ?)..... not a single semiconductor in her, just 'battery and bulb' technology... DIY heaven   

[Strube09]

Thanks for some quick responses,

Some units have been in the field for a matter of years... The most recent unit was only in the field for 6 months.

The part that is failing is the main control FET of the DC/DC converter. Unfortunately I can't do much post mortem as the unit is burnt so bad even the PCB and copper are ruined.

I think one good question for this customer is to ask if the cigarette adapter is a directly connected or one that is attached to the accessory. This might help indentify if it was subject to transients during the cranking.

thanks,
Strube

[Simon]

I would guess that you have less spikes on the battery (as it will act a bit like a capacitor), hopefully the sockets are connected straightish to the battery ?

[tyblu]

I asked a relevant question: http://electronics.stackexchange.com/q/8798/2118

[Simon]

I asked a relevant question: http://electronics.stackexchange.com/q/8798/2118

hm interesting, I saw the 120V and 400mS figure when i was trying to resolve problems with equipment being damaged in a military vehicle but that was 24V. So basically automotive stuff needs to withstand spikes well in excess of 10x the rated supply

[Zero999]

Load dump, in automotive electronics, refers to the disconnection of the vehicle battery from the alternator while the battery is being charged. Due to such a disconnection of the battery, other loads connected to the alternator see a surge in power line. The peak voltage of this surge may be as high as 120 V and the surge may take up to 400 ms to decay.

Why on earth would the battery be disconnected from the alternator while it's charging?

Surely this would only happen when there's a loose connection? If the electrical system is sound, a load dump will never happen?

[Simon]

this would be a big issue on older cars where the crude alternator regulator will hardly function without the balast of a battery. It may be more of a historical reference although still something the device must be prepared for in modern applications

[Psi]

I always put a transient voltage suppressor in anything i build for my car.


Also, as someone has already said, you do have to watch the voltage, as 14.5V is common while the car is running, 14.5 is to much voltage for most 12V devices.

Using a car PC's for an example.  years back VIA produced a motherboard TC10000 i think, one of the reasons it was designed was because of requests from people wanting to build carPC's and only having access to 12V.

They responded with the TC10000 but sadly they failed miserably. It turned out to be only rated from 11.4V to 12.6V so was useless for cars. To fix the issue I ended up buying a specially designed voltage stabiliser,  6-24V input  12V out @ 12A .

[oPossum]

The part that is failing is the main control FET of the DC/DC converter. Unfortunately I can't do much post mortem as the unit is burnt so bad even the PCB and copper are ruined.

I see this frequently in automotive electronics. The most common cause is undervoltage. There is not enough gate drive to keep the MOSFET out of saturation, so Rds On rises and it smokes.

Often caused by corrosion of power connectors or improper install (bad ground, wiring undersized, etc..)

[Strube09]

The part that is failing is the main control FET of the DC/DC converter. Unfortunately I can't do much post mortem as the unit is burnt so bad even the PCB and copper are ruined.

I see this frequently in automotive electronics. The most common cause is undervoltage. There is not enough gate drive to keep the MOSFET out of saturation, so Rds On rises and it smokes.

Often caused by corrosion of power connectors or improper install (bad ground, wiring undersized, etc..)

Hmm Interesting, I had not considered and undervoltage situation.... I am pretty sure that the way I design the circuit that it just wouldn't start up. But I never really tested and undervoltage situation. I Think I know what testing I am going to do tomorrow

[Simon]

I dbet during start up the voltage plumets, might be repeat undervoltage situations that have dome the damage over time

[jahonen]

I think LTC3728 Datasheet from Linear Technology says it pretty straight:

But before you connect, be advised: you are plugging into the supply from hell. The main power line in an automobile is the source of a number of nasty potential transients, including load-dump, reverse-battery and double-battery.

One possible method for surge protection is to have a MOSFET (or two if you want also polarity protection) to disconnect the supply in case of overvoltage, if your device does not need much power. Clamping approach is not very practical if the device is small, unless you can ensure that there is enough impedance in supply line to dissipate the surge.

Regards,
Janne

[Strube09]

Thanks for all of your replies.

Any good ideas to simulate this issue? I need to be able to create these pulses and show the issue is repeatable.

I am going to play around with some under voltage situations today but I need to come up with a good demo for the transients.... Hmm Maybe I need to buy more equipment ?

Oh and I might add a bit more that makes this issue even more serious... The unit that blew up, burnt the board past the point of testing, and melted the outer plastics happened in a fire marshals truck.... The customer isn't too happy :/

Thanks,
Strube

[jimmc]

This any help? 'Suppression of Transients in an Automotive Environment'
(http://www.littelfuse.com/data/en/Application_Notes/an9312.pdf)

Jim

[Neilm]

I am going to play around with some under voltage situations today but I need to come up with a good demo for the transients.... Hmm Maybe I need to buy more equipment ?

There are pieces of test equipment that are designed for EMC testing that will probably do what you want. They would not be too expensive to hire for a couple of weeks.

Yours

Neil

[Galenbo]

...Be carefull if you decide to connect an oscilloscope to a cars 12V rail... some of the engine cranking transients can be high peak voltages and potentially bad for your scopes health !

Can you (or someone else) please share more info (=experiences) about this?
I have a normal basic Rigol scope, and did measurments in the past with a basic handhold Velleman Scope.
The frontplate of my scope says: 1Mohm, 18pF, 300V rms, Cat I
The probe says 10x 300V Cat II

Are there cheap workable solutions for this? I don't need the 200MHz or 0.xxx1V precision/accuracy for my cars, so the cheap/sure solution can have multiple disadvantages. Do inline BNC surge protectors provide a solution?

[Simon]

maybe put a tvs in parallel with the scope input that trigers at 80 or 90% of your scopes limit.

[jitter]

Funny, in the Beginners section a thread with more or less the same topic is going on: https://www.eevblog.com/forum/beginners/vehicle-transient-protections/. Perhaps info relevant to this topic can be found there too...

[Rerouter]

Automotive is where building to a cost gets kicked in the face,

Load dump also occurs on a number of newer cars, and there are a surprising amount of european cars that can be bricked if cranked for more than 5 seconds on a dead flat battery... to the extent that the dealers bundle in a wall charger for the battery over the weekends, equally some australian cars suffer the same fate, the 2002 ford falcon, the battery begins to go high resistance, the coil pack driver goes erratic and it fries the instrument cluster through the rpm sense wire.

The rule of thumb is built for a repeated 1 second transient at 5 times the voltage you expect, so 60V for a 12V device and 120V for a 24V system, your device doesnt have to work while encounting this, just survive,

Load dump is generally in situations where the battery is just on the edge of being too flat to start and someone tries to crank it over, the surface charge gives an initial surge of current to the starter, then once that is depleted it goes high resistance, and the inductance of the starter pulling large currents now encountering this increased resistance give a nice big voltage spike

The other common thing industrial vehicle electronics has to try and survive is welders,

And in probably the most extreme case one device i worked on had to work over a 22KV insulation gap, the installer figured the relays where too expensive and direct wired, so 22KV @ 8A flowed down the cherrypicker via control wires, through our electronics, blew it to high hell and then entered the vehicles power wiring, before arcing out on the road base. the truck was well and truly bricked,

[Andreas]

Also, as someone has already said, you do have to watch the voltage, as 14.5V is common while the car is running,

The spec is 4-16V for normal operation (cranking to charging in winter).
With a dead battery the voltage can be 18V with a 2V AC overlay.
On jump start (start from a truck battery) you have to withstand the 24V for 1 minute.
Load dump was already mentioned.
With a central protection device you have to withstand around 50-60V for 400ms.

And there are further low energy pulses up to +/- 200V in a 12V system.

http://www.fordemc.com/docs/requirements.htm

With best regards

Andreas