Why you should remove the negative battery terminal before doing ANYTHING!!!!

He said there was a spark when he tried to remove the bulb. The bulb's
glass was broken, and he tried to grab it without touching anything else,
but there was a spark, and when he tried to start the truck...nothing.

He had it towed to a Nissan dealer. They're trying to find a used ECU and
told him that before doing ANYTHING electrical with the truck he should...

disconnect the negative battery terminal...

(Next week he's removing the terminal and installing a Negative Terminal
Shut-off...)

 

Guys, guys... go get a schematic and find out if the dome light is in
the same circut as the ECU. Even if it is, I don't think the changing
of the bulb is what caused the problem. I do think that whatever
caused the light to function intermittantly MIGHT have cused the ECU
problem. Either way, if the light was working sometimes, that would
tell me that there was nothing wrong with the bulb. Funny thing about
light bulbs... they either work, or they don't. The filament is either
complete, or it's not. It doesn't go back once it's blown. Okay, so if
there's a short, it would be either in the fixture, or in the wires
going to it. Furthermore, That's what fuses are for! I've done it many
many times. Shorting things out and blowing fuses and replacing them
again. The fuse opens the circut in the event of a short before
ANYTHING gets damaged (unless the fuse is either disregarded, or
replaced with a fuse of a higher amperage). And furthermore, since
we're on the topic of fuses, this is the whole reason that things like
dome lights and tail lights and headlights and radios and ESPECIALLY
ECU's are all on thier own circuts with thier own fuses. Kind of
insurance, if one thing blows a fuse, you don't lose everything at
once. I do agree with the titlehead of this issue, being "disconnect
the negitive before working". But this rule is mostly in place because
sometimes people like to play with the wires and accidently drop the
exposed end of a hot wire onto the bare frame of body of the vehicle.
The result can be a blown fuse at minimum, or a small fire at most
(I've been there, too). However, I regress (and conclude), the dome
light bulb cannot be the reason for ECU problems. ECU's may or may not
be "idiot proofed," but they are not on the same circut as the dome
light, the dome light cannot cause "power surges" and niether can
anything else on a vehicle, and even if anything were to happen, the
fuses would have caught it. Case Closed. Go get a new ECU, and re-wire
your dome light. quit griping about crap that is not relevant. The
more you gripe, the more other people despise you, and the job still
is not done.

 

Yeah, but we're talking a Nissan here! I've seenthings in Nissans I've
never seen in other cars.

And, he said the glass was broken. How it even lit without burning out is
a mystery!

 

A welder and a dome light are vastly different things. An arc
welder generates a HUGE reactive voltage spike when striking the arc,
and that spike can wander all over the entire vehicle and fry
sensitive electronics and not-so-sensitive things, too, like
alternator diodes. Standard procedure there is to disconnect the
battery whenever doing any welding on the thing. Shorting a dome light
will NOT generate any sort of spike. Period. You need a coil to
generate spikes, coils like those found in starters, alternator
rotors, ignition coils. You might as well try to generate a spike by
disconecting and reconnecting the battery.

Dan

 

Hmmm,
Lamp filament is a tiny coil, LOL!

 

with an air core and virtually no inductance. if you have instruments
that can measure any voltage spike you get from a coiled bulb filament,
i'd love to see your readings.

 

Hmmm,
You are so bone headed, can't even take a joke.
The more you babble, the more you reveal your lack of knowledge:
pratical or thory. My job used to deal with fraction of nano amps. VLSI,
ASIC, etc. on mil-spec.

 

bullshitting idiot.

 

I do.

BUT, you don't need one. If you have an 1141 lamp with 50 loops of
1/16" (that's estimated by eye here) about 1/2" long, you can use
Wheeler's Formula:

50^2 * (0.065)^2 10.5
--------------- = ----- = 0.27 microhenries
18 * 0.065 + 40 * 0.5 28.0


That's a lot less than the hundreds of henries that an arc welder ballast
will have, and it's probably small enough to be compensated for by the
capacitance of the wiring, even. But it's enough that you could measure it
carefully with a scope and a pulse generator, even though it's actually
going to be swamped at any reasonable voltage by the nonlinearity when the
filament heats up and its resistance increases.
--scott

 

I personally have seen a bunch of blown ECU's caused by wiring shorts,
mostly fuel pumps, but other things too.

Sure the ECU is 'protected' by a diode, but no one sells the damn diode
to replace the blown one nor do many 'authorized' places exist to fix
them. Half are sunk in a block of plastic also.

I have repaired a few by using a cheap diode with a +/- 10 to 15%
tolerance so they fell close to the original specs. Close enough to work.

If the dome light is on a timer, it is connected to some kind of
computer....

Mike
86/00 CJ7 Laredo, 33x9.5 BFG Muds, 'glass nose to tail in '00
'New' frame in the works for '08

 

First you said that the bulb was in pieces, not that the glass was broken.
Was it really broken?
I have seen several times that the glass part of a bulb can separate from
its socket but still hanging by the leads it may still work.
And I have also seen that the ends of a broken lamp filament can dangle
together and get get kind of welded together, so it works again, but usually
not for long.
What may kill an ECU is the very very short transients in voltage when a
cirquit somewhere in the system is abruptly opened or closed, especially
when it is high current (a short) and/or repeated many times. If he made a
momentarily short cirquit while changing the bulb that MAY be enough.
That is why with modern cars we are advised not to use jumper cables without
transient suppressors.

Asbjørn

 

Hello:

This might seem odd to you but, newer cars use bulbs that are a specific
impedance. A 2157 is the same candle power as an 1157, but the impedance is
higher. That's why if you use a 2157 in place of a 3057 for argument's sake.

You'll get a check engine light in an OBD2 compliant vehicle.

RK

 

The very reason that using a wrong bulb:

Will set a code on an OBD2 system.

RK

PS
I'm not speculating on Jim's method of earning a living. But there are such
instruments available.

 

that's load measurement, not inductance from a bulb filament coil.

of course! but i bet t.h. doesn't have any, and i'll bet he doesn't
know how to measure whether it has any noticeable effect!

 

excellent. and it's /well/ below any level that could /possibly/ impact
the ecu!

indeed.

 

You seem to forget that the dome light is 'live' all the time using a
ground as a switch.

They can have the glass fall out of the base and the loose internal
connecting wire(s) can and will cause a short circuit. This happens all
the time with bulbs such as the 1157 in taillights.

Short circuits can do nasty things to computers like blow the input
diode 'before' the fuse has a chance to blow.

It's not like the designers of modern vehicles are just BSing when they
say to unhook the negative before working on them.....

Mike
86/00 CJ7 Laredo, 33x9.5 BFG Muds, 'glass nose to tail in '00
'New' frame in the works for '08

 

Hachiroku ãƒãƒãƒ­ã‚¯ wrote:
\

I don't buy it. ECUs are actually *very* well protected against
intermittent faults, voltage spikes, etc. (Zener diodes are a wonderful
thing). 99.9% of the "failed" ECUs people pull out of cars are actually
fine and the problem lies elsewhere. There's no WAY you can hurt one
just changing a light bulb.

 

That much is not true in all cases. Any care that has "theater fade"
interior lights, for example, has them powered from a computer module
(not necessarily the engine controller, but a "body controller" or some
other computer module.)

And on all of THAT, I agree completely!

 

Think "fuse" in this case.... unless it was one of those Chinese-made
fuses from Harbor Freight... ;-)