This video by Machine Design magazine shows what is wrong.
The detent plunger was too short and it's spring too weak to firmly hold
the switch in the "run" position. So, extra weight on a keyring could
turn the ignition switch off. (By swinging I suppose.)
The way I understand it your key chain would have to look like the
men's room key at a highway gas station.
It still takes a significant amount of weight to turn that key.
This is just typical of modern automotive engineering. They are not
looking at how big would be best, only how small they can get away
That is true of just about everything in a car these days and explains
why a 6 MPH bump will do thousand of dollars worth of damage. They
only have to design for 5.
On Sunday, May 18, 2014 12:59:00 PM UTC-4, firstname.lastname@example.org wrote:
But in this case, I haven't seen any evidence that there
was any profit motive for the ignition switch plunger being
1/8" too short. An eighth of an inch additional length of a
tiny pin in a switch shouldn't make a difference in the cost
of the switch. Especially when you're GM and buying millions
Two other things about this. One is that there is a family
suing GM over this. I believe the parents were killed. They
were in the front seat, not wearing seat belts. Their infant
was in the back, not sure about in what. The child survived
and ia a paraplegic. Here's the interesting part. They
were driving along, apparently normally. A drunk driver
crossed the road, hit them head on. It was very major damage,
crumpled all the way to the fron seats. The airbags did not
go off and supposedly the key was found in the accessory
position. The only plausible scenario is that the force of
the collison caused the key to turn *before* the airbag
system could fire? I guess it's possible, maybe
because the airbag system needs enough force for it to go
off and by that time the decleration has caused the keychain
to move the key. So, it's a race between key turning and
The other crazy thing is GM had an instructional presentation
to it's employees on what they shouldn't say in memos discussing
engineering, safety issues, etc.
When I first heard about this, I didn't think it was so bad,
as I can see a reason for not using some phrases that could look
bad in court someday. But if you look at that list, it's a joke.
Some of the forbidden words make some sense:
Some I don't even understand:
Some I've never heard used in my life:
And the worst part is that on that same list are:
It looks like some nutty lawyers are running GM. I was thinking
of how difficult it would be to write a memo about something
without using those words. And those words were just examples.
When always and safety are on the list, IDK how you figure out
what should and shouldn't be on.
On Sunday, May 18, 2014 1:32:20 PM UTC-4, trader_4 wrote:
I was talking to someone else about the above and they came up
with a scenario that may explain not only the above, but many of
the other crashes. IDK about you folks, but watching the media
reports they showed some of the crashed cars where people died,
and talked about the key moving, causing the power assist to the
steering and brakes to stop, and the airbags failing to deploy.
From that I assumed the scenario was that the key moved *first*
causing the loss of power steering and braking making the car more
difficult to control. The steering being more difficult, I get,
because that's going to happen instantly. The braking, given that
there is a vacuum reservoir, I would expect to be perfectly fine
for at least one full application of the brakes. If you pump, pump
a few times, then you'd use it up. So, I was wondering how the
key moving initiated the crash, since that is what the reporting
seemed to suggest.
But talking to a friend, he came up with a different scenario
that explains the specific odd crash I brought up in the case
above and could explain many of the other crashes too. In the case
above, the car was working fine, until a drunk driver came across
and hit them head on. What would a driver in the GM car likely do
at some point prior to impact in that case? Slam on the brakes.
And doing so, with a heavy keychain, the extreme deceleration
causes the key to move. That cuts off the airbags. So, a similar
thing could have happened in many of the other 13 deaths, ie
it wasn't the key moving that instigated the accident, through
the steering/braking being more difficult, but some other event.
The scenario being you're running off the road already headed to
a tree, whatever, you slam on the brakes, that moves the key, turning
off the airbags.
On Sun, 18 May 2014 10:32:20 -0700 (PDT), trader_4
That is why I say it is just the culture of modern engineers. They
make everything as small as it can be, even if there is no economic
reason to do it.
I preferred the olden days when they did not do that much engineering,
they just put in a big chunk of metal and knew it would never break.
I bet someone thought those bigger detent "clicks" did not project the
smoothness they wanted to project.
perhaps it would be safer if airbags always had power. since the worst that could occur and deploying some airbags in parked cars with no one inside
I have sat in cars waiting to meet someone with the ignition key off, or even removed.
ideally people should still be protected
If you are in a parking lot, the impact would not be hard enough to
trigger the airbags. If you got rear ended, the airbags would not deploy.
I'm not even sure if they would deploy in a front collision of a parked
car. I think rate of deceleration is a factor in the trigger. Standing
vehickles would not have deceleration, only impact.
On Monday, May 19, 2014 2:23:49 PM UTC-5, Ed Pawlowski wrote:
Physics 101: Impact IS deceleration.
Acceleration/deceleration is a change in speed. You are going from zero to
a negative forward speed if struck from the front while standing still. An
accelerometer doesn't know the difference, unless it is somehow hooked up t
o the speedometer, and I don't think that's how airbags work.
I am sure it is not this simple in reality, but one would think that if fro
nt-ended while idling, the airbags SHOULD deploy.
If a car is stopped and the engine is shut down; that is another issue enti
rely, because you would have to determine whether the airbag system still h
as power while the engine is not running.
On Monday, May 19, 2014 7:05:04 PM UTC-4, Nicholas Kriho wrote:
Wrong. Deceleration is a slowing down. In the example you're replying
to, a parked car being hit was cited. That is an *acceleration* of the
parked car. It doesn't matter from which direction it's hit. From
Newton, F = MA. Hitting the car applies a force and it accelerates.
That may be true, but it's not deceleration. The parked car is accelerating.
It's going faster 100 ms after being hit than it was at 5ms.
An accelerometer doesn't know the difference, unless it is somehow hooked up to the speedometer, and I don't think that's how airbags work.
An accelerometer would record an acceleration, not a deceleration. The
parked car is speeding up, not slowing down. It's going faster 100 ms after being struck than it was 5ms after being struck.
And actually I'm pretty sure speed is also a component the computer looks
Logically you would think so, but it depends on the exact criteria that the
airbag computer uses.
Apparently from the media reports, the allegation, which I haven't seen
GM refute, is that the airbags did not deploy in these fatalities because
the ingnition key had moved from run to acc.
Does anyone know if that GM ignition switch recall extends to Canada or
not and whether the Oldsmobile Alero is one of the models being
recalled? My sister has an Oldsmobile Alero, but I don't know what
model year it is.
excuse me my wife owned a cobalt, it had all sorts of lost power, engine qu
it, etc etc. then local dealer MUST of known the car had a design defect, i
t must of been to the dealer 15 times. my wife was very short and heavy and
her key ring had everything but the kitchen sink on it..... power steering
she fit the profile for the problem exactly. we got divorced....
if you ask me GM should be required to pay every original owner who had com
plaints that fit the defect 5 grand each. so no manufacturer will ever cove
r up such a defect again!
Aren't cars made during the past 10 years (and moreso 5 years) supposed
to have sophisticated "black-box" data logging systems that record a lot
of what's going on with the car (steering wheel movements, braking,
speed, etc) compared to 10+ years ago?
So much more logging that it's become standard practice for cops to have
sophisticated data recovery tools sold to them by third-party equipment
makers with slick marketing that tells cops that there's valuable
evidence just waiting for them at crash scenes?
If so, then it should be a simple matter to know the state of the
ignition switch or airbags prior to, say, loss of main battery bus or
engine RPM offscale or zero (the loss of either one being the indicator
of impact damage in progress).
I did a bit of reading on air bags tonight, and whether or not an airbag
deploys, and which of the airbags deploy is not just a matter of how
strong the impact is.
If the ignition is off, then there's no power to the vehicle's computer,
and so no amount of impact will cause the airbags to deploy because the
computer will not generate a signal to deploy the air bags.
In a collission, or if the vehicle's airbag sensors sense an impact,
then within 15 to 20 milliseconds the vehicle's airbag computer decides
whether or not to deploy an air bag and which air bag(s) to deploy. It
takes into account the direction of the impact and the weight of the
passengers in the seats. If the car gets hit from the back, then the
passengers are pushed back in their seats, so deploying the front air
bags won't help at. In that case, no matter how hard you're hit from
the back, the front air bags will never deploy. If the car is moving
slowly when the impact occurs, or the driver in the front seat is very
small (like 70 pounds or less) then the front air bags may not deploy
because the explosive deployment may cause more injury to the driver or
passenger than the collission. In any collission or impact, the car's
computer does a lot of calculations and within 15 to 20 milliseconds it
decides whether or not to deploy air bags, and which air bags to deploy.
Most of the time when airbags don't deploy it's not because the impact
wasn't severe enough, it's because the car's computer decided not to
deploy any air bags because the conditions didn't warrant it.
Air bags have an redundant deployment system. They require electrical
power from the car's battery or alternator to deploy, but if electrical
power to the airbags is interrupted as a result of the collision, air
bags have a power capacitor that provides enough power to deploy the air
bags for up to 90 seconds after the crash. So, if your car gets hit in
the left front fender which causes the car's battery to go flying, the
power stored in that capacitor still operates the car's computer and
provides enough power to deploy the airbags in the car for up to 90
seconds. When you park your car, if it gets hit from the front by
another car, the air bags can deploy for up to 90 seconds after you turn
off the ignition. After that 90 second window passes, then there won't
be sufficient power to run the car's computer or deploy the air bags, so
no matter how hard you're hit after 90 seconds of turning off the
ignition switch, the air bags won't deploy.
Or, at least, that's what I understand now.
On Monday, May 19, 2014 6:10:46 PM UTC-4, nestork wrote:
I think here you mean that if the key has been off longer than the
stored energy period that you talk about below.
I saw on a car talk website where they said a similar thing,
ie that the bags have some reserve energy to deploy for some short
period eg couple mins, after the key is off. If that's true though,
then how do you explain the GM car crashes at issue, where the airbags
did not deploy? All the media reports I've seen say it's because the
key was not in the on position.
On Mon, 19 May 2014 16:39:57 -0700 (PDT), trader_4
The air bags are fired electrically and I assume they are wired to the
ignition position not the accessory position.
GM had enough "dead battery" problems that they try to turn off as
much as they can whenever they can.
On Monday, May 19, 2014 9:00:44 PM UTC-4, email@example.com wrote:
That's what I thought too. But according to Nestor, he found where
there is an energy reserve of x minutes, for the airbags. I found a
But I think the real answer is that it likely varies from manufacturer
to manufactuer and even NHTSA is confused. NHTSA testified to Congress
that they do have a reserve, like Nestor said, but GM says it's recall cars
airbags go inactive 150ms after the key moves from run.
"There might actually be a bit of a silver lining to General Motors' igniti
on switch recall of 2.6 million cars. In the end, it may mean safer vehicle
s on the road from every automaker. The debacle has shined a light on how l
ittle the National Highway Traffic Safety Administration actually understan
ds about airbags and their deployment. The regulator is now working to chan
ge that, and it's investigating how to make the devices even safer.
In its Congressional testimony, NHTSA said that it believed the General Mot
ors recalled vehicles had 60 seconds of power for the airbags after their i
gnition switches were turned off, according to The Detroit News. The automa
ker has denied this - its own research has found that the bags will only wo
rk for about 150 milliseconds after the ignition has been moved from the ru
This discrepancy between GM and NHTSA investigations may lead to major impr
ovements in how airbags work. The regulator has begun surveying automakers
and airbag suppliers to learn more about how ignition position affects acti
vation, according to the report.
At the moment, these airbag activation parameters remain unregulated. While
the pyrotechnic devices legally have to be in cars, automakers tune their
deployment differently, depending on their own internal standards. NHTSA's
analysis could eventually lead to new regulations determining whether these
safety devices should continue to work for a period of time after a vehicl
e is shut off. "
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.