How does computer detect a misfiring cylinder?
Jeff Wisnia
My daughter's 98 Olds (V6) failed state inspection because of a huge
number of misfires recorded as having occurred in one cylinder, and the
"check engine" light was on.
I tried a new spark plug and lead and swapped coils, but the condition
remained so I did a compression check on that cylinder and found it 40
lbs lower than the ones on either side of it. Looks like I'm into
pulling the head and hoping it's just a valve problem and not a piston
ring/cylinder one, and that getting the compression back up to normal
will correct that "misfire" indication.
But I'm wondering, what mechanical or electrical phenomena gets
interpreted by the computer as a "misfire" in a specific cylinder?
I can't guess what it might be and three supposedly knowledgable auto
techs I've asked all thought about the question, wrinkled their noses
and said, "Gee, I don't really know that."
Thanks guys,
Jeff
--
Jeffry Wisnia
(W1BSV + Brass Rat '57 EE)
"Rudeness is the weak man's imitation of strength."
AZ Nomad
>My curious mind wants to know...
>My daughter's 98 Olds (V6) failed state inspection because of a huge
>number of misfires recorded as having occurred in one cylinder, and the
>"check engine" light was on.
>I tried a new spark plug and lead and swapped coils, but the condition
>remained so I did a compression check on that cylinder and found it 40
>lbs lower than the ones on either side of it. Looks like I'm into
>pulling the head and hoping it's just a valve problem and not a piston
>ring/cylinder one, and that getting the compression back up to normal
>will correct that "misfire" indication.
Perhaps the crank sensor is accurate and fast enough to detect the slowdown
every time the bad cylinder fires?
Mike Romain
cylinder. It flickers and misses the regular light pulses. I would
suspect the electricity passes differently across the gap, but am not
sure the 'mechanics' of it...
Mike
86/00 CJ7 Laredo, 33x9.5 BFG Muds, 'glass nose to tail in '00
88 Cherokee 235 BFG AT's
Canadian Off Road Trips Photos: Non members can still view!
Jan/06 http://www.imagestation.com/album/pictures.html?id=2115147590
(More Off Road album links at bottom of the view page)
Shep
sequences as measured in comparative crank speed changes.
"AZ Nomad" <azn...@PmunOgeBOX.com> wrote in message
news:slrneao6v6....@ip70-176-155-130.ph.ph.cox.net...
----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----
Kevin
"Jeff Wisnia" <jwi...@conversent.net> wrote in message
news:mLmdnf3fmIK9izHZ...@conversent.net...
The crank shaft has a sensor on it that tells the computer the speed of
crank rotation. Some sensors work by producing a voltage signal as a sensor
ring tooth passes by a magnetic coil. The sensor ring is mounted on the
crank shaft and rotates at the same speed as the crank. The change in the
magnetic flux as one of the ring's teeth pass the sensor causes a change in
the signal voltage produced by the sensor coil. The sensor can also tell
which cylinder is at TDC at any given time by the number of teen on the ring
and the spacing between the teeth. Usually one tooth is wider or left out to
indicate the number one cylinder TDC position. When a cylinder fires, the
sudden increase in rotational speed of the crank shaft is sensed by the CPS
and the computer knows which cylinder just fired. If there is a problem with
that cylinder that causes its increase in crank speed to be less than the
other cylinders, then that is interpreted as a misfire. Yes, we think the
crank shaft is turning at a steady speed, but in actuality it is speeding up
and slowing down for each cylinder event. As a matter of fact this is why
there is a need for a crank shaft harmonic balancer to dampen those harmonic
vibrations in crank speed that occur with each power stroke.
Hope this helps,
--
Kevin Mouton
Automotive Technology Educator
"If women don't find you handsome,
they should at least find you handy."
Red Green
Shep
nice job. Shep
"Kevin" <kmou...@cox.net> wrote in message
news:XFVqg.64484$9c6.11112@dukeread11...
----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
Jeff Wisnia
Very understandable now, thanks.
I had thought of sensing the uneven acceleration of the crankshaft but
was stuck thinking that the inertia of everything would make those too
small to reliably measure.
Seems I was wrong, 'eh....(And that wouldn't be the first time for me.)
Kevin
"Shep" <djsl...@capital.net> wrote in message
news:1152134...@sp6iad.superfeed.net...
> Geez Kevin, that is what I said, I'd have made 20 typing errors by then,
> nice job. Shep
That's what spell check is for.<grin> Besides, I'm sick and confined to the
house. Nothing better to do. If it weren't for the pain pills making me
goofy I''d try to post more.
Kevin
TeGGeR®
news:XFVqg.64484$9c6.11112@dukeread11:
>
> The crank shaft has a sensor on it that tells the computer the speed
> of crank rotation.
<snip>
An interesting story related by a Toyota MDT tech a few years ago in
another group:
It seems a new Echo was coming into the dealership repeatedly for an MIL
illumination with a misfire code. The MIL would come on first thing in the
morning or last thing at night, but the rest of the day it was fine. The
dealer could neither figure it out nor duplicate the problem. After much
back-and-forth and many days of diagnostics, somebody rode home with the
customer to witness the problem as it happened for him. Sure enough, as the
customer drove into his driveway, the MIL came on with the usual code.
The problem turned out to be the customer's very bumpy cobblestone
driveway. It jounced the car and crankshaft around sufficiently to fool the
computer into thinking there was a misfire.
--
TeGGeR®
Shep
diagnosis, of course if there is no misfire than at least the car is
running ok while we are all trying to figure it out!
"TeGGeR®" <teg...@tegger.c0m> wrote in message
news:Xns97F7E170...@207.14.116.130...
----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
C. E. White
"TeGGeRŽ" <teg...@tegger.c0m> wrote in message
news:Xns97F7E170...@207.14.116.130...
> An interesting story related by a Toyota MDT tech a few years ago in
> another group:
> It seems a new Echo was coming into the dealership repeatedly for an MIL
> illumination with a misfire code. The MIL would come on first thing in
the
> morning or last thing at night, but the rest of the day it was fine. The
> dealer could neither figure it out nor duplicate the problem. After much
> back-and-forth and many days of diagnostics, somebody rode home with the
> customer to witness the problem as it happened for him. Sure enough, as
the
> customer drove into his driveway, the MIL came on with the usual code.
>
> The problem turned out to be the customer's very bumpy cobblestone
> driveway. It jounced the car and crankshaft around sufficiently to fool
the
> computer into thinking there was a misfire.
My SO has a Chrysler mini-van with the horrid Mitsubishi 3.0L V-6. A couple
of years back, we had big problems with misfire codes (cylinder specific and
random). While attempting to fix the problem I discovered that Chrysler had
issued a TSB for that engine that warned that the random misfire code could
be set by a bad serpentine belt tensioner. Apparently the vibrations set up
as the belt slipped and caught could affect the engine speed enough to be
interpreted as a random misfire. Unfortunately that was not her problem (or
at least her only problem, she did actually have a bad serpentine belt
tensioner). In her case the problem was a burned exhaust valve. The engine
seemed to run fine and the difference in compression between cylinders was
marginally OK (low within 75% of the high). If not for the local emissions
testing program (you can't pass with a check engine light on), we would not
have had it fixed. We had a local independent garage do a valve job and that
resolved the problem (Van is well over 200K now). The Chrysler dealers only
solution to the problem was a new engine. They said they didn't do valve
jobs.
Ed
Don Stauffer
now merely notice a sudden, temporary change in rpm immediately after
each cylinder fires. The timing on current engines is accurate enough
to note that.
Other ways that are/have-been looked at include measuring the voltage
across the spark plug during ignition in each cylinder, having a rapid
response exhaust gas sensor at each exhaust port, having a pressure
sensor in each cylinder, and having an optical (via fiber optic) sensor
look for the 'fire' in each cylinder.
The spark plug measurement is not too expensive. The others all require
a new, special sensor at each cylinder, so to my knowledge none other
than the high-bandwidth monitoring of rpm are being used.
Actually, rather than rpm, maybe I should say "instantaneous crankshaft
speed", since the the angular velocity of the crank is monitored in much
less time than a full revolution. The ones I was aware of before we
dropped the project required more than one timing gadget on the crank,
but I understand there are engines out there that have such timing
systems. I think the misfire sensing is part of OBDII. The company I
was at dropped our OBDII project before the effective date of OBDII, and
then I retired, so I am not up on really current stuff.
Don Stauffer
>
>
> An interesting story related by a Toyota MDT tech a few years ago in
> another group:
> It seems a new Echo was coming into the dealership repeatedly for an MIL
> illumination with a misfire code. The MIL would come on first thing in the
> morning or last thing at night, but the rest of the day it was fine. The
> dealer could neither figure it out nor duplicate the problem. After much
> back-and-forth and many days of diagnostics, somebody rode home with the
> customer to witness the problem as it happened for him. Sure enough, as the
> customer drove into his driveway, the MIL came on with the usual code.
>
> The problem turned out to be the customer's very bumpy cobblestone
> driveway. It jounced the car and crankshaft around sufficiently to fool the
> computer into thinking there was a misfire.
>
>
Indeed, the folks developing those type of systems were aware that
exceedingly rough roads would cause such false alarms. Even on a
perfectly smooth road the crank angular velocity is not perfectly
uniform, and the roughness of the road additionally perturbs the
velocity. The key was to set a threshold in speed changes that would
accept bumpy roads but detect misfires. I am not surprised that that
compromise is not perfect.
Steve
> My curious mind wants to know...
>
> My daughter's 98 Olds (V6) failed state inspection because of a huge
> number of misfires recorded as having occurred in one cylinder, and the
> "check engine" light was on.
>
> I tried a new spark plug and lead and swapped coils, but the condition
> remained so I did a compression check on that cylinder and found it 40
> lbs lower than the ones on either side of it. Looks like I'm into
> pulling the head and hoping it's just a valve problem and not a piston
> ring/cylinder one, and that getting the compression back up to normal
> will correct that "misfire" indication.
>
> But I'm wondering, what mechanical or electrical phenomena gets
> interpreted by the computer as a "misfire" in a specific cylinder?
>
The first mechanism used by carmakers for detecting misfires was that
the computer could see an abnormal current shape in the coil primary
winding for that cylinder. That's an excellent indication of a misfire
*caused* by a bad plug or bad wire. It can sometimes detect a misfire
due to low compression or air dilution (vacuum leak, etc) but its not
very reliable because the change in cylinder pressure has a very subtle
effect on the spark current profile.
Later engine managment systems detect the slight deceleration of the
crankshaft after a misfire, which is much more reliable. Given that your
engine is a '98, thats late enough that its looking at crankshaft speed.
~^Johnny^~
>That's what spell check is for.<grin> Besides, I'm sick and confined to the
>house. Nothing better to do. If it weren't for the pain pills making me
>goofy I''d try to post more.
Two men he air oars sir, come vent the spell, check her.
That said:
I wonder if there's a way to get the tortional vibration out of my
mouse wheel? It's the first thing to always go, long before the
monitor or keyboard. ;->
--
-john
wide-open at throttle dot info
~^Johnny^~
<jwi...@conversent.net> wrote:
>I had thought of sensing the uneven acceleration of the crankshaft but
>was stuck thinking that the inertia of everything would make those too
>small to reliably measure.
The mass of the flywheel actually accentuates the problem.
Of course, it's worst with longer crankshafts.
Good writeup about vibration sensing issues here:
<http://www.prosig.com/signal-processing/TorVibAliasing.html>