Ignition cut-out

This evening (yes - Christmas eve) my daughter's '93 Accord LX (225K miles)
suffered a short bout of engine failure. She described it as surging a
couple times when she tried to accelerate, then the engine died when she
braked. I quizzed her about the tach and she was clear it was jumping rather
than swinging with the engine. It didn't start right away, but would sputter
occasionally and eventually lit off as though flooded (which it probably was
by then).

The Main Relay and ignitor are each about a year old (but both are NAPA
aftermarket, since the car insists on breaking down on Sundays and
holidays). The timing belt is 4 years, 60K miles old. My son and I replaced
the radiator about 2 weeks ago, so I am wondering if we stirred up evil
spirits in the process of that. He bled the system carefully, but a leak in
the lower radiator hose attachment may have let air in. I haven't done a
visual inspection yet, but will in the morning.

What comes to mind? Ignition switch? Some particular underhood wiring?
Distributor? Something I'm forgetting? She is committed to driving to San
Diego in a week or so and I'd sure like to sort this out by then.

TIA

Mike

 

How old's the ignition coil?

I'd try to do a check of it early on. Dying when
braking--then being able to restart it after a cooldown
period, at least until the coil fails completely--is a
symptom.

I know you have a lot of experience with Hondas. This is
just where I'd start, based on the car's age and symptoms.

 

I'd forgotten about that! The coil is the original. I might be wise to
shotgun it out, whether or not I find something else wrong. I should even
have time to get an OEM part before she leaves. At least that is something I
don't have to worry about disturbing unnecessarily - I hate doing invasive
work so soon before a big trip. Thanks, Elle!

Mike

 

I wonder if the problem is electrical in nature, because the RPMs were
jumping around wildly.

Since the tach and engine timing both get their signal from the crank
sensor, I'd also look in that area.

Remco

 

I thought I read on here recently that an erratic tach that precedes a
cut-out is a sign of a bad igniter.

 

Not sure about any previous post on this issue, but the ignitor seems
to just be a larger power transistor (probably a darlington, but that's
besides a point).

The way this transistor is configured, it simply acts as a switch that
that closes/opens a contact when its input changes states (to simplify
it all, imagine a relay with a coil and contact - that is not quite
what a transistor is, but behaves quite similarly the way an ignitor is
hooked up).
When a level is applied to its input, it switches the primary side (the
12V side, basically) of the coil to ground and does this at a very high
rate of speed. The secondary of the coil is what generates a high
voltage spark.
The output of the ignitor does not seem to attach to anything else so
iti it breaks you won't have spark.

I don't think the tach is getting its signal from the ignitor, so don't
see how one can tell by the tach that the ignitor is bad. If you
somehow can detect this on the tach, it must be some sort of weird
artifact.

Remco

 

Actually, I believe the tach does get its signal from a separate lead on the
igniter. The current symptoms do sound like igniter trouble, but the igniter
(and tune-up parts) are a little less than a year old... from March, IIRC. I
might be wise to re-heatsink the igniter, though.

What makes me think it isn't distributor sensor trouble is that there is no
"check engine" light, suggesting the problem is past the ECU... i.e. igniter
or coil, or high tension side (like the rotor or cap.)

Mike

 

The electrical diagram for the ignition system suggests you
are remembering correctly. See the top schematic at
http://home.earthlink.net/~honda.lioness/id5.html . Tegger's
version appears to confirm it:
http://www.tegger.com/hondafaq/igniter-operation/index.html

Plus ISTM igniters don't exactly seem to die slowly, at
least not the way coils often do. Igniters seem more like an
all-or-nothing deal, with symptoms only for the very
observant, at best.

 

I have a Bosch coil ordered through alleurasianautoparts.com, should be here
by Wednesday. Since there really is no way to determine whether the coil is
failing intermittently and 225K miles is a lot of service to expect from
one, it seemed like the prudent thing to do. I can check the ignition switch
by seeing if the voltage across it fluctuates from one time to the next, but
the coil keeps its secrets.

In my experience, the more power a device handles the more likely it is to
have a limited life. An ignition coil has to transform hefty currents into
hefty high voltage jolts, so I don't expect it to last forever. Thanks again
for the suggestion.

Mike

 

You're absolutely right - I stand corrected and didn't mean to mislead
anyone.
It has been a while since I looked at the ignitor and didn't have my
manual handy, but that's a bad defense <blushing>

This is why we need those pros, huh

Remco

 

I'm with Elle on the ignitor's failure mode. One would imagine that
these things just fail very quickly (unless it is thermal in nature,
oif course)

Could it be that the power to the ignitor is intermittent - A bad
contact in that path somewhere?
Imagine a bad intermittent contact on the 12V (or ground) to the
ignitor while the car is running. It would certainly translate into bad
spark (too narrow) and getting noise on the tach lead..

Remco

 

I've looked at the schematic I cited and Tegger's igniter
site closely at least half a dozen times each this past
year, and I forgot about the tach output until Michael P.
mentioned and I re-checked. That the Tach signal comes from
the igniter doesn't exactly pass the common sense test. At
least not at first blush, afaic.

I trust you see Tegger's site confirms the rest of your
conjecture, about the igniter being mostly a Darlington pair
transistor, yada.

Yes, plus repair manual hounds.

 

Aw, shuks - now you're just trying to make me feel better

Yeah, that is kinda weird how they have the tach input coming from the
ignitor. One would imagine they could have just as easily taken the
signal going to the ignitor (possibly current boost it a little) and
feed it into the tach directly. Less wire and things that can go wrong,
imo...

Maybe they did that as a trouble shooting measure or we can use at as
such: This morning I cranked my wife's integra and noticed that while
cranking the tach moves very slightly. Maybe if a honda cranks and the
tach does not move it is an indication that the ignitor is bad..? That
would be an interesting thought.

Wonder how Michael is making out with this thing.

Remco

 

You and anyone else who posts or lurks here. This was not an
easy catch, IMO. But as importantly, a group like this will
thrive only with open-mindedness and polite honesty. (Not
that I excel at these, but I know I'm supposed to try...

I thought maybe this might be happening anyway. The igniter
gets input from the ECU, of course, after all.

On further thought, the RPM signal really had to be
someplace in the vicinity of the camshaft or crankshaft. And
someplace protected, like the distributor housing. I am
thinking many other cars have their tach input coming from
somewhere within the distributor as well.

bad..?

Tegger's site certainly puts emphasis on this notion. See
http://www.tegger.com/hondafaq/igniter-operation/badigniter.
html

I see on further examination that Tegger's drawing (at
another of his sites on the igniter) has the Tach signal
coming out of the igniter's control chip, for one thing.

<Whispering: Why's one of our regulars buying a non-OEM
coil?>

 

Lol -- yeah, I was kinda wondering the same thing. That aforementioned
mentioned politeness thing had me keeping it to myself.
Sorry for getting on your case, Michael.

 

Thanks for asking! Kinda in a holding pattern. My daughter got back with it
just now, so I can check out the ignition switch and the engine ground. No
failures since Friday, though. Whether that's good news or bad news is a
matter of perspective :-}

Mike

 

==============================================================================
TOPIC: Ignition cut-out
http://groups.google.com/group/alt.autos.honda/browse_thread/thread/e8a5fd8cba927459
==============================================================================

I thought coils last forever. I never had to replace a
coil, rotor, cup, or wire on any of my cars. I have an
Accord that soon will be 22. But on this list it is
very frequently recommended to replace, even on young
cars, all kind of ignition parts. To me this sounds
like "it won't do any harm, so lets replace it."

I have a question about the original message:

Surging - meaning the rpm increased (suddenly,
audibly?) more than expected, right? With the cars I am
familiar this would not happen when any of the ignition
parts fail, to the contrary, they would slow down.

In order for the engine to rev up, without any
corresponding increase of the speed, there must be a
disconnect between engine and road, for example
slipping of the clutch or AT - which most likely
happens when they cannot transmit the increased power
during acceleration.





..

 

I doubt a starter will crank the motor fast enough to make any significant
reading on the tach.
What's the smallest RPM indication on the tach,100 RPM?
No starter turns the motor that fast.

 

I wonder if using a non-integral distributor cap/coil like the separate MSD
racing coil and plain distributor cap would be a more durable system?

MSD is supposed to have a good reputation.

I never liked the idea of an integral cap and coil.

 

Well, FWIW on my wife's Integra I see the needle bump around zero when
it starts. It would be interesting to see if it still does that with a
bad ignitor..

Remco