: Ignition drop-outs



Will's Fiero
10-10-11, 11:00 AM
I am having problems with "ignition drops" in my Northstar.

The engine is a "Frankenstein" with '93 block and '95 heads/intake. The block is basically fully built (11.5 compression CP pistons, gapless rings, ceramic/moly coatings, Eagle rods, balanced crank, MLS gaskets, etc.). Stock throttle, intake mani, heads, cams, sprockets, and exhaust manifolds. I'm using NGK plugs one heat range colder than stock.

I'm using an OBDI computer. The body is a Pontiac Fiero and it has a manual transmission.

The problem is this:
When the engine is below 2000 RPM, the ignition "drops out" momentarily. The tach flickers when it does this. When it happens at idle, the engine usually stalls. When it happens under load the car bucks hard. I tried a couple of different coil packs I had on the shelf. The first was the same in terms of the ignition drops, but had a bad coil and dead miss. The second was much better in terms of ignition drops and did not have a miss. With the newest coil pack, the engine will pull cleanly from 1000-2000 RPM in fifth gear about half the time and stays lit about two thirds of the time when the ignition drops at idle.

An "Ignition drop" is a split second interruption of ignition. When it happens, the tach flickers. This indicates that the tach is loosing signal. This means that there is a problem with the crank sensors or coil pack. Either the coil pack is dropping offline momentarily, or its losing synch with the crank sensors (BOTH of them?) or the crank sensor signal sags at low RPM. The fact that the phenomenon was present with multiple coil packs makes me think that either it isn't the coil pack at all or it's a very common failure mode with the coil packs. I suspect that since I haven't heard about it before, that the coil packs don't fail so much that all three of mine would be bad. I don't think it's an interruption of ignition power because I would expect that to show up at all RPM, and the phenomenon is completely absent above 2000 RPM. In fact it runs like a raped ape, pulls like a freight train and makes better than 45 horsepower more than the stock shortblock did...

Ideas? Is this a common failure mode of the coil pack? Crank sensors? PCM?
Thanks!

89falcon
10-10-11, 09:32 PM
Will....please put a link to your engine build forum string in your sig line.....I'm pretty sure I've read it and it's pretty f'n impressive. It's nice reading about one of the engines in something other than a 2 ton tank.

No ideas on your problem other than to say it isn't likely your coil pack.....the fact that the tach "dies" when it happens makes it sound like an electrical issue.....

Something to think about.....the engine is "smoother" at high RPMs.......due to the cams, you get a little more "shake" at idle....and under load, it flexes in one direction.....is it possible that you are getting an intermittent ground when the engine is "shaking" at lower rpms? Do you have a wiring harness ground near the coil pack? Is it tight? Start it up, and while at idle, start yanking on and flexing the wiring harness in several places....

Will's Fiero
10-10-11, 09:51 PM
The coil pack is wired correctly (worked before I put the new engine in).
Mounts are stiff urethane and engine motion is pretty minimal.
The coil pack is mounted to the cam cover like stock and the baseplate is grounded with the stock 8 ga (or whatever it is... really big) ground jumper. That goes to a bolt on the cylinder head that it shares with the negative battery cable. I can check how tight that bolt is, I guess...

The dropouts make it buck pretty hard at low RPM/high load. It can smooth right out and keep pulling, so I don't think it's a wiring issue. Besides, I have the harness routed such that it doesn't get pulled when the engine moves what little it does move.

No harm in trying though...

The tach dropping just means that the tach's losing signal. That could be a wiring problem OR a component malfunction.

I don't *think* I've posted a build thread on this forum... Cams are stock L37, so it doesn't have any lope at idle. Puts about 300 to the wheels with my built short block, though.

89falcon
10-10-11, 11:09 PM
I believe you have a thread on a fiero forum....and yes, some of us read those to see what you guys are doing to the engines!

Ranger
10-10-11, 11:36 PM
"Coil Pack"? Sounds like you are talking about the ICM. I'm not sure if it will help any, but there is a lengthy ICM sticky at the top of the page that may give you some insight.

CKP sensors could be a possibility.

Will's Fiero
10-11-11, 09:34 AM
Yes, Coilpack is colloquial for ICM. I'll take a look at the sticky.

Build thread is here: http://www.fiero.nl/forum/Forum3/HTML/000121.html
It's long and boring as I learn the ins/outs of building a Northstar and have a couple of false starts. Pictures are scattered throughout and dyno results are on page 17.

Edit: The sticky looks like a transcription of the shop manual... which I read just about cover to cover when first doing the swap.
Forgot to add that I also have Magnecor plug wires.

Will's Fiero
10-11-11, 02:42 PM
Maybe I'm incompetent, but I can't find how to create my own sig... I have found an option that allows me to display everyone else's or not.

89falcon
10-11-11, 11:12 PM
Yes, Coilpack is colloquial for ICM. I'll take a look at the sticky.

Build thread is here: http://www.fiero.nl/forum/Forum3/HTML/000121.html
It's long and boring as I learn the ins/outs of building a Northstar and have a couple of false starts. Pictures are scattered throughout and dyno results are on page 17.

Edit: The sticky looks like a transcription of the shop manual... which I read just about cover to cover when first doing the swap.
Forgot to add that I also have Magnecor plug wires.

I know this is COMPLETELY counter-intuitive, but I'd get the thing as close to "normal" (stock) and work from there.....ie, I'd grab an old set of stock wires, install them and see if you still have the problem...

Will's Fiero
10-26-11, 09:40 AM
I got a reply on another forum from a Cadillac dealership mechanic that he's seen a similar failure mode on the DIS coil packs. I should be able to scope the crank sensor and ICM<->PCM signals this weekend.

Pete1996
10-26-11, 01:43 PM
I'm an EE and since you've checked so many things I'll just mention
one that comes to mind. It is possible that the ECU is sensitive to
grounding, in the sense of how (physical order, shared grounds, etc.)
you physically wire the grounds. Think of current flow, which will cause
a drop in any ground wire making it not exactly ground anymore. Say
the ECU has a direct connection to the battery ground, and one of the
sensors shares a ground with say the coil packs. The peak current
charging the coils is probably high, causing a lift of the associated
grounds, if a sensor shares that ground it will have the noise super
imposed on the signal since the ECU ground is referenced to the
battery. A larger ground wire might not help since these noise
components are high frequency and the inductance of the wire comes
into play. Issues like this can be highly dependent on operating
conditions such as system voltage, alternator charge rate, or
any other thing that might also contribute to the ground shift.
A "star" grounding system is usually best in order to avoid cross
contamination of noise components. A partial star can also be
used where low noise sections share a star connection, and each
high current noise source has a dedicated line to the star point.

Look at how Cadillac wired the grounds and try to duplicate it as
much as possible if you think this might be an issue.

aeronca36606
10-26-11, 01:56 PM
I don't know how the Fiero Tach works but if it plugs into the coil pack i have seen a bad tach cause just this problem on an old Firebird.

Will's Fiero
10-26-11, 07:24 PM
Thanks! I'll take a look at that.
If I'm scoping the signals, I should be able to reference the scope to different ground locations.

The current to ground shouldn't fluctuate based on RPM, should it? Even though the frequency goes up with RPM, the DC stays the same...

The coil pack has 16 ga (IIRC) +12 supply and ground wires which share a 3 pin connector with the tach wire. The baseplate of the unit is bolted to the cam cover, which is isolated from the block. There's a large (8 ga maybe) ground jumper that goes from the baseplate to a ground point on the cylinder head which is shares with the battery ground, which is even larger. All the grounds have direct connections among them.



I'm an EE and since you've checked so many things I'll just mention
one that comes to mind. It is possible that the ECU is sensitive to
grounding, in the sense of how (physical order, shared grounds, etc.)
you physically wire the grounds. Think of current flow, which will cause
a drop in any ground wire making it not exactly ground anymore. Say
the ECU has a direct connection to the battery ground, and one of the
sensors shares a ground with say the coil packs. The peak current
charging the coils is probably high, causing a lift of the associated
grounds, if a sensor shares that ground it will have the noise super
imposed on the signal since the ECU ground is referenced to the
battery. A larger ground wire might not help since these noise
components are high frequency and the inductance of the wire comes
into play. Issues like this can be highly dependent on operating
conditions such as system voltage, alternator charge rate, or
any other thing that might also contribute to the ground shift.
A "star" grounding system is usually best in order to avoid cross
contamination of noise components. A partial star can also be
used where low noise sections share a star connection, and each
high current noise source has a dedicated line to the star point.

Look at how Cadillac wired the grounds and try to duplicate it as
much as possible if you think this might be an issue.

Will's Fiero
10-26-11, 07:26 PM
I don't know how the Fiero Tach works but if it plugs into the coil pack i have seen a bad tach cause just this problem on an old Firebird.

The Fiero tach is a conventional air core meter, BUT the output from the DIS is a digital 0-5V... It should be very well isolated from the other circuits in the ICM.

I think I've started the engine without the instrument panel connected and it still had the problem.

Pete1996
10-26-11, 11:06 PM
Yes as long as no other instrument is grounded to the car then you
can move the scope ground around as your reference point. You could
even ground it to the battery and check the grounds at the coil pack for
example with the scope on a fairly sensitive scale to observe the ground
shift.

The DC supply voltage is roughly the same, but the currents all vary
depending on RPM and many other factors. I hope this is not your problem
since it is difficult to debug and can be avoided with good electrical design
and noise margins but it is something to watch out for. Again, think of
current flow, anything that draws high current, even pulsed will shift a
given ground return.

Pete1996
10-28-11, 08:23 PM
I gave your build thread a quick read Will, nice work, very nice!
I'm impressed and curious about your high RPM gains.
I wonder if the higher valve spring pressure might have anything
to do with it?