#5 not firing right

[Walt Fricke]

I have a 2.8 LS race motor with twin plug and Electromotive TEC3 fuel injection and ignition actuation. Clewett offset crank toothed wheel pickup and his pulley and wheel. I bought the motor running, replaced the 40mm throttle bodies with 46s and appropriate manifolds, and with help from a guy who knows this stuff worked out some irritating bugs and got it running great. After I built a replacement SS 2.8, I pulled the LS down to refresh what needed refreshing. Turns out many valves were seating sort of off center, so other than replacing bearings which really didn't need replacing despite 125 hours of racing, most of them shifting at 8,200 rpm, valves were about the only change.

The SS had the intake stuff bolted on with no changes. Seemed to run OK until it pulled a piston apart on the chassis dyno. So back to the rebuilt LS. To make the swap easier, I undid the 12 TB to manifold bolts, and left all the intake stuff hanging. Shorter profile motor for R&R, no spilled fuel, much easier deal. And put the LS in and bolted everything back up.

I now know I should, when doing something like this, go through the usual checking and adjusting - do the throttles on each side open at the same time. Check and rebalance the idle air flow as needed. Check and redo the programming for sensor values. After working my way through issues related to those and other not surprising things, I am down to just one mystery.

The #5 cylinder isn't firing as well as the others. The exhausts are much cooler for #5 than its neighbors. #5 and #2 have EGT probes, and #2 rises appropriately but #5 does not, though once or twice it did rise as it should compared with the other, but then got tired and dropped back down. The plugs are a bit wet when pulled, and it seems to be gasoline, though I'm not getting black smoke.

I have pulled both plugs, reconnected to the plug wires, and both spark. They sparked the same when I swapped in a different plug wire (Magnecores). The plugs are the fancy Bosch with the small pin electrode and a tapered side pin rather than the more common overlapping ground lug. I bent the center pin in a bit to close the gap on these very well used plugs (as are all the others on this motor), but no change in behavior.

I swapped the #4 and #5 injectors. Same issue, so it oughtn't to be a bad injector. On dry running (cranking with no fuel pump) both #4 and #5 injectors click the same, so I don't think I have bad wiring for the #5.

The problem is clear at idle-2,000 rpm. It seems to clear up when I take the motor up higher.

A quick CR check indicated 130 psi at about 6,000' altitude, which should be good enough for consistent firing. Haven't gotten around to leakdowns or CR checking of everything, but not sure that's going to tell me much. I can measure plug wire resistances with neighbors, to see if enough for a spark out in the open isn't enough for idling. I should have swapped plugs between cylinders. I haven't done the spray a combustible gas around the intake tracts exterior to see if I've got an air leak somewhere (I didn't leave out a whole gasket), but don't such leaks usually lead to faster idling and a leaner burn? Beyond these seemingly unlikely possible causes, I'm out of ideas.

TEC3 allows individual cylinder trims, but without fire I don't see how I could use that to help. The TEC3 gives more fuel for the 1st second (start enrichment), and also for the first 20 seconds (startup enrichment). But if every other hole is getting enough fuel and spark for idling beyond 20 seconds, why wouldn't this one fire also?

Suggestions or speculations about possible causes and cures gratefully accepted.

[Green993]

Ignition wire?

[HawgRyder]

Air leak?
It might upset the fuel/air ratio.
I would look for a gasket leak...or worst case...a crack in the runner?
Next up...perhaps a bad dist cap if there is one...or is it direct separate coil to plug?
Thats about all I can think of at this time.
Bob

[Lapkritis]

Depends on the installer. Tec is normally wasted spark with three re-stickered GM two post coils. I would pick up quantity 6 of 12v LEDs to test all injector outputs first. You'll need a helper to crank while you observe the LED fires in place of the injector. If still misfire and injector channels test out then I would check the matched cylinder on the wasted spark coil to ensure it is firing. If it is, swap the plug wire with #5 and try again.

If spark and fuel check out then I suspect a valve may not be behaving as intended.

[Walt Fricke]

I need to check resistance on the wires just to be sure. And for the lower plug, I pulled the wire from the coil, attached another one, and tested the plug. It sparked, just as it did with its regular wire and the plug out.

This is twin plug run from two banks of coilpacks. It is waste spark, with three coils, each with two plug wire terminals, on each bank. One bank serves the uppers, and one the lowers. So if a coil is bad, it should be bad on both sides of the motor. And for both plugs in one hole to look about the same, you'd pretty much have to have the same problem on the hole on the other side which fires 360 degrees out.

Could an air leak so upset the A/F ratio that the gasoline injected into the TB just above the port wouldn't even ignite? I need to do the standard tests, though cracked manifolds (stout PMOs) or TBs (TWMs, or is it TMW?, now Borla) are unlikely candidates.

How might one check a spark to see if it was powerful enough, assuming that one could have a spark with the plug out in the air, but under idle compression pressures it wouldn't spark? I can see some new plugs in my future to see if maybe these gaps (not easily adjustable) have grown so wide that with higher compression, those extra air molecules might be just enough extra insulation so the spark might not jump? But do cylinder pressures rise as RPMs go up, at least until the torque peak?

Keep those thoughts coming.

[Lapkritis]

Also double check your firing order, possible a couple are crossed. I've seen that before... normally spark error but could be injector too if you're set for sequential injector control.

[Walt Fricke]

When racing my VW beetle convertible in the Carrera Panamericana in 1991, I passed a for real 4 cam 550 Spyder on one leg. Made me feel like a really hot shoe. Two days later the German co-piloto/mechanic of that car asked me what his firing order was. When I explained I had zero experience on any twin plug car (at that time), much less an exotic one like his, he said it was the same as my VWs. Oh, I knew that by heart. And yes, they had two leads going to the wrong plugs. Walked right by me next time we met on the road.

But my wires are of lengths and loomed so attaching to the wrong plugs isn't really possible, and they never come off the coil. And no other cylinder exhibits any signs of distress.

The word on the race motor street is that sequential firing isn't useful, because the injector can't get enough in during its one carefully timed squirt. It seems that batch firing, sort of like waste spark I suppose, is simpler. And it doesn't require a cam sensor. Might well be different for street motors. The TEC won't sequential fire without a cam signal anyway, so it can't be that.

A friend who is using Megasquirt wondered if I was having a minimum on time issue at idle, but that seems improbable for just one injector.

I've been wondering about a valve issue. So I'll do a leakdown in addition to the compression test I did. Valves all checked out good after I reground them and seats, but maybe something is going on, and somehow affects things more at idle speeds than higher RPMs.

[Lapkritis]

I've seen injectors jam shut (killed a piston like that myself last year-Siemens Deka 60lbs) and it was only noticeable at idle. The engine would fool me with acoustics above 2500rpm. I went through the exercise of replacing the piston and test driving before I found it by reading the plugs compared to neighbor plugs. I think this may be quite likely here.

[Walt Fricke]

Well, I swapped the injector on the bad hole for its neighbor on #4. 4 still good, 5 still bad. Acoustics may have fooled me into thinking my issue clears up with RPM, because the EGT doesn't really support that. Test conditions weren't conducive to much higher RPM running, though.

No fuel killed the piston? Wonder how it did that, though obviously it did. There is lean burn, but no fuel ought to be no burn?

[Lapkritis]

Originally Posted by Walt Fricke

Well, I swapped the injector on the bad hole for its neighbor on #4. 4 still good, 5 still bad. Acoustics may have fooled me into thinking my issue clears up with RPM, because the EGT doesn't really support that. Test conditions weren't conducive to much higher RPM running, though.

No fuel killed the piston? Wonder how it did that, though obviously it did. There is lean burn, but no fuel ought to be no burn?

Walt, I would move forward with the LED test at the injector harness. Might be wiring between the connector and ecu for that one cyl. Not sure if you've done this test before but be sure not to use a conventional bulb as it may overload the ecu injector driver and burn it out. LED only, RadioShack will have them in 12v.

The failure happened at a really bad time; ~25lbs of manifold boost and high rpm. I felt it break up and lifted the throttle immediately and went to the clutch.

First found the spark plug broken, oil soaked and quite lucky it did not come apart and drop the bits into the cylinder to be pounded into the head. Only piston in the engine to have any damage...

[Walt Fricke]

Well, I can try the LED bit. I can make them up with a resistor (forget just what value, but can find it) and LED from my stashes. Or just purchase. Since it clicked, same as its neighbor, I'm not holding out much hope for something as simple as wiring. My TPS issues were a crimp connector which had somehow pulled apart - bad crimp and I guess tension while the assembly was hanging from some bungees as car sat in trailer as it moved, etc. Calls for opening up the carefully zip tied loom, but what the heck.

Is the damage to rings and lands on that cylinder due to detonation? Looks like the kind of thing which led my wife to veto my 2.16L turbo motor project - didn't want to learn by breaking. But that can follow you anywhere.

[HawgRyder]

Walt...before you go for resistor/LED combos...be advised there are such things as 14V LEDs.
There are also bi-directional LEDs....meaning they light up one color when (+) is applied to a lead...and a different color when reversed.
It might make things quicker and easier for you to have some of these around.
As long as you rely on the LED just lighting up (dont care about color) they work great.
Bob

[Walt Fricke]

Sunday I got in the trailer to check things out a bit more.

Leakdown on #5 (problem hole) ~3.75%, #4 4%. All leakage was through rings, which is normal. At 100 psi, valves should not leak at all. Combined with compression check of 130 psi, I don't think I have a low speed valve sealing issue.

Intake valve had proper lash. Didn't remove exhaust covers, but next time in there will see if I can get a brass drift down to edge of exhaust rocker. It should wiggle, though obviously it isn't being held open or it would leak and I'd have heard it in exhaust, because I was listening with my stethoscope's rubber tube. I'll have to take the car out of the trailer to jack it up enough to get at exhaust valve covers.

Upper plug coil wires for #4 and #5 had about same resistance - ~7.5Kohms. Coil packs really can't be at fault, because if #5 has a weak or no spark, #2 should also as it runs off the same coil. Next I'll check the #5 exhaust plug wire after bringing a longer test lead, as the lowers are sort of built-in where they pass under the frame rail.

Plugs look OK, and at least center electrodes are patent, and no VOM level leakage between center and shell. To address possible spark problems, I found I could purchase the Bosch plugs which have a pointy side electrode which is no higher than the tip of the ground electrode. This is what is in there now. Expensive plugs, but needed for clearance per the previous owner, who built the motor, so I'll replace those in this cylinder just in case.

Used borescope. Nothing remarkable down the throttle bodies to the valve stems. I felt around the exhausts to see if perchance I'd let a gasket get crosswise and block things, but all seemed OK. But the intake pocket of the #5



looked cleaner, less carboned, maybe more fuel washed? than the #4



Both pistons had dark crowns, as compared to the valve pockets. Not sure what this means, other than possibly a spark problem.

Someone suggested I check valve springs. While my experience is that these motors will run up to 7,600 rpm happily with broken inners (I ran most of an hour endure that way), and they all seemed their normal hard to compress selves when I reassembled the heads after doing the valves and seats, I can't afford to overlook anything at this point of bafflement. Maybe something broke during its very short running time post rebuild, while I was trying to sort it out.

[Lapkritis]

I might be tempted to swap the entire coil with the coil for another pair and see if the problem follows to the new pair. Could be weak enough that it won't fire both and #2 is getting just enough... I had this experience on a gm style two pin coil from Jegs out of the box. They replaced free of charge.

Inline tester takes guess work out:

http://www.jegs.com/i/MSD+Ignition/121/3503K/10002/-1

[WERK I]

Walt,
Is it possible 5A and 5B leads are reversed going to the wrong DFU, respectively?

Edit: The more I think about it, the idea doesn't seem to hold water, but I would go over the DFU to plug wiring just confirm everything is right. Sure sounds like an ignition issue.

[WERK I]

Here's an illustration for a Porsche twin-plug, wasted spark, Electromotive TEC-3.

[Walt Fricke]

It does sound like an ignition issue. Maybe if a coil is on the ragged edge, small differences in impedance or resistance on one plug run of a pair might lead to one getting just enough, and one not?

I've got a new coil. Carried it for years as a spare just in case. I can swap it for the 2/5 uppers, and see what's what. And if no change, for the 2/5 lowers. Since by chance 2/5 have the EGTs, any change will jump right out.

I can also find and reinstall the ground wires I used to have attached to one of the mounting screws for each coil pack. Since the packs are bolted to brackets I fabricated and welded to part of the roll cage, I've never thought these were all that useful, and when I switched from HPV1 ignition only to the TEC3, I didn't reinstall - with no apparent bad effects.

For a while, my friend and I wondered if we might be getting a minimum on time complication, but I've kind of discounted that.

[Walt Fricke]

I'm getting pretty stumped.

#5 compression good, leakdown good, intake valve spring (levered against it and #4 for comparison, wasn't budging with pushing with fingers)and lash good, exhaust lash good, and spring seems OK (checked indirectly).

Sprayed brake cleaner around outside of #5 TB. No change in engine. Sprayed some down throat of #5 TB - engine picked up for a while.

Swapped spare ignition module into upper, then lower 2/5 slots. No change. Swapped right and left side wires at module. No change. Put in new wires for #5 upper and lower. No change. Measured resistance on old wires and new ones - same at about 7.5K ohms. New plugs right side upper and lower, no change. Plugs out, no fuel pressure, connected the upper #5 and grounded it, watched the upper #5 plug spark away.

All injectors clicking away with engine running. 12V LED attached to #5 injector plug, blinked away. Plugs out, throttles open, crank engine with one injector attached. #5 sprayed just like the three or four others I checked. Checked continuity of injector wiring to the positive wiring point - just fine. And pulled the white plug and checked from ECU to #2 and 5 injectors. About 0.3 ohms resistance, same with one or two others.

But the brand new plugs in #5 show no signs of ignition. EGT for #5 doesn't rise. IR gun heat readings show header for #5 significantly cooler (more like base engine temp) than the five cylinders which are firing.

Only oddity I noted was that #5 on the Unisyn was pretty close to all the rest (adjusted mid-range with idle air screws on TBs). But the idle air screw made no difference at all. Though without fire, perhaps that is what one would expect?

About all I can think to do is reduce the ignition advance at idle. Starting (600) is 14 degrees, 1000 is 16, 1500 is 18, and 2000 is 26. I made a map with starting the same (it starts right up, and I'm now using a remote starter so I can start and futz from behind), 1000 at 12, 1500 at 14, 2000 at 20. This on the off chance that for some reason #5 (which has the best VE because it is shortest) is having trouble dealing with the big cam (even though #2 isn't).

I'm open to suggestions. Next beyond futzing with the programming is to check cam timing. If it is a bit off on the right side, maybe somehow that affects #5 more than #s 4 and 6? 4 does some backfiring, and the plug is blacker than #6, which has the best looking plug, considering all my starting and testing has put almost an hour on the engine meter.

Open to more suggestions.

[rsscotty]

Walt,

I had an engine brought to us to dyno. I go through the normal hookup, of which one thing we do is crank the engine over without spark plugs until the oil pressure comes up. It didn't dawn on me at the time, but I remember thinking .....gee, this engine seems like it cranks slower than what I am used to hearing. So everything is hooked up and it starts up fairly normal, but sounds off key, like maybe down one cylinder, and a little weird popping now and then. So I am leaning into the engine and the left side of the muffler was cool, like I could put my hand on it. This engine had stock heat exchangers and stock early dual in single out muffler, so I could not isolate the cold cylinder easily. All of the spark plugs looked weird, with only one wet one and the others on that bank were black but not wet. My gut instinct said the cam was out 180 degrees, and after a quick check, sure enough it was. The engine builder had done these engines before, but somehow it got put together 180 out. The cam was reset and everything was good..... and the engine cranked faster with plugs in, now with good cam timing , than it did before with the plugs out.

After reading through your posts on this, I would have to think this is the only thing left. You have done the proper diagnosing, but the cam being out 180 is something you don't commonly see...first time for me in many years. So this one really throws everyone a curve ball.

You will get it sorted..be strong.

[Lapkritis]

Originally Posted by Walt Fricke

Sprayed some down throat of #5 TB - engine picked up for a while.

This tells me fuel related because brake clean is super combustible. How old is the fuel in the tank?