Stumble at 3000 RPM 2-10 degrees throttle.

[magnusfeuer]

I give up.

For the longest time now, I've had cut outs/stumbling at light throttle at 2700-3200 RPM. Since I have a wideband lambda as a part of the instrumentation, I've tried running both rich and lean and nothing helps. The ignition for that area is set to 20 degrees BTDC.

It seems to come and go as I drive. Sometimes it is barely noticeable, while at other times the whole car is jerking. There is no correlation between engine temperature, ambient temperature, or other external factors and the miss.

Apart from this specific issue, the engine runs clean and strong over the entire map.

My guess is that it is an ignition issue, but I can for the life of me not figure out why it is happening just there and nowhere else.

Any help would be greatly appreciated.

/Magnus F.

[magnusfeuer]

Reading on blatchat, it seems like the TPS is a possible culprit.

I'll do some data logging to ensure that it is smooth.

/Magnus F.

[Clark]

Hi Magnus,

Do you remember when I had the engine control problems a few years ago? I tried a number of things in the intake and ignition systems: cleaning the mass air flow sensor, cleaning the throttle body, replacing the idle control valve, fuel pressure regulator, throttle position sensor, and pre-cat original (from Caterham) oxygen sensor on my OEM exhaust system (broken by hitting a speed bump) with one made by Bosch. None of these attempts at fixing the problem worked. Finally in a desperation move, I went to my Ford dealer and bought a Motorcraft oxygen sensor and put that in. Now my engine runs as smooth as can be. Through all of these trials, my OBD-II problem codes were normal. It never gave me a hint that the Bosch oxygen sensor was giving the ECU bad information.

I know that your engine control system is vastly different than the OEM system on my car, but your fix is probably going to be something similar. I think that Justin's drivability problem that occurred shortly after the start of the Buttonwillow Enduro two years ago was eventually traced to a connector. My guess is that you're going to find something like this.

One more thing. Before the modern era with fuel injection and ECUs, we had carburetors. Some, particularly the famous SU carburetors fitted to many English cars would develop a small air leak around the butterflies and throttle shaft. Webers would also develop varying size leaks in where the butterflies closed against the carburetor bodies. With these, the problems occurred intermittantly, usually at tip-in or very light throttle, kind of similar to what you described with your car. The fix for SU's was to replace the carburetor or have the throttle bodies sleeved. For webers, we'd get creative with little tiny files in an attempt to balance air flow between the pairs of throttles on each dual-throat carb.

I don't know if any of this will help, but some ideas to think about. Best of luck with your search!

[magnusfeuer]

Thanks Clark.

While I do have a O2 sensor, it is for monitoring only and does not affect how the ECU runs the engine.

On the other hand, throttle body leakage may be an issue. I'll check if there is any spindle play. I will also look at the coil-on-plugs to see if there is anything suspicious going on there.

Thanks for the tips!

/Magnus F.

[slomove]

The TPS sounds like a probable culprit and easy enough to measure (or just exchange). Not sure about the air leak, you would see that mostly in idle and very low throttle.
If that does not help you got to start monitoring all other parameters, ideally while driving.

I hope I am not mentioning the obvious:

- fuel pressure, if not on your tablet gadget, a long hose to a gage in the cockpit.
- ignition (timing light clamped on), watch for erratic rev count display
- double check the ignition and fuel maps for errand entries
- do you have a vacuum sensor hooked into the ECU? If so, might be bad or misconfigured.
- do all plugs look similar or is one cylinder very different?
- compression or leakdown test
- did somebody plug a potato into the exhaust?
- injectors dirty?

[magnusfeuer]

The idle is stable, low throttle is fine outside the problematic 2700-3200 rpm range.

1. Fuel pressure
   Rock solid at 58 psi
2. TIming.
   I haven't checked the timing manually. However, the only way that the timing would be off is if the crank sensor had died, which would have stopped the engine completely.
3. Ignition and fuel maps.
   Checked and double checked. 21 degree BTDC and running stoichiometric. (Tried both lean and rich, to no avail.)
4. Vacuum sensor
   Throttle bodies = no MAP. TPS-driven map.
5. Plug state.
   I haven't checked the plugs. That is next up.
6. Compression.
   I don't have a compression tester. The engine did, however, perform fine at the last dyno run while having the issue.
7. Potato culprit.
   No. The exhaust still doubles as a leaf blower.
8. Injectors dirty
   Possible. I will have them cleaned, just to be sure.

Thanks for the check list. I'll get back to you with a plug update.

/Magnus F.

[slomove]

Good luck.

I mentioned the timing light not so much to check the timing (I can see you aiming the light at the pulley sitting on the radiator while somebody else is driving at 60 mph). But the better timing lights have a plug lead clamp and rev counter display. Just a wild idea and you might see revs dropping out if the plugs don't get spark.

I would offer you my compression tester as well as a leakdown tester (you need a separate compressor for that) but you will probably spend more gas getting it than buying one.

Injectors are a real wild idea and very unlikely. But who knows.

Can you log the AFR while driving? Anything unusual in that range?

Edit: most engines of that size have some resonance going (depending on the trumpet length) that may screw up the fueling. E.g. my engine has a distinctive ridge in the fuel map around 2500-3000 rpm. I am somewhat skeptical though if that would be enough to cause a stumble. It is only 10-15% and mostly at full throttle.

[magnusfeuer]

The AFR is fine at that range. I have the lambda value on my digital display, and it is within 0.96-1.04 range depending on how I've fiddled with the map.

I am fairly sure that it is an ignition issue since the engine cuts out, and I immediately see a very lean condition for about a 10th of the second before the engine comes back on. That is in line with no spark where the O2 goes straight through the engine without anything being used to oxidize the gasoline.

I'll check the plugs and run some tests with my timing light.

/Magnus F.

[Doug Liedblad]

can't you do datalogging with the T6?

[magnusfeuer]

So, I undid the plugs and found the possible culprit - grit in the spark plug threads possibly messing up the electrical connection between the plug and the engine.

The grit, present in plug holes 2 and 3, came from an uncapped central hole in the valve cover where I believe the knock sensor (or head temp sensor) used to sit. This hole is right under the vents on the bonnet, thus providing an easy path for dirt to enter.


The question now, obviously, is how to get the grit out of the spark plug threads in the head. I'll vac out the holes and use a long q-tip with grease on it to pick up/bind what is not sucked up. The threads themselves may be a bit more tricky though.

I am also paranoid about grit having gotten into the engine when I removed the plugs. Maybe open one of the valves, attach a small hose to the compressor, feed it through the plug hole and blow out as much as possible.

I will buy a leak-down tester to check for damage once I am done.

>sigh<

/Magnus F.

[slomove]

http://www.harborfreight.com/cylinder-leak-down-tester-94190.html

Not bad for a Harborfreight kit.


The grit stuff is surely not good and should be removed if possible. If it is aluminum grit it may not be so bad.
But I doubt it would cause spark failure. After all it is also metal and would not isolate the plug.

[magnusfeuer]

I was checking out that leakdown tester, but it had received really crappy reviews. I'll borrow one from a friend.

The grit is sand that has been ground down to a fine powder as it was making its way down through the threads.

There was no oil on the spark plug (same color as those in cyl 1/2), so at least things are not completely shot.


/Magnus F.

[Doug Liedblad]

Magnus

Make a small tube that will if thru the spark plug hole and vacuum out the bore.

Spend some money on a borescope you can get inside to look. They can be had for under $300 I think.

If in any doubt pull the head. It will be cheaper than a rebuild.

Having said that Mark Donahue used to blip the throttle to suck rocks thru the engine that were caught in his throttle plates, he didn't need the engine to last more a few hours so I am not sure that's the safe way.

I always check for dirt before I pull the plugs. I doubt that it was the cause of problem as it probably got on the threads as you were pulling them out.

[slomove]

I was checking out that leakdown tester, but it had received really crappy reviews....../Magnus F.

I noticed that but I suspect these folks just did know know what to do with it or how it works. I have it and while not a masterpiece of aerospace engineering it works just fine. It is only a pressure regulator, 2 gages, and a small orifice.
I am sure there are better ones but they do the same thing.

[magnusfeuer]

I'm thinking of filling the cylinder with isopropanol, stir it thoroughly with compressed air, and then immediately suck it out with a vacuum from my oil change suction kit.

Oil change will, of course, follow.

/Magnus F.