350i - Misfire from 4.5k RPM onwards with WOT

[wombat172a]

As per the title, is this likely to be a bd to diagnose? Car is a 350i Flapper, and it won't rev past 5.5k at all. I took it out for a run today after getting it running again, and didn't realise I had a problem until the traffic died down and the road opened up. However, if I take it to a point where it's misfiring and then lift off from WOT it can maintain the RMP without mis-firing.

The last problem was Ignition Module, which stopped it from running for more than 20seconds. Although in the process I also replaced the coil and the Fuel Pressure Regulator.

Previously I've had a stuck hydraulic lifter, and I'd like it to not be that again.

Where's best to start?

Cheers,
Chris

[honestjohntoo]

I would say there are five main threads you can pursue.

1 failing ignition passing 5500 rpm - easy to check inside the cabin, with a small temporary circuit wired to the coil connections - probably unlikely as the engine is OK after lifting the throttle.

2 Full load enrichment circuits not firing at WOT - should be triggered by the throttle pot from about 3.7 to 4.3 output volts - easy to check throttle pot function

3 Same problem, could also be the ECU - test by substitution.

4 Cam worn - limiting fuel/air intake to one or more combustion chambers - easy to check without dismantling any part of the engine (except spark plug lead(s) ) just using a multimeter with accurate RPM feature.

5 Hydraulic lifter(s) sticking - again - cannot be ruled out due to prior occurrance.

Suggested test solutions

For #1 - read how to test for 'Intermittent Ignition Faults' (scroll to the titled section) here:

http://www.vintagemodelairplane.com/pages/Download...

For #2 check the Throt Pot according my Efi Ops Manual or the Efi component archive at:

www.vintagemodelairplane.com

For #3 easiest way is check by substitution of an identical spec ECU

For #4, an amazingly simple process, go to:

http://www.vintagemodelairplane.com/pages/Snippety...

For #5 - I'm told that an engine cleaning oil additive may be a solution, but not tried it myself.

[Wedg1e]

Last time I had a rev-limiting glitch like that it was the pickup coil in the distributor; went for months before it just died one day. In the meantime I'd stripped and rebuilt the injection system, convinced it was my home-brew collection of parts that was to blame, but no. With the pickup replaced it was business as usual.

[Danny Hoffman]

When I had this it turned out to be a dodgy rotor arm

[ElvisWedgely]

To me it sounds electrical. Either the distributor, cap, leads, rotor or coil. Like all diagnosis you have to use elimination. Just one question. Are your ignition leads tucked up neatly? It could be a stray spark from the leads (arking) best seen at night in the dark when the engine is running at high revs. (at high revs the spark is strong and can jump to other metal surfaces).

Tony. TCB.

[maston]

Check the throttle position sensor , should be about 0.33v throttle closed and about 4.3v throttle wide open , easy to check too , ignition on voltmeter on tps

[wombat172a]

Wow, thanks for all the replies guys.

Unfortunately I've not had a chance to do anything on it today, but hopefully with the lighter evenings I can start eliminating some bits and pieces this week.

[wombat172a]

ElvisWedgely said:

It could be a stray spark from the leads (arking) best seen at night in the dark when the engine is running at high revs. (at high revs the spark is strong and can jump to other metal surfaces).

Okay, just had a bash at this whilst it was convenient (and the neighbours are away). No sign of any arcing, although a couple of the exhaust manifold pipes were visibly glowing red - 2nd from front NS, and 3rd from front OS. No idea if this indicates that those cylinders are running lean, but I'll carry on with the other tests when I can.

[Hamish400]

Has the car been sitting for a long time? If so you may have a fuel/sediment problem.
Can you check the fuel rail pressure on full load eg by using a pressure gauge on a long hose from the cold start injector connection.
If you fill the tanks with fuel does it improve the situation. If so the fuel pump may be worn, (the increased suction head helps).

Rgds
Hamish

[wombat172a]

honestjohntoo said:

1 failing ignition passing 5500 rpm - easy to check inside the cabin, with a small temporary circuit wired to the coil connections - probably unlikely as the engine is OK after lifting the throttle.

For #1 - read how to test for 'Intermittent Ignition Faults' (scroll to the titled section) here:

http://www.vintagemodelairplane.com/pages/Download...

Okay, so today I attempted this one, and I got some interesting results.

With the ignition on, engine off, both POS and NEG are steady ON. POS flickers when starting. When running both POS and NEG remain steady ON, when misfiring both POS and NEG remain steady on.

Which isn't an observation in the list of things. Also it turns out I've used 12V LEDs rather than 2V LEDs.

Thinking the LEDs are a problem to the experiment, I then wire in 2 No. 5w bulbs, in parallel to the LEDs.

With the ignition on I have the 2 LEDs steady on and the 2 bulbs are steady on

What happens then, is the car won't start at all. It cranks but has a weak spark - I put some flashy spark testers on the HT leads to confirm this.

If I disconnect my ignition tester from the POS coil side, the car starts with a strong spark, and as soon as I make contact with ignition tester to the POS side of the coil again it starts to cut out.


So, assuming the LEDs are causing something weird to go on I've disconnected them and run the experiment with just the bulbs. Result is as above, i.e both lights are on, and then prevents the car from starting. Remove it, and the car starts fine.


Essentially, I don't understand exactly what's going on, but it looks like the ignition is a bit dicky.

Edited by wombat172a on Tuesday 30th April 20:28

[adam quantrill]

Not necessarily, the 5W bulbs you might think will take half an amp, but when wired in series across 12V they may well take an amp each, perversely, as they have cold filaments, and start to interfere with the action of the ignition amplifier.

The 12V LEDs will be fine and won't interfere so stick with them.

When you say that both are on, is that both on strong as in full-on 12V, or both on half (i.e. flickering)? If in doubt, use your fingers as a stroboscope - with your hand between the LEDs and your eyes, spread your fingers apart half an inch each, then wave your hand up and down so your fingers interrupt the light. If it's flashing (like as in a fluorescent lamp or an old CRT telly) you will notice.

[wombat172a]

Hi Adam,

By on, I mean they're both on fully. I did try to record it, but I seem to be having problems with the camera.

This is essentially what I get:

http://youtu.be/eUF72D2HvQM

Apologies for a seriously crap video! I'll try again tomorrow.

Edited by wombat172a on Tuesday 30th April 23:28

[honestjohntoo]

The test process was not designed for 12 volt LED's so I dont know how they would respond.

When a 2 volt LED energises the current through the coil is limited to just 15 ma by the 680 ohm resistor.

How would a 12 volt LED work with a 680 ohm resistor in circuit. I dont know!

Wiring bulbs in parallel with the wrong LED goes beyond my ability to explain.

The circuit as designed is very simple but clearly does not like being redesigned

After removing the 12 V LED's one expects the 5 watt bulbs to respond as described but perhaps they are drawing too much current thro' your coil not allowing the coil magnetic core to charge/collapse properly.

In which case the 2 Volt LED option is the next step.

Look at the various aspects of the circuit this way:

Coil positive is normally a solid 12 volts from the ignition switch.

Putting a 5 watt bulb between there and earth will light the bulb but cannot kill the ignition - IF it does then check the local wiring.

Add a 12 volt LED in parallel, and the ignition fails to work.

Ergo 12 volt LED's are deffo causing a problem.

Moving on:

When the amplifier is working as a switch for the coil, then the 2 volt LED across the amplifier (ie the switch) should respond to the voltage changes at coil negative. - ie going from near short circuit to something more than 2 volts. I cant explain how a 12 volt LED will respond?

Obviously, if the amplifier goes permanently short circuit then a bulb or an LED will extinguish.

From the results reported I would guess your primary side ignition is probably OK - after all it works OK under most conditions.

You can still test all the primary side components individually, ignoring the intermittentcy testing.

If necessary you can move to testing secondary side components, also covered in the essays.

What you have determined however, is a possible problem with the circuit design which I will check out on my known good SD1 ignition.

[Wedg1e]

LEDs by design expect to see 2 Volts (well OK, technically they want to draw 20mA or so). LEDs intended for 12V have a resistor built-in (470R to 680R or so depending on maker). If you're putting a 680R resistor in series with a '2V' LED then all you are doing is creating a '12V' LED. If you use a '12V' LED and put 680R in series you would halve the current available to drive the LED so it would tend to be dim.

[Wedg1e]

wombat172a said:

Thinking the LEDs are a problem to the experiment, I then wire in 2 No. 5w bulbs, in parallel to the LEDs.

With the ignition on I have the 2 LEDs steady on and the 2 bulbs are steady on

What happens then, is the car won't start at all. It cranks but has a weak spark - I put some flashy spark testers on the HT leads to confirm this.

Just running through this in my head... how is the lamp connected to the coil (-) side? From the coil (-) to ground? If so, the car won't start because the low resistance of the lamp is holding the coil (-) at or near ground, so the ignition amp will have minimal effect.
If it's coil (-) to a 12v supply, then the same applies because as the ignition amp is trying to pull the coil (-) to earth to make it fire, the lamp keeps pumping 12v in, so the coil doesn't see a nice square-wave...

Might sound daft but have you checked the engine to chassis and battery earth leads/ straps?

[adam quantrill]

I think 12V LEDs will be OK too - you don't need the extra resistors - but they will do no harm either.

Looking at the vid yes it's truncated so not much to see there. However I did notice that the battery voltage jumps up to the high side....

Can you do the next vid starting with ignition on for a few seconds, then cranking, then revving till you get the problem?

[Danny Hoffman]

From your tests so far I'd say the low voltage side of the coil is ok

[wombat172a]

Thanks for the replies again chaps.

I'll try and get into town today to pick up the 2V LEDs, so that the experiment can be run again as per instruction.


In the meantime how is this for a thought:
Last year the ingition was re-timed as it wasn't running quite right, I can't remember the exact symptons now though. This was done with the old Ingition Module, Coil and FPR valve.

With new parts fitted, could it be as simple that the ignition timing was compensating for the old parts, and now with new bits (mainly the Ignition Module) the timing isn't correct, so would benefit from being re-timed?

[Danny Hoffman]

Hi,

You don't need to get 2v LEDs as per Ian's email. The ones you are using are fine.

Definitely worth checking the timing when changing the ignition amp, did you remove the distributor to get access to it?

Danny

[honestjohntoo]

wombat172a said:

could it be as simple that the ignition timing was compensating for the old parts, and now with new bits (mainly the Ignition Module) the timing isn't correct, so would benefit from being re-timed?

Maybe not, because timing is a purely mechanical process, but where doubt exists, checking idle ignition timing at 600 rpm with the vacuum feed disconnected is so simple, requiring just a bog-basic timing lamp, it would be prudent to verify its still OK, before testing for other more spurious faults.

However, because you have detected a problem at full load WOT it may be your timing is sufficiently off to cause the problem thereabouts?

So, because you are seeking the best possible ignition timing under full load, it could mean an expensive journey to your local rolling road or Dyno shop. ALTERNATIVELY, at nil cost you can adjust/reset your timing to a point just shy of pinking. This is my dissertation on the whole timing process - for your consideration.

Timing Procedure can vary according to the equipment being used but basically, with a simple stroboscopic timing lamp, ensure the setting is made below normal idle speed (say 600 rpm) and with the vacuum advance tube disconnected from the ported vacuum outlet. Loosen the clamping bolt just enough to allow the distributor to rotate. Follow the manufacturer’s instructions and set the timing to the recommended (or recalculated) figure. Reconnect the vacuum advance tube and readjust the idle speed to the normal setting. Secure the distributor clamping bolt, unless moving on to the next optional, but wholly appropriate empirical step.

Empirical Timing is a dynamic process performed with the distributor clamp bolt loosened just enough for the unit to be rotated without undue force. Drive the car on a selected route where one can subject the engine to heavy uphill acceleration at low engine speed in a higher gear to listen for pinking, the high-pitched knocking noise, caused by premature detonation. If pinking is heard, stop the car and rotate the distributor very slightly to retard the ignition and test the car again under the same conditions. Use small adjustments each time until the pinking is eliminated.

Conversely, Pinking may not be Observed at the start. To achieve optimum burn efficiency with a chosen fuel grade, advance the ignition under the same conditions until pinking begins then back off until it just disappears. Tighten the distributor clamp bolt and to finalize, back at the garage, re-measure the timing and note the actual setting for future reference. A good wheeze is to scratch a location mark on the distributor body to aid re-assembly after any future removal. As mentioned, this is not the whole story these days because fuel quality varies even between the same nominal grades and definitely between vendors who favour different additives, so decide upon the best grade and vendor to suit ones car, wallet and shopping habits, then stick to it.

There is a school of thought that the empirical process will not reach the very best timing position for maximum torque AND power. Quite right, I would expect, but perhaps owner enthusiasts would be happy to get it very close, WITHOUT spending money for a specialist to have a go.

Edited by honestjohntoo on Wednesday 1st May 12:00

Edited by honestjohntoo on Wednesday 1st May 12:02