Mystery Misfire

[geoffbeaumont]

I've been playing hunt the misfire for the last few days on the range rover (it's had the misfire for ages, but I've had to put it back on the road so I'm quite keen to sort it out...).

Hotwire EFI system, Landi Renzo LPG kit (seems to be another Leonardo clone, at least the electronics side of it).

The misfire affects both LPG and petrol. Petrol won't idle unless I unplug the mass air flow meter and force it into limp home mode.

I've been through the hotwire diagnostic routines (the ones that don't involve a testbook, anyway), and most things check out okay. The exception is the voltage across the throttle position sensor - I can't find any... The TPS checks out okay separately, and I'm getting a 5V input on one of the connector pins. All three pins on the connector check out okay back to the appropriate pins on the ECU plug. Wierd...

The LPG kit is wired into the TPS and can't easily be removed as it's hard wired on the sensor side of the connector, but I swopped it out for another TPS which behaved exactly the same.

There is an additional bit of wierdness going on - the alternator charge light is glowing faintly most of the time. I've fitted a new alternator, which has made no difference at all (old one had knackered bearings, so it needed doing anyway) - it's charging fine, though

Ignition, so far as I can tell, is fine - it's running EDIS, good strong spark, good leads and the fairly new NGK plugs look fine. The Megasquirt is away getting recased by a friend at the moment, so fixed (8deg) ignition advance. Gutless, but should be perfectly stable, especially at idle. I can't check it out any more thoroughly until the weekend as I get back from work fairly late, and running a misfiring V8 with two thirds of a free flow exhaust (thanks to the welder who still has my tailpipe ) outside the neighbours houses probably won't go down too well...

I do have one area of utter dodginess with the electrics at the moment - the old ignition coil is still in place, sinking its spark via an old spark plug cable tied to the air filter bracket. It's still providing the timing signal for both fuels, but has been like this for about a year. When I get the Megasquirt back this becomes redundant but it's not worth doing a proper fix in the meantime*.

So far as I can tell I've got a good earth from the engine block - had problems with this before and fitted decent earth straps.

I know I've got a couple of things here I need to attend to/check out if only so I can eliminate them, but is any of this ringing bells? Particularly the rather wierd electrics round the TPS.

* - yes, I know...it was worth it a year ago. Wasn't planning on taking this long to finish the job

[jwriyadh]

If your misfire occurs under acceleration then the TPS could be causing it. No TPS signal means no acceleeration enrichment.

Check it out using the LR hotwire guide method:

Pull ECU plug and check resistance between pins 3(ref voltage source) and 25 (sensor ground). This should be around 5k ohms.

Reconnect ECU plug and turn on ignition. Now back probe pin 20, then operate throttle and watch the swing. This should vary from 0.5 volt to 4.8 ish and smoothly.

jw

[jwriyadh]

Alternator warnig light glowing.

With engine running take a critical measurement of voltage across battery lugs (not clamps) and then across the alternator output stud to alternator frame. If these are different then the connections between the two are iffy. To find the bad point measure volts from alternator frame to battery neg lug and from alternator output stud to battery pos lug. Both should be less than 0.5 volt. Where a larger drop is found then by moving one of the voltmeter probes the problem area can be found.

If there are no voltage losses in the charging area check the voltage with reference to the battery neg lug at the ignition switch input and output to the charge warning lamp feed with the switch on and engine running. Again this should read very close to battery voltage.

It's not unusual for ignition switches to fail just as the connectors suppling it with battery power do.

jw

[geoffbeaumont]

If your misfire occurs under acceleration then the TPS could be causing it. No TPS signal means no acceleeration enrichment.

It's worst at idle, opening the throttle generally helps - not sure if that's just because it gives it some momentum though.

Check it out using the LR hotwire guide method:

Pull ECU plug and check resistance between pins 3(ref voltage source) and 25 (sensor ground). This should be around 5k ohms.

This bit works fine, as does testing the changing resistance across the TPS pins. The TPS connector pins test out fine to pins 3, 20 and 7 (via coolant sensor - should have metered to 25) as well.

Reconnect ECU plug and turn on ignition. Now back probe pin 20, then operate throttle and watch the swing. This should vary from 0.5 volt to 4.8 ish and smoothly.

This is where it goes wrong. I have no voltage at all at pin 20, even though I have 5V at the TPS connector

Looking at the wiring diagram, the TPS shares pin 25 (5V?) with the coolant temperature sensor, fuel temperature sensor and MAF, so I guess any one of these could cause strangeness. Both temp sensors are around 1Kohms, which is 40degC - seems reasonable, as the truck was running not long beforehand. I'll try unplugging them, and I've spares of both (though not known good), so I can try swopping them out if necessary. Unplugging the MAF doesn't seem to make any difference to the voltage.

[geoffbeaumont]

Alternator warnig light glowing.

With engine running take a critical measurement of voltage across battery lugs (not clamps) and then across the alternator output stud to alternator frame. If these are different then the connections between the two are iffy. To find the bad point measure volts from alternator frame to battery neg lug and from alternator output stud to battery pos lug. Both should be less than 0.5 volt. Where a larger drop is found then by moving one of the voltmeter probes the problem area can be found.

If there are no voltage losses in the charging area check the voltage with reference to the battery neg lug at the ignition switch input and output to the charge warning lamp feed with the switch on and engine running. Again this should read very close to battery voltage.

It's not unusual for ignition switches to fail just as the connectors suppling it with battery power do.

So far I've just checked voltage between battery positive and negative/alternator body/engine block and resistance from battery neg to block and alternator, so I'll check the above tomorrow and see what I find.

Cheers.

[jwriyadh]

Pin 25 provides a common ground for the MAF, coolant sensor, fuel sensor and TPS.

Pin 3 provides a reference voltage for TPS.

Pin 20 inputs the TPS signal.

At the TPS this translates into final wire colours as;

Pin 25, green

Pin 3, yellow

Pin 20, red.

at the TPS.

Slice open the sleeving and check voltages from pin 25 at ECU plug while plugged in with ignition on, use a very pointed probe and pierce insulation. Yellow should be steady at 5 volts, red should swing with throttle movement.

Be wary, IIRC, pin 25 is not connected to vehicle earth with the ECU plug out. Always use pin 25 for ground with the sensors.

jw

[geoffbeaumont]

Pin 25 provides a common ground for the MAF, coolant sensor, fuel sensor and TPS.

Pin 3 provides a reference voltage for TPS.

Pin 20 inputs the TPS signal.

Ah! I had them the wrong way round then (pure guess work - I was just looking at the wiring diagram).

Be wary, IIRC, pin 25 is not connected to vehicle earth with the ECU plug out. Always use pin 25 for ground with the sensors.

Now that bit they don't tell you... Kind of essential information, that!

That may well be all that's "wrong" with my TPS.

Thanks,

[geoffbeaumont]

Didn't get as far as investigating the alternator light any further tonight, but I think I'm getting somewhere with the misfire.

Thought I'd eliminated ignition trouble a while back, but decided to swop back to the dizzy just to eliminate it. The coil was...well, a bit charred, to be frank , so I swopped that out for a spare and tried firing her up. No difference, swopped the HT leads onto the dizzy and it ran very badly (wouldn't idle and gutless on the road) but definitely felt different - mainly a bit more even - that on the EDIS, so swopped the plugs out for some used champions gapped to 0.6mm (what I used to run) as the NGKs are gapped to 1mm which is a bit optimistic with the standard spark, and fiddled with the timing.

It's still pretty poor, idle is if anything worse but it's much smoother on the go, just feels like the ignition timing is miles out. It'll get me to work tomorrow and I'll get the timing light out on Saturday. Won't know for certain until then, but it looks very much like something on the EDIS rig is dieing.

[geoffbeaumont]

Saturday morning got off to a fairly promising start - pulled out the distributor, set it correctly (it's only supposed to drive the oil pump now, so I just shoved it back in after doing the head gaskets) then dug out the timing light and adjusted the timing. It now ran fine on petrol (although still only in limp home, so there is something else wrong as well ), and lumpy and a bit lacklustre on LPG. That's fine - it never did spark well enough for gas, that was one of the reasons I replaced it.

Coming back from London this evening it started briefly dieing - switched to petrol and it was okay, but wouldn't switch back to gas. More like fuel starvation than ignition trouble, but as it was pissing down with rain I assumed it was the other reason I got rid of the old ignition and stopped at Oxford services to purchase WD40. Still wouldn't switch to gas so I forced it to start on gas and carried on - not far up the road it started cutting out again and I had to switch back to petrol. It wasn't until some time later that it occurred to me that my feet were rather cold...

Oh, yes! No heat from the vents - I've got an air lock in the cooling system It was fuel starvation - the vapouriser was freezing up - and it wouldn't switch to gas because it (rightly) didn't think it was warm enough. So now I've got to bleed the cooling system and try and figure out why it got an air lock in the first place Head gaskets were done recently, so shouldn't be that. Suppose it could be a leak that's sucking in air as it cools?

Sometimes I really think I could do without a land rover...

[geoffbeaumont]

Well, the air lock or whatever it was remains a mystery and probably always will - couple of days after I posted the above, having not had time to do anything about it, I noticed I'd got warm feet on the way back from work. Double checked there was heat from the vents then switched over to LPG and it's been fine since

[COLD SURFER]

I got a new set of bosch wires at AUTOZONE (US, I know, but they were about $39 instead of the $80 special order ones). then while I replaced them I used BLUE RTV on each plug and cap terminal end, ensuring no water or spark leaking. installed new plugs too. I sprayed a good dose of wire dryer in the distributer and RTV'd that on too. RTV'd the coil to the cap AND both terminal sides of the coil. The result was OUTSTANDING and only $5 more than installing new wires. no misses, no pops, no questions if I submerged the engine.

Ever test it at night? open the hood away from street lights and spray a misting of water around the coil, cap, plugs. If you get a light show then I recommend the RTV silicon job. Call it a boob job, it does wonders for an old sagging rover.