Classic Hotwire V8 Runs Rough at Idle

Hi all,

Still trying to get to the bottom of my running issues. Hotwire 3.9 V8 in a ‘93 Vogue LSE.

Symptoms are rough running / stumble / intermittent random misfire at idle, rich smelling exhaust at idle and big mixture corrections. Runs pretty nice I think under any load. Reving cleanly and returning reasonable economy (21MPG on a recent run to Scotland cruising at 65 lightly loaded and on all terrain tyres).

I’ve done so far…

New lambda sensors both banks
Repaired wiring issues to lambda heater on odd bank
Smoke tested inlet for vacuum leaks - replaced inlet manifold gasket as precaution in case it was leaking to the crankcase, replaced throttle spindle gaskets. No leaks remaining.
Checked catalysts for blockages as i knew they’d failed and were rattling - they’re both completely empty and the blockage has been removed from the centre silencer.
Fitted ignition kit from RPI including amp, coil, plugs, leads, cap and rotor.
Checked base idle, ignition timing and MAF sensor voltages. All seem fine to me.

I have rovergauge on a laptop so can check codes and do live monitoring. Generally it will idle pretty well from cold and get a little worse as it warms up. At idle the short term mixture trims are maxed out +100% on both banks. If allowed to idle for long enough it will throw a code for one or both lambda sensors and then immediately start running much rougher. When codes exist for both lambda sensors it may even stall when coming to a stop in drive. Even then it still cleans up and runs nicely enough off idle.

Anyone have any thoughts where to look next? MAF sensor? Compression test? Injectors? Fuel pressure? I’m coming to my wits end. Much more experienced with diesels really and haven’t really got he money to fire off the parts canon and replace the lot! have a picture of some catalyst extracted from the silencer for your entertainment (or not - apparently I don’t have permission to do that any more).

All help and advice appreciated.

Rich

Check for exhausts leaks - if it's sucking in air, the ECU thinks it's running lean (too much O2 in the exhaust) and will add fuel trim - hence the +100, maxed out. Then it starts running super rich.It doesn't take much of a leak, either.

Once the Lambda codes are triggered, it defaults to the limp map, which is really rich on a Cat tuned car.

Which MAF voltages are you talking about? the CO trim is entirely different to the Air Flow voltage. Rovergauge should show you about 30% at Idle on "Direct" for MAF, and the fuel table indicator should be in rows 2, maybe 3 but fairly stable. You can also fit a non-cat tune resistor, then read what Rovergauge sees for the CO trim voltage (More accurate than a meter reading, but more ****ing about). The CO trim is less critical on a cat-tune car anyway - its "replaced" by fuel trims at idle and in the lower revs/load parts of the map.

Hi,

Thanks for the message back. When I said id checked the MAF voltages I meant both the output and CO trim. Apparently the CO trim should be set to a specific voltage on lambda cars and have an anti-tamper cover. Mine doesn’t have this so I checked and set it to whatever it was meant to be. Frankly I’ve also tried adjusting across the whole range and it seems to make naff all difference to how the car runs anyway. In direct mode it’s reading 28-29% at idle and the map cell jumps around but is always around those rows.

Your suggestion about exhaust leaks is a good thought too. I have been battling with a head to manifold gasket leak on the even bank. Two sets of gaskets have leaked straight away so I think I’m going to have to source a new manifold for that bank - the spot where it’s leaking on the back side of the no. 8 port has very little gasket in contact with the manifold. Seems odd that my issues are affecting both banks though so may this is yet another one of the issues, and not the only remaining one.

Rich

CO is usually 1.8V on a lambda car. It only comes into play in limp mode or if you select a non-cat tune. Either way, both CO and Lamba fuel trims only work on those low RPM/Load cells, it progressivly moves to open loop, based on the fuel tables alone as those criteria increase.

You could also check for a leaking injector - see what the fuel pressure does after shutoff?

I think doing a compression and fuel pressure test might be worthwhile so I know a bit more what I’m working with wouldn’t it.

I’m still 99% certain it’s an injection issue and 99.999% certain that it’s not ignition so you can guarantee that the engine itself will be in fine health and there will be something obvious wrong in the ignition system!

Rich

Doo@ ,Whos not been on for a while had a similar prob ,,He stripped all the wiring to the maf ,played wi cats and the like think he even changed the crank sensor ,coils leads ,,,,Turned out to be head gasket between two cylinders ,last i seen him was going to do a full nut n bolt restoration ..

Compression test may well be worth while then! If there’s anything up with the engine then it will be unceremoniously dumped and replaced with a 4.6. It’s already the wrong engine for the car so I’m not going to waste too much effort on it.

Someone correct me if I’m wrong but I’m not aware of a crank sensor on hotwire injection cars? I thought that the injection system got its engine speed information from the ignition coil and had no idea of engine position. Since it’s a batch fire system which injects to all four injectors on a given bank at the same time it doesn’t need to know engine position?

I’ve already done all the metal work on the car (nearly all) and reinstated the air suspension plus gone through the running gear and suspension. It’s not had what I’d call a nut and bolt restoration but it’s definitely had a lot of work over the last few years. There’s a thread in the projects section with photos. I’ll get it right but I’m struggling with this at the moment.

Richard

IF you have ignition issues, that can also cause the ECU to add fuel trim - A misfire actually shows as a lean condition, because the fuel doesn't burn, excess air comes through.

What is your timing at? I often run 12BTDC, but have it on 10 right now, because I was towing in Summer. It's a 9.35:1 Disco 1 3.9. I find it idles better with more advance.

Can you get the exhaust manifold face machined if possible?

Well I've received a compression tester today - Mityvac MV5530 kit from Amazon - feels very good quality if anyone is on the lookout for one. Run a test this afternoon and the results are as follows...



1 - 170psi
2 - 165psi
3 - 160psi
4 - 170psi
5 - 145psi
6 - 160psi
7 - 170psi
8 - 165psi

I didn't check the service manual until afterwards so didn't follow the procedure to run to operating temperature first. Test was completed on a cold engine with the bores dry, injectors and coil unpluged and all cylinders got 6 compression strokes. I repeated cylinders 3 and 5 with a squirt of oil in both to see what effect that would have and both picked up around 10psi. Not sure what to make of that really. The spread is a little wider than I'd like but I don't see a smoking barrel there for my idle issue? The service manual says that for a 9.35 compression engine (mines a 36D so that's applicable) results should be 170-180psi and within 10%.

Had a look at the plugs while they were out. I'll try to pop a photo in here in a moment if I can find a way around the 'you don't have permission' message. Basically they all look a touch on the rich side to my eyes but I'm not that familiar with petrol engines. Straps look clean and seem to indicate the right heat range (I'd hope so too since RPI Engineering supplied the plugs). Ceramic look spotless but the tops of the bodies are a little sooty. All 8 were dry which seems to line up with the engine using no oil at all.

Mark Piercy said:

Can you get the exhaust manifold face machined if possible?

Click to expand...

I don't know. Do you think that would be cheaper than getting a replacement manifold? I'd be inclined to just do the even bank as the odd bank is sealed nicely at the moment plus I already have one set of gaskets in stock. My worry is that I wouldn't be able to take enough material off but it's definitely worth a thought. I'll see if I can find a machinist who might have a go.

Anyone have any more thoughts?

Rich

8 - 6 - 4 - 2


1 - 3 - 5 - 7

I dropped 3 on the floor and managed to smudge the ground strap. Must have landed strap up as the gap wasn't affected.

Looking a touch rich?

Rich

RAW111111 said:

View attachment 284083
8 - 6 - 4 - 2

View attachment 284084
1 - 3 - 5 - 7

I dropped 3 on the floor and managed to smudge the ground strap. Must have landed strap up as the gap wasn't affected.

Looking a touch rich?

Rich

Click to expand...

I'd of said lean as they are white /grey not biscuit brown?? I'd sort the manifold first and then see... Put a straight edge on the manifold face. If it is convex then that's not bad as long as it's not really bad! Then tighten the bolts from the middle outward to each end.

Can you read the fuel trims for each bank or just as a whole?

Mark Piercy said:

I'd of said lean as they are white /grey not biscuit brown?? I'd sort the manifold first and then see... Put a straight edge on the manifold face. If it is convex then that's not bad as long as it's not really bad! Then tighten the bolts from the middle outward to each end.

Click to expand...

Oh, I thought that the ground straps were more an indication of heat range and the ring around the top of the threads showed more about mixture. As I've probably said before I'm more comfortable with diesels to be honest. As for the manifold, I don't think it's so much warped as eroded. Around the port on no.8 there's not a lot of overlap between the gasket and the manifold. Maybe the gasket's I've been using have oversized ports to which wouldn't help but that specific cylinder seems to be either eroded or just have had a void in the casting.

Mark Piercy said:

Can you read the fuel trims for each bank or just as a whole?

Click to expand...

Rovergauge lets you see the short and long term trims for each bank live. Both are +100% at idle and within about 10% either side of neutral on the move.

Rich

Hmm, air filter?
Fuel filter and once the manifold is sorted I'm not sure if you reset the value on the lambdas and fuel trims to see if it stabilizes?

Both were new genuine parts during the rebuild so less than 3500miles ago. I hope they’re ok. I have a fuel pressure testing kit arriving today so hopefully I can check the pump and regulator next.

Manifold I think is a good next place to sort and then see where I go from there as you say. There’s no specific button in rovergauge to clear the adaptations but the long term fuel trims clear when you reset the fault codes I think and certainly seem to respond fairly quickly.

Rich

How does the maf read? Does it répond to the pedal and airflow?

I'd suggest dealing with the exhaust manifold leak before you do anything else.
Need to eliminate known faults before doing anything else.
If you're using the individual gaskets - one per port - then they do have a right way round.

For your problem port try Firegum.
I have s/s sports manifolds on my 3.5Efi & the corrosion on their gasket faces had to be seen to be believed.
Head to gasket faces were perfect & unmarked & never had the issue with the old iron manifolds so the s/s manifold metal quality is clearly crap.
Refaced to get rid of some of the pits but others were so deep a sealant was the best option.

You could unplug the maf .it might run better or worse on preset values ,if it flags up a fault ,at least the car knows it been disconnected ...

Ratae said:

I'd suggest dealing with the exhaust manifold leak before you do anything else.
Need to eliminate known faults before doing anything else.
If you're using the individual gaskets - one per port - then they do have a right way round.

For your problem port try Firegum.
I have s/s sports manifolds on my 3.5Efi & the corrosion on their gasket faces had to be seen to be believed.
Head to gasket faces were perfect & unmarked & never had the issue with the old iron manifolds so the s/s manifold metal quality is clearly crap.
Refaced to get rid of some of the pits but others were so deep a sealant was the best option.

Click to expand...

I think you’re right that I should deal with the known issue first. It’s the logical way to move forward. I have a friend who’s main earner used to be motorsport exhausts and manifolds who recons we might be able to reface the manifold just with a linisher if it’s not too corroded. I’m using the two pairs gaskets from rimmer brothers called sport which I had read somewhere tended to seal more reliably but I’ve also been told that leaks aren’t normally an issue. Isn’t exhaust paste meant to release some gas which can damage lambda sensors or is it cats?

dubbleRR said:

You could unplug the maf .it might run better or worse on preset values ,if it flags up a fault ,at least the car knows it been disconnected ...

Click to expand...

I actually have a spare MAF but it seems to run the same on either so far as I’ve noticed and definitely throws a code without one. Both so seem to detect changing air flow as I would expect but I don’t have a known good one to compare to.

I may test the fuel pressure anyway as I now have the kit but I’m going to try to sort the manifold leak as no.1 priority before any more chasing. It might be a while before I manage that.

Thanks for the guidance so far.

Rich