Help with misfire fault please

[taylormj4]

Misfire started on way home in very wet weather yesterday. Had to use throttle to prevent stalling.
It varies from lumpy running to a complete cut out and then sudden recovery to full power.

I thought something had got wet and left it to dry out overnight.
This morning starts and idles OK but then misfires when driven or revved up even at relatively low revs. If you let it come back to idle, idle is generally OK.

I have made a quick check that all plug leads and wiring to the AFM, igniter and lambdas look sound and have not come off. Can you help please with a list of possible faults. I have Rovergauge so will be plugging that in tonight.

Otherwise:

Faulty wiring terminations
Faulty igniter
Fauty coil
Rotor arm / dizzy cap
lambda sensor
faulty ignition relay
faulty fuel pump
faulty fuel pump relay
ECU circuit board cracked (thinking unlikely as it wouldn't have been affected by damp driving unless hole in the floor!)
fuel starvation (although pump primes OK)

A clue maybe that it came on whilst driving in wet weather - didn't really go through any big puddles but spray etc was bad enough to make the brakes initially less effective on first application after a run. The other clue maybe that there seem to be less problems at idle.

Cheers, Matt

[taylormj4]

Hi ButtonIt, yes sounds very similar, except that my car is a daily driver except for weekends, so couldn't be due to damp issues inside the car.

So they could be very similar if it is the rain / spray getting into the engine bay that is causing the problem.

One other thing I remembered was that there is an odd smell too. Not unburnt fuel, more like an electrical fault smell. Can't quite place it, or decide where it is coming from annoyingly.

One good thing that has come from it is whilst looking round the engine bay trying to find the fault, I noticed a thin whisp of steam coming from the radiator so I have at last found the source of the very slow leak in the cooling system !

Cheers, Matt

[taylormj4]

QBee said:

Sounds a bit like a plug or plug lead shorting, or something similar. I get those symptoms quite frequently either in heavy rain or after a rather too vigorous session with the bucket, sponge and hose, but it usually clears as the engine warms up, or after a short run and ten minute stand for the wet to evaporate.

One other random thought - If the battery seems a tad sluggish on engaging the starter it may be worth checking the 100 amp fuse (under the car), as running on battery rather than alternator can tend to exaggerate any slight deficiencies in the HT system (if it has indeed blown).

Hi QBee,
I know what you mean. I too have had that after splooshing through a big puddle or something but this is different. The whole engine stops firing for a second of two and then comes back with full power or it can also splutter on-off-on-off at a faster frequency. It is more violent than when you run out of fuel or fuel starvation, so I am suspecting an electrical problem but one that takes all cylinders out simultaneously I think. Could be wrong though.

When I left the car on the drive last night, it was well warmed up (steam coming off the wet bonnet surface) so should have dried out well but still had the same issues this morning.

Battery / starter seem fine. Will check alternator output though. Thanks.

[taylormj4]

So last night I pulled all of the electrical connections off the igniter, the coil, the dizzy king lead and the AFM. None were wet or corroded.

I changed the coil for a known good one - no change.

Following discovery of the leak in the radiator yesterday, I checked the level in the swirl pot and found alarmingly that I needed to add 2 litres of coolant to the engine. I am hoping that this is nothing to do with the poor running.

I tried without success to get Rovergauge to work on our new laptop that is running Windows 7. The drivers for the USB-to-ECU lead off the disc don't seem to like Windows 7. I tried to download the latest from the internet but no success. Have PM'd Mark so hoping for some help there.

The engine will always start straight away but will either:
1. cut out immediately
2. run at idle until the throttle is touched and then cut out
3. if throttle is opened widely it will rev up but run very badly with lots of misfiring

Cheers, Matt

[taylormj4]

FlipFlopGriff said:

Had about 18 months of this plus stalling when the clutch was dipped which was extremely dangerous. Have you checked the engine earth strap - mine was 90% through.
FFG

That's an interesting one FFG, will look at that. Where is it located ?
Wouldn't I see other problems with starting if that was poor. Or does the starter have another earth connection ?

[taylormj4]

fatboychim said:

Have you determined which cylinder it is missing on?
You may need to bleed the coolant system out to make sure there are no air locks if there are this will affect the running as the water temp sensor will not be functioning correctly.

It changes so rapidly, it's hard to say. Sometimes all cylinders misfire / complete loss of power.

Also now tried swapping the ignition / fuel relays over - no difference.

[taylormj4]

Update:

So after much support from Mark/Blitz over the weekend on how to get Rovergauge to work with Windows 7 64-bit, I was embarrassed to find that the COM port driver did not install because I didn't have admin rights. That now resolved and many thanks again to Mark for the fantastic support.

So results from Rovergauge:

Two fault codes logged: 12 (Airflow meter) and 23 (low fuel pressure)

I think 23 may be a red herring, being from a fault several months ago when the car ran out of fuel when a faulty earth on the fuel gauge causing falsely high readings. So ignoring that one.

With engine off, throttle pot range is healthy.

Codes cleared and engine started. Idles OK.
Engine temp, fuel temp, rpm, target rpm, volts all seem correct.
Inj duty cycle is 6%, pulse width 3.35msec (not sure if those are good or not).

Fuel map position at idle was 56, which seems usual.
MAF was 7% (linear)

Strange thing initially was that the lambdas (short) did not kick in on Rovergauge so I opened the throttle slightly. The engine speed then started hunting around and I saw violent changes in the fuel map position.
The lambda (short) readings kicked in and were both maxed at -100%.
So this means running very rich and the lambdas are telling the ECU to lean out but probably not able to correct that much ?

Open the throttle more and the fuel map position goes right to the bottom of the matrix.
MAF reading 0 to 34%

I then tried running the engine with the MAF disconnected. The fuel map position oscillates up and down the column between 1B and FD.

Then a strange thing:
I turned IGN off then ON again and the pump primed but it cut out much earlier than normal and the Rovergauge communication light started oscillating red/green/red/green repeatedly. Anyone know what this means ?

Looking at fault codes, 12 (Airflow meter) is up again but presumably due to running without the MAF connected.

Ign off/on again and the red/green lamp oscillation occurred again and several odd readings were displayed, e.g. speed=56mph, rpm=7317 and fuel temp=133deg C !
Is this just a comms error or indicative of a faulty ECU ?

Suspecting the MAF, I fiddled with the wiring where it runs into the MAF connector - never been very impressed with the appearance of this joint since I bought the car. No waterproof boot and wiring forced to enter the connector at a sharp angle due to limited clearance under the bonnet. I taped the joint up to waterproof when I go the car but that will never be 100%. Untaped the joint and it looks dry and clean. Wiring in good condition but a bit crumpled.

The engine now started up a ran as normal: stable fuel map, lambdas slightly positive and cycling as usual. Drove the car a short distance and all seemed normal so thought I'd found the fault (bad wiring into MAF connector). However, I then ran the engine at high revs for a short period just to check and the problem returned.
However, this time I noticed that the MAF reading was jumping from 0% to 100% (direct)
Fault code 19 popped up (throttle pot lo/MAF hi)
So again the MAF or its connector is suspected but is it the connector or the actual MAF unit that is faulty ?

This evening I tried in vain to remove the pins from the female MAF connector to cut out the crumpled wire sections and remake the joint. Managed to get the hinged cover on the plug open but couldn't extract the pins. I realise there are lugs on the pins to hold them in, just couldn't release them.
Instead, I bared off a bit of insulation on all 4 wires and made continuity checks between all 4 and the pins in the female plug - all tested OK.

I then tested the MAF volts, with IGN on but engine off. The voltage between the red/black (ground) wire and the blue/green (+ve) was 1.11V, which I believe (according to ACT's diagnostics info) indicates a faulty AFM (should be 0.3-0.34V). Cycled ign on/off a few times and MAF voltage between these two wires settles at 1.1V each time.

So diagnostics suggest faulty AFM. Does that seem right to everyone or have I gone off on a tangent ?
Is it possible for an AFM to fail, then suddenly work until you go for high throttle and then fail again ?

Cheers, Matt

Edited by taylormj4 on Monday 9th June 23:10

[taylormj4]

blitzracing said:

Id suspect the odd readings and the red / green coms light is corrupted data- the ECU relies on the all the data being sent within a set time frame- and if it falls outside this window the data gets corrupted and there is no parity correction- its all very basic comms wise, but a disconnect and reconnect should be about the worse you have to do.

As for the AFM- its possible the ECU is dropping in and out of fall back mode if the AFM output is inconsistent- The readings are very odd to get 100% however - its like its shorting to 5 volts intermittently- but that standby voltage is far to high anyway. The only other thing to try is put the AFM reading to direct and see what you get at idle- its more accurate than linear low air flows- it should be between 32 and 35%. Above this it will over fuel horribly.

Thanks Mark,
I'll try that tonight. I have to say that Rovergauge and your lead and drivers are (for me anyhow) one of the biggest advancements for DIY TVR owners in the 14 years I have had it. The advice to new owners used to be buy Steve Heath's book but I would say it is now buy the book and the Rovergauge/lead kit. Fantastic work and for the software to be available for free is just incredible.

[taylormj4]

ButtonIt said:

Good work Matt, interesting read. There seems to be a lot of rough running issues at the moment! Hope you get it figured out.

At £160 a pop, it's not easy to justify just buying a new MAF sensor. Whereabouts are you? If you're close by I'll come over and you can try mine.

Hi ButtonIt, Yes was reading your thread last night and we sound to be having the same problem. It always amazes me how you can come onto PH intending to start a thread and you find someone else has beaten you to it with the same problem within the last couple of weeks. Very surprising when you think how many Chimaeras there are. I just it's just that there are lots of us on PH.

I'm in Cardiff BTW.

Cheers. Good luck with yours.

[taylormj4]

Anyone know if the 5AM MAF is serviceable ?
Have read about the wire getting gunked up and people using Easy Start to clean it up a bit.

[taylormj4]

blitzracing said:

It wont be dirt- it will be a failing semiconductor on the circuit board inside, having been cooked by the TVR exhaust manifold- Im surprised they last as long as they do without a heat shield.

Thanks Mark,
I'm reasonably handy with a soldering iron. Is it something that is repairable / common component that goes ?

I ran Rovergauge again tonight with the MAF reading on "direct" and was getting oscillations from around 29 up to in the 50s. Caused a stall at one point from idle.
I have videoed the Rovergauge screen to show everyone what I am seeing and will try to get a link to it sorted soon.

[taylormj4]

blitzracing said:

You have to dig your way in through the cover with a sharp knife- its well glued in with silicon. There is an Op-amp in there you could try swapping if you can get a 20 year old component- but Ive never managed to get a circuit diagram- also the hot wire sensor has steel connections that dont resolder well- and you have to clip the connections to remove the circuit board anyway.

Thanks Mark.
What I am still a bit confused over is how fiddling with the connector wires could have made the system suddenly work again and then a short period of high revs made the problems restart. Wouldn't you expect a MAF circuity failure to remain once it had started ?
Very odd.

If the MAF is shot. Is it worth considering alternatives to the 5AM ?

Edited by taylormj4 on Wednesday 11th June 12:53

[taylormj4]

Hi Blitz,
I'll take a look at that kit - think I found it a day or two ago actually.
However, I'd thought I'd read that the AFM isn't the place to start with increasing a standard engine's ability to breathe. So thought I'd be better off sticking standard and looking at the inlet manifold / trumpets if I wanted to go down that route. I've already fitted the smooth bore pipes.
I don't want to go too crazy as it's a daily driver and I'm quite keen on originality but I would love to solve the low throttle opening shunting behaviour so this may be an opportunity to do that.

In the meantime, I understand that I can run the engine (maybe) with the AFM disconnected with fueling done of the TPS. Presumably fueling will be a bit approximate like this. Do you think there is a possibility of wrecking the CATs if I drive it like that ?

Edited by taylormj4 on Wednesday 11th June 17:33

[taylormj4]

blitzracing said:

The bosch AFMs are no more to buy than the Lucas so if you are buying a new AFM it would make sense to go to Bosch, but the chipping is the issue. If you could get a chip for £50 it would be a very viable swap out for the 5AM with a bit of wiring, but until someone comes up with an alternative option to the MA chip its too expensive just to replace the 5AM. And no dont run without the AFM-its get you home, not drive it around mode!

Edited by blitzracing on Thursday 12th June 12:35

Thanks. So what benefits would I likely see with a Bosch AFM on a stock engine? Also, I thought that others were now offering alternative chips to MA. Is that not the case ?

[taylormj4]

Bit of an update:
I have cut off and replaced the AFM connector completely. Reconnected and teh problem seemed to have been sorted. Took it for a drive, let it warm right up, gave it some revs and still fine.

Then I gave it a real boot up a hill (nearly full throttle and up to around 4500rpm) and as I lifted off the throttle, the problem started again.

RG was linked up and found fault codes of:
AFM fault
Throttle pot low/MAF hi

Is it possible for an AFM to fail under high throughput of air or when it gets hot but then work normally again when it cools and at low revs ?

The connector repair certainly seems to have helped but not remedied it completely, which is very odd.

What does the ECU have to see to register and AFM fault ?

Edited by taylormj4 on Wednesday 18th June 15:23

[taylormj4]

blitzracing said:

The ECU software is smart enough to know you cant have high airflow with the throttle shut, hence the error code with both devices listed. You could have a normal AFM output, and a break in the throttle pot track, so it gives a low voltage reading with a high AFM reading- so it might not be an AFM going high, but an intermittent throttle pot. And yes electronics can fail when hot and recover when it cools. You could try heating the throttle pot a bit with a hot air gun, and check you dont get drop outs with the RoverGauge check when you move the throttle.

Edited by blitzracing on Wednesday 18th June 21:43

Hi Mark,
I should have added that on return from the road test where the high revs caused teh spluttering to reappear, I tested the throttle pot with IGN off as follows:

Watching the RG display, I gradually opened the throttle in 1% steps from 0 to max (97%) and then back again, making sure that every 1% step was displayed. The test was successful. So it appears that the TPS is healthy (although engine bay would have cooled a bit).

I have a new AFM on the way. Went standard 5AM. Would like to have looked at a Bosch unit but I've looked at going through the MA process in the past and the guy is too busy to return calls, so it would be months probably before the car got back on the road.

[taylormj4]

New 5AM AFM finally arrived and fitted. It's actually date stamped 1997 so exactly the same age as the car funnily enough.
Car now working well again throughout all the rev range and loads.

Very thankful to Rovergauge/Blitz's lead for helping me diagnose the fault and saving me from the dreaded gradual replacement of parts until resolved technique !

I notice that the long-term lambdas are still showing -99% and -94%, caused by running with the fault. The short terms are compensating in the +ve range so car runs OK.

I presume the ECU will return the long-term lambdas back to their usual values or I could reset the ECU. Is that correct or should I be adjusting the pot on the AFM ?

Now I need to make a heat shield for the new AFM. Using some tin foil at the moment but doesn't look great.

Edited by taylormj4 on Tuesday 1st July 13:54

[taylormj4]

QBee said:

£19 would get you this off Classifieds:

http://www.pistonheads.com/classifieds/parts-and-p...

Ha ha thanks QBee, that looks like Brian off Confused.com !

[taylormj4]

blitzracing said:

It should reset after the car idles for a while.

Thanks Blitz,
So let it warm up to get lambdas pulsing then idle.

Presume once that has happened, I should adjust the trim pot to give the central position with the short terms as its a new AFM ?