After exhaustive testing I am no closer to a solution. Please help! Let me begin…

Car:

1994 (M) Mk3 – AAV engine 56,000 miles

GT injection setup (inc ECU, AFM etc..)

GT Cam

GT Dizzy

GT spark plugs

Charcoal canister and purge valve removed (and hoses blocked)

Standard air box (but no hot air selector, vacuum temp bit on elbow also blocked off)

(Only vacuum tube remaining is from back of throttle body to fuel pressure regulator)

Poloace std limit chip (though the symptoms definitely occurred before the chip was added too)

Original AAV exhaust manifold and exhaust system

MPI in-line fuel pump

Full service carried out as the conversion took place inc. fuel filter

Symptom:

The engine appears to ‘miss a beat’ around once every 1-4 seconds. The engine rocks very slightly, revs dips a little and the exhaust makes a very low tone cough – also the pressure of the exhaust gasses dips slightly at this point.

This occurs throughout the rev range (but with increased frequency as the revs increase).

Immeadiately when the engine is started from cold, for the first minute or so the problem does not seem to occur.

Aside from this the engine idles well at roughly 950rpm (adjusted it to be slightly high with the idle screw on the throttle body) and doesn’t have issues at any speed on the road

History:

The GT injection system was added by myself immediately after the head gasket was replaced (you may remember the thread of many questions about a month ago!). During this the GT cam was swapped for the AAV and the cambelt, water pump replaced. Though I definitely lack experience I like to think the bits were put together in a thorough manor.

Testing so far: (one at a time obviously)

All vacuum hoses checked closely and tested unplugged & blocked just to be thorough

Rotor arm and dizzy cap were replaced with the GT injection and still look to be in good condition

Dizzy swapped for SPI version – lumpy as leper but the symptom was still present

Dizzy cap and rotor arm swapped for the old ones

HT leads (inc. king) swapped for another set (though these were used)

A single lead swapped one at a time across all 4 cylinders

Spark plugs quickly changed for AAV version (both sets are 3-prong NGK?s)

Throttle body immaculately cleaned (not dewedged or ported)

All ECU connectors check for corrosion and sprayed with WD40

All engine and ECU earths checked

Alternator function – (13v with all electrics on, I think about 13.7+ without)

Inlet manifold visually checked for leaks (though admittedly, not the best check)

Throttle body checked for leaks

ISV removed at inlet elbow and plugged

Injector cleaner put into full tank and used

AFM played with/tested with multimeter, doesn’t seem to be the cause

Adjustment screw on the side of AFM turned fully each way, doesn’t seem to make a difference either way

Timing – Only have access to a simple strobe gun – no advance degree facility. However timing seems good. Have timed as per unusual GT digifant procedure. Also the symptom is less noticeable the more the timing is retarded, though it is still present.

Blue temp sensor – reads 224ohms at full temp and agrees with values listed on here and another (used) sensor I have in stock

Lamda sensor plugged/unplugged no difference

Throttle position sensor plugged/unplugged no difference

Throttle position sensor gives good, steady resistance based upon throttle position.

Removed and blocked oil breather

SPI in-line fuel pump doesn't change the issue (I originally though this was the problem so upgraded to MPI pump, but it still shows the symptom with an MPI pump)

The engine doesn’t use water or have any sign of repeat head gasket failure

Giving the engine more air after the AFM certainly doesn’t help

When the blue temp sensor is removed the issue is much reduced, though it is still present.

So, I’m thinking possibly a fueling issue??

Not yet checked:

ECU wiring loom

ECU (need one to swap with!)

Vagcom readings

Swap AFM

Swap GT dizzy

Could be:

The above unchecked bits

Injectors (how can I test these?)

In tank fuel pump (doesn’t make too hideous a noise currently)

Coil

Cambelt 1 tooth out??

Knackered valve?

Exhaust issues (too small a bore?)

Currently the ECU is disconnected overnight in the hope of some magic

Not yet taken it to a garage…. but will do if necessary – anyone recommend a good one in/near Coventry?

Any help of advice gratefully received… component replacements, rechecks to make, other suggests…all very welcome!!

Many thanks

Edit: Symptoms adjusted.

Edited 8 June, 200917 yr by cheys03

buuuuump!

done a compression test??

done a compression test??

Nope, does this have to be done at a garage or can it be done with a cheapish tool?

From the symptoms what could be wrong that a compression test would show?

Edit: Aha! Google says tool £15-£30, shows if the piston rings/head gasket is at fault.

Edited 7 May, 200917 yr by cheys03

Sounds like you have tried most things, have you tried anything with the injectors? Check the rail for loose connections, its known that the main plug pushes the pin out sometimes. You can check by taking an injector out making sure they all fire, obviously do it in a controled manner dont be spraying petrol on the engine

I'm pretty sure it won't be mechanical, it will be down to one of the sensors. Might be worth getting the fault codes on the ECU removed.

How did you do the ignition timing if you haven't got a dial-in advance light?

Updates:

Ignition timing - you're right Andy I can't be certain it's perfect. Apart from this issue the engine idles well and climbs revs well throughout the range. Even if it's out though the symptom is still present with the dizzy fully advanced or retarded.

Went to a couple of garages on Friday.

First garage couldn't diagnose the fault, but put it on the emmissions reader where everything was OK at idle but at higer revs the HC reading was massively too high (600-800!). They advised me to look into the ignition system. No charge - top blokes.

Second garage checked the ECU diagnostics with a simple fault reader - 3 faults for missing sensors - coolant, throttle, lamda. All gained during my testing. Fault codes cleared. Problem not fixed. They advised ignition system too and to replace the HT leads. No charge either - again top blokes.

So off the GSF I went.

Bought new HT leads and fitted in GSF car park, no better

Bough new coil, fitted at home, no better

Then this morning I changed the hall sendor from the old SPI dizzy into the GT dizzy, no better.

So with the earlier new rotor arm/dizzy cap I think that rules out the ignition system, apart from the ECU!

Injectors are the next I think - possibly one is spraying too much fuel? What do you think guys?

Edit: Also the ECU took 2 attempts to read the fault codes and 2 attempts to clear them...though this could be the reader at fault... I've ordered a VAG-COM cable so the laptop so should hopefully have better info soon.

Edited 9 May, 200917 yr by cheys03

try these guys out maybe dude?

Company CARWOODS LTD

Address HERALD WAY

COVENTRY

CV3 2RQ

Tel 01203 449533

i would check the injectors flow rate. you might have a weak or bad injector. Such as, one will be injecting too much or too little fuel when it needs to, casuing the "miss beat". if you see what i mean.

There is a flow rate test you could do (i cant seem to find the tool to do it). but it measures each injector individually. From doin taht you can see which is the bad one.

Or you could swap the four injectors and injector rail for another and see if that makes it any better.

One way is to take the rail off, 4 jam jars and run power to them with some fuel and they should fill equally. You can do a soloution of injector clean and petrol and keep thm going over night. Just put make a return hose for each jar.

One way is to take the rail off, 4 jam jars and run power to them with some fuel and they should fill equally. You can do a soloution of injector clean and petrol and keep thm going over night. Just put make a return hose for each jar.

as above good idea...

also go to a garage with an ossciliscope

this is a tool for measuring milliseconds.

there you could read you injection timming ruling out ecu faults if all are equal

what fuel are you using? cheap supermarket crap or shell?

russ

Edited 10 May, 200917 yr by GING77

Thanks for the injector test and cleaning ideas :-)

Currently the car has cheap Tesco fuel in the tank (3/4 full) but for the last month has been running BP (admitedly though it's the cheapest 95 octane stuff) out of convenience as the BP garage is the closest to where I live.

Hopefully there is another full injector rail (with injectors, fuel pressure reg etc.) heading my way so will try that as a swap-out and see if it makes things any better/different. Will let you know how it goes! Thanks for the help so far :-)

Get your ignition timing set up properly and see if the problem's still there - Digifant controls the idle by varying the spark advance/retard, so if the base timing is wrong it has a very hard time keeping the engine idling.

Andy, I bow to your knowledge but for the life of me I can't see it being timing-related. The problem is the same whether the engine is at idle or higher revs and whether the dizzy is fully advanced, retarded or fractions of either. I see your point though and will invest in a decent timing light, but like I say can't see it solving this particular problem. If I'm talking bollocks please feel free to correct!

In my experience ignition timing has never caused me many problems on GT's. I had some crap running issues a while ago, black smoke out the back, foot flat to the floor and no power. Was damn right dangerous. The problem was the blue temp in that case (not saying yours is) but the timing hadn't been done on it and I thought it was that, after doing the timing it drove better but it idled fine with it being out.

It may well be an additional problem, but incorrect timing may well be muddying the waters somewhat. What have you done with the breather pod in the back of the block?

One way is to take the rail off, 4 jam jars and run power to them with some fuel and they should fill equally. You can do a soloution of injector clean and petrol and keep thm going over night. Just put make a return hose for each jar.

it's the spray pattern that is gonna cause more of a problem if the injectors are dirty and the only real way to clean them is ultrasonicly. Also if you get them sent to a place like this :http://www.injectorcleaning.co.uk/spray.htm they'll replace filters, pintle caps and o rings too

It may well be an additional problem, but incorrect timing may well be muddying the waters somewhat. What have you done with the breather pod in the back of the block?

Ok, there's a chap down the road I can try to borrow a decent timing light from while cash is tight. (He's ex-Jaguar and has a whole garage full of snap-on tools! )

The breather pod is the original AAV one with the 3F pressure valve in place as per your original walk-through (again, this was soooo useful, many thanks. Plan to repost properly with some other useful bits with your permission once all is working ok). The breather functions well, when it's disconnected from the inlet elbow you can see light oil vapours escaping when the engine is hot. Oil is not escaping from the block anywhere at an alarming rate, so don't think there is a build up of pressure. The throttle body is still immaculate so no huge amount of oil entering from the breather I guess.

The plan with the injectors is to change the entire rail and see if it makes a difference. Though the replacement injectors will be used and not guarenteed perfect, any difference is good and may help with the diagnosis. If it turns out to be the injectors I'll most likely have a set cleaned properly and then fit those to the car.

Cheers chaps!

Update...

Different injectors, fuel rail and fuel pressure regulator fitted - no difference.

Timing check with advance timing light - timing was spot on any way.

Man this is driving (or not!) me insane!

Next possibility:

I'm thinking as a consequence of the head gasket failure the lambda sensor is fried and mis-reporting excess O2 in the exhaust causing the ECU to overfuel. By disconnecting the lamda the ECU overfuels by default so could be the reason I'm not getting any change when it's unplugged.

Does the ECU have the ability to see the lambda as faulty and list an associated fault code reliably?

My next check would be to look at the value the ECU sees for lambda using VAG-COM, but of course my cheap-ass cable has fried itself and now smells of burnt electricals having worked fine for over a week. I shan't be playing the lottery anytime soon with this luck I can tell thee

Have spent some time this afternoon between rain spells testing the Lambda sensor.

From t'internet research here and here it seems the type of lambda sensor attached to the Mk3 polo will report lean mixture as ~0.2v and rich mixture as ~0.8v.

I did the testing with the ECU lambda connecter disconnected - so you would expect the reading to be high as the ECU goes to a richer fuel map I gather.

I found that indeed the mixture was high in this state (~0.75-0.80v)

To test reactivity I simulated lean and rich mixtures by:

Lean:

Allowing more air into the mix (by partially disconnecting the main breather). This reduced the voltage (to 0v-~0.2v)

Immediately after it was reconnected the mix became rich again.

Rich:

By disconnecting the fuel pressure regulator (I believe this raises fuel pressure to 3bar?) whist blocking the vacuum tube from the throttle body. This did seem to change the lambda accordingly, but it was difficult to check for rich response as the engine was already running rich because the lambda was disconnected.

Each time the change responded quickly within 1 second, so although I was only testing with a basic digi multimeter (ideally you need an oscilloscope) it appears the lambda sensor is ok.

Also swapped the ECU chip back to std (which weirdly seemed to rev better, but not driven with it yet). But this didn't solve the issue.

I think it's time to really admit defeat and try to get a mobile diagnostician to have a look with all the relevant gadgets.

Ooooh I just hope it's cheap.

I replied on polo3.

If we really have the same problem i hope someone can shed some light as to the problem!

Cant remember if i asked already but did you check each spark plug like i did? I still cant work out what the shiny yellow chemical deposit on plug no.3 is...

Have you checked your cam timing?

Do you have a spare AFM to test?

Have you done a compression test?

Have you checked all the wiring you changed?

Have you checked your cam timing?

Do you have a spare AFM to test?

Have you done a compression test?

Have you checked all the wiring you changed?

Hi Andy, thanks for the ideas. So far:

- Cam timing - I've only checked that the timing marks match up, i.e. at TDC the crank pulley and cam pulley match their TDC marks. Would you recommend checking to see if the problem improves one tooth either way?

- Unfortunately I don't have access to a spare AFM

- Compression tester is on it's way, hope to arrive tomorrow.

- I've done some rudimentary checks on the ECU loom (this was the only loom I changed). Referencing the Haynes manual all wires from the ECU multi-way connector to the relevant connectors report as ~1ohm. As do the earths on the ECU connector. Admittedly though these are hardly thorough checks, but the best available without ripping the loom out (it's only a matter of time!).

Isnt there a way of checking the AFM? The bloke i was talking to down the Ricoh said about checking the points in the multiplug with a multimeter for resistance. Cant remember what it was supposed to be though.

Not sure about moving the cam timing forward or back, not sure how far out it has to be before the valves and pistons strike... its pretty close as it is from what i understand!

The AFM doesn't seem to be failing mechanically - if it is having issues they are electrical.

With the top of the AFM removed you can see that the arm moves easily and well in response to different airflows...