Big ABF problems (Update 15/5... FIXED... or so I thought!!!)

I'm really hoping this is the problem (I can't think of any other reason!).

I'm away until Saturday morning so will suss this out on the weekend. If this is ok, though, then I'm officially out of ideas

 

Might be worth shot then. Will get one this week

 

Sorry, forgot to add - have changed this too, but still no difference

 

toyo, you stated: "the vaccum line inside the ECU is compromising the signal to the internal MAP sensor. You may want to check this.
At throttle openings higher than much idle you will be close to 100kpa (WOT) anyway and the car will appear to drive ok"
can you explain better how can i check that? do i need to open the ecu box and measure the map sensor pins while the engine is running? im asking cos i have the same problem, it runs fine for a couple of seconds than the engine visibly rocks, i ve always knew its sth to do with the vac. i have double checked almost everything mentioned in this thread. if i finally give a solution to this problem, i'll send a beer by DHL to everyone of you!

 

ECU relay is #30, but I'll try another one just in case

Thanks for those pin outs - will check them on Saturday!

Just pull the board out of the ECU, unplug the vacuum line from the barb to the MAP sensor and check there's nothing stuck in it.

Will check this too!

 

P.S. I've thought of something else, which I'm really thinking might be a major point in all this

The ECU I'm running was driven with a turbo installed for a few minutes (long story) and the MAP quite likely saw positive boost pressure a couple of times. From what I understand, this could potentially kill the sensor (??)

Every load sensor on the engine is new, except the MAP, so this surely is the only problem... I need to find a spare ECU, which is an 024BE and has the same Siemens p/no...

What does everyone think? Can positive pressure kill the MAP, or at least reduce its functionality?

 

you can sort of track what the map sensor is doing in vagcom.

on measuring block 4 there is a binary block.

1000 = over run
0100 = idle
0010 = part throttle
0001 = full throttle

 

From memory, the standard MAP sensor can handle boost to a limited amount after which the range measurable reaches maximum i.e measure up to 1.5 bar of boost over atmos and thats it. But I don't know if you can damage the sensor from seeing too much boost.

Matts idea is good to see if your getting approx the right value.

Gurds

 

ABF ECU internal map are rated 120kPa or 0.2 bar positive manifold pressure max.

Yes you can try another ECU and see if that works.

 

These are more 'engine state' related, which would result from a combination of inputs but not from the MAP sensor per se.

 

when i had my ecu open the other week i saw the map sensor was connected to the board by 3 solders.

5v, ground and signal

ive seen a few after market 120k maps that are 3 pin also, im sort of tempted to try and stick a new one on the board

 

http://kms.vankronenburg.nl/download/PDF/KMS pressure sensor 75-120 kPa.pdf

im tempted

 

The only thing I can think of which gives lean AFR's but yet fowls the plugs is unbunt fuel ie low compression! You say you got this engine of EBAY with a KR head and cams. Maybe the 'cams'(chain end) are not timed up right. Have you done a compression test?

 

When I first started the car up, there was a "MAP sensor out of range code", which I cleared and it never came back... Perhaps there was something more to this I'll try to track down an ECU...

Yeah, true. Compressions all ok - the plugs are sooty and not wet, though, so they are firing. I can only think that perhaps the mixture is all over the place and running rich occasionally, which is causing them to foul up.

It's definitely time to try a different ECU me thinks

 

That would make sense if the ECU was used in a boosted application previously.
It would be interested to get a direct read from the MAP sensor when the engine is running as already suggested. It should measure around 35kPa at idle for a standard engine. If for some reason the internal MAP membrane is damaged it could be reading closer to ~ 85-101kPa or simulating WOT conditions and cause the engine to overfuel during cruise or idle and cause the lambda correction function to clip in its correction.
This is possible if the stock ECU was incorrectly used in a boosted application.

 

Thanks Toyo. Fantastic info as usual

Have ordered a secondhand ECU from you chaps in the UK, so I'll report back as soon as it arrives.

In other news, I've found that 60psi (instead of 45) of fuel pressure makes the engine run almost perfectly

 

what do you mean matt? the ecu just reads this 4 situations from the map?