AAN buggy engine overboosting/backfiring

[Stefen]

We are having a few issues with the buggy engine. The main problem seems to be overboosting. I did a series of runs and loaded the engine, with the VAG Com scope going.

On the first graph here, the boost went up to 1.8 Bar at high revs, and as the revs/boost went up, the TDC value went from -44 to -8 @ 1.8 Bar. There is a simliar pattern for every time it boosts.
When under no load the ingition is retarded as the revs increase.

The engine will backfire at anything at and over 1 Bar. There is a new fuel pump in it allegedly capable of 500+ bhp. The fuel pressure gauge shows that the fuel pressure is okay. It has a tweaked version of homefries in the ECU.

The boost is should go to 1.5 Bar, and feather off to 1.0 Bar at high revs.

The TDC value usually stays around -20, but sometimes it goes to -10.

1.8 Bar Banging


1 Bar Banging


0.5 Bar No banging


0.5 Bar No banging


No load


The Buggy!


The pipe going to the ECU is free from obstruction and leaks as far as we can tell. I have also worked the N75 manually from the ECU plug and it switches. There doesn't seem to be any boost control at all.
When the N75 is disconnected the boost goes to circa 0.6 Bar.

I don't really know what to look at next, except to plug the ECU into another car to see is the behaviour replicated.

[Tonks4x4]

Is it boost cut you are getting?

It sounds very much like a boost leak, do a leak test using a tyre gauge.

[Stefen]

How would I know it's boost cut or not?

Good idea, must bring the compressor around and test it throughly.
It does sound like a boost leak, but I have connected a short pipe from the inlet manifold to the ECU (maybe 2 or 3 feet) and still got simular results. That's not to say it isn't leaking somewhere else or around the joint.

[prj]

When the N75 is disconnected the boost goes to circa 0.6 Bar.

That's your problem right there.
It should be 0.4 bar - you have a stiffer spring, so no wonder your boost goes up too high.
Fit the correct spring in the wastegate or unscrew the bolt.

And it's weird that your timing jumps between 44 and 8. It should not really get to 44. But your buggy is very light, so parts of the map are used that would normally never be used.
8 degrees timing sounds about right on a knocking engine through torque peak, but - it means it's knocking. I wouldn't try this much boost on stock turbo.

As for loud banging and flames out of the exhaust - that's because there's no ignition.
The only time that happens is if your plugs are bad (and if you do 1.8 bar often, your plug electrodes will fall off), your ignition system is bad or the ECU resets/loses signal, but since vag-com seems to be operating, that doesn't seem to be the case.

[Tonks4x4]

Hope you get it sorted it's an awesome looking buggy have you got a build thread anywhere?

[Stefen]

That's your problem right there.
It should be 0.4 bar - you have a stiffer spring, so no wonder your boost goes up too high.
Fit the correct spring in the wastegate or unscrew the bolt.

And it's weird that your timing jumps between 44 and 8. It should not really get to 44. But your buggy is very light, so parts of the map are used that would normally never be used.
8 degrees timing sounds about right on a knocking engine through torque peak, but - it means it's knocking. I wouldn't try this much boost on stock turbo.

As for loud banging and flames out of the exhaust - that's because there's no ignition.
The only time that happens is if your plugs are bad (and if you do 1.8 bar often, your plug electrodes will fall off), your ignition system is bad or the ECU resets/loses signal, but since vag-com seems to be operating, that doesn't seem to be the case.

Re: Wastegate spring - I thought it was the case that, as long as the mechanical limit set by the wastegate spring, was below the electronic limit set by the N75, then the N75 would be able to operate correctly. This would mean that the N75 always has enough boost available to override the wastegate spring.

Re: Timing - The loading was done via the handbrake. Sometimes you see a dip while rising to the main peak. This is me varying the load and throttle to maintain a constant boost level (as possible!).
Is it correct to reduce the retarding as the boost/revs build?

Re: Banging/flames - The plugs have only done maybe 10 miles at the most and it it rare to go to 1.8 Bar, and it is very brief as well.
I have read about common problems with the ignition system, but as at this point I think the N75 is not functioning correctly (need to check for leaks and adjust spring yet). I was thinking along the lines that the ECU is not reading the correct boost, therefore not opening the N75 and not giving the engine enough fuel, so running too lean was causing the the banging?

[prj]

Re: Wastegate spring - I thought it was the case that, as long as the mechanical limit set by the wastegate spring, was below the electronic limit set by the N75, then the N75 would be able to operate correctly. This would mean that the N75 always has enough boost available to override the wastegate spring.

No, you are completely wrong.
Motronic operates off WGDC, which is corrected for by adaptation. This adaptation has it's limits, especially if WGDC is set high already - it will overboost. Fit a completely stock spring and unscrew the screw - your chip is not meant for a stiffer spring.

Re: Timing - The loading was done via the handbrake. Sometimes you see a dip while rising to the main peak. This is me varying the load and throttle to maintain a constant boost level (as possible!).
Is it correct to reduce the retarding as the boost/revs build?

I suggest reading a book on FI engines, as everything can't be explained in a few sentences on a forum.
That said - 8 degrees advance means strong retard due to engine knock, even at torque peak on this ECU.

Re: Banging/flames - The plugs have only done maybe 10 miles at the most and it it rare to go to 1.8 Bar, and it is very brief as well.
I have read about common problems with the ignition system, but as at this point I think the N75 is not functioning correctly (need to check for leaks and adjust spring yet).

I am sure your N75 is working fine.

I was thinking along the lines that the ECU is not reading the correct boost, therefore not opening the N75 and not giving the engine enough fuel

Wrong. Engine fuelling is not affected by the MAP sensor at all on this ECU.

so running too lean was causing the the banging?

No, banging happens when there is no ignition - fuel is dumped out of the exhaust valves where it contacts the manifold and explodes.

[Stefen]

Great, that has really helped to build up a picture of what is happening. Just a couple of more questions:

a) What is the purpose of the boost sensor if its value is not used for fueling? Ignition?
Surely the more the boost, the greater fuel is requried? The ECU is hardly running open loop with regards to the fueling?

b) So it sounds like the banging is caused by a faulty coil or POS? I attempted to diagnose which coil it was last night, by disconnecting one injector at a time. The boost wouldn't build up to one Bar (where it starts to bang) as the turbo may be too big, so I couldn't diagnose that way.
I guess the only other way is to swap in/out known good coils and POS until the banging stops. Unfortunately we don't have any! Is there another way?
I have also read about using 1.8t coils which seems a viable alternative.

I will definitely get a book to learn some more about this!

[Jamo]

Great, that has really helped to build up a picture of what is happening. Just a couple of more questions:

a) What is the purpose of the boost sensor if its value is not used for fueling? Ignition?
Surely the more the boost, the greater fuel is requried? The ECU is hardly running open loop with regards to the fueling?

b) So it sounds like the banging is caused by a faulty coil or POS? I attempted to diagnose which coil it was last night, by disconnecting one injector at a time. The boost wouldn't build up to one Bar (where it starts to bang) as the turbo may be too big, so I couldn't diagnose that way.
I guess the only other way is to swap in/out known good coils and POS until the banging stops. Unfortunately we don't have any! Is there another way?
I have also read about using 1.8t coils which seems a viable alternative.

I will definitely get a book to learn some more about this!

Motronic system uses the MAF aka Mass Air Flow to work out how much air is going into the system. Map sensor is for boost control.

1.8t coils i'm not a big fan prefer this kit or the 2.0l coils http://www.efiexpress.com/catalog/pr...products_id=18

[prj]

a) What is the purpose of the boost sensor if its value is not used for fueling? Ignition?
Surely the more the boost, the greater fuel is requried? The ECU is hardly running open loop with regards to the fueling?

Fuelling and ignition is calculated according to the MAF sensor. You have that connected, right?
MAP sensor is only for boost feedback.

b) So it sounds like the banging is caused by a faulty coil or POS? I attempted to diagnose which coil it was last night, by disconnecting one injector at a time. The boost wouldn't build up to one Bar (where it starts to bang) as the turbo may be too big, so I couldn't diagnose that way.
I guess the only other way is to swap in/out known good coils and POS until the banging stops. Unfortunately we don't have any! Is there another way?
I have also read about using 1.8t coils which seems a viable alternative.

Use 2.5TFSI ones....
And what do you mean by "the turbo is too big"? Are you using chips for stock turbo and fitting a bigger one? In that case your boost will be too high as well! You need a remap when fitting a bigger turbo and also bigger injectors.

[Stefen]

Yes, the MAF sensor is connected.

I mean that there is not enough exhaust gas produced by 4 cylinders to spin the turbo to 1 Bar. I will try this again to be sure.
The chips were adjusted for a bigger turbo. The injectors are standard, but we are thinking about using bigger ones.

Thanks for all your help guys!

Here is an early picture of the build. I should really do a project thread.

[prj]

Yes, the MAF sensor is connected.

I mean that there is not enough exhaust gas produced by 4 cylinders to spin the turbo to 1 Bar. I will try this again to be sure.

Yes, possible. Always unplug the injector as well though!

The chips were adjusted for a bigger turbo.

By who? How? What turbo are you running?

The injectors are standard, but we are thinking about using bigger ones.

Will need a remap.

If you keep running the engine like that - excess of 1.5 bar boost etc on stock injectors with turned off limiters, you /will/ blow it up.
Knock management alone without other limiters in this ECU is not sufficient to save you.

[Stefen]

There is a tweaked version of homefries in the ECU, I'm unsure what the turbo is.

Yes, it will need remaped again for bigger injectors, but we need to have the engine running well before any more upgrades.

Don't worry it is only a very brief squirt, but I appreciate the warning.

[prj]

Who did the tweaking? Paul?
There are very few people in this world who know how to map M2.3.2 boost chip.
Plus, you need to do test runs on the car, you can't map the turbo by guessing.

Has the MAP sensor in the ECU been changed? Or is it the stock 250kpa unit?
If it is the stock unit, you should not be going over 1.4 bar boost at all.

[prj]

Also - how can you not know what the turbo is? Read the numbers off it, there must be something written on it...