Misfire with mapfilter

[z0tya]

Hello!

I have a strange misfire problem with 1.1.74 after upgrade from 1.1.27.
Engine:
Cosworth yb with zetec wasted spark coil, 36-1, semi sequential, spark gap around 0.8 mm ( but tested with another plug with 0.6)
First run with mapfilter settings:
Triggers Num For MAP Filter :1 MAP Samples Per Trigger: 4
The car has a massive missfire on boost:



Next run with mapfilter settings:
Triggers Num For MAP Filter :1 MAP Samples Per Trigger: 1
The car has no massive but map fluctuate very much, and lambda too (because of wrong measured map -> false Ve -> false pw I think)



Trying to raise dwell not help, with more samples per trigger (suggested 10) wrong to.
With 1.1.27 there was no similar problem, map in a log a bit fluctuate but no misfire pull hard.
Logs:
http://cosworth.hu/misc/sapphire/v3.3_n001244-2010.10.06-15.38.36.vemslog
http://cosworth.hu/misc/sapphire/v3.3_n001244-2010.10.06-15.49.38.vemslog

I have no idea
We went back to 1.1.27 while no solution for this.

Can somebody help?

[mattias]

The MAP fluctuations while on full boost are typical for a wastegate valve that is binding. I would take apart the wastegate and check the valve and valve guide - without the spring the valve should run freely with little friction, ream/hone if necessary to get proper clearance. Imagine that it may bind when the components are hot. Use copper grease and lube it up.
Of course, it could just be a badly working boost control - remove all of that junk and just run with the spring and hose from the intake to the bottom part of the wastegate actuator, that will determine what's causing the fluctuations which probably can't be cured by software.

Boost is building very linear too, that is not how it usually works. Do an intake leak test. Plug the compressor inlet, pressurize the system through the fuel pressure regulator signal hose, and with WOT.

I can see that it is hard to tune the engine with these problems, but at least approach those boost levels with a richer mixture and lean it out, rather than the other way around. A turbo engine can also mis-fire and stumble if run too lean, especially at low speeds.
Set max/min for lambda and lambdatarget to 0.7 and 1.1 to get a better resolution, or use VemsTune log viewer which already does this (open log in other frame).

I would use most recent firmware, which is 1.1.81, it has proven to the best 1.1.x for me. It is very good at revealing if you have trigger errors.

[GintsK]

May be your w-gate solenoid tube shares with MAP hose?? Where MAP is connected?
Log shows strange boost build-up as Mattias mentioned. And unusual fluctuating at this period too.

What type of intake you run? What turbo? Cams? Nothing exotic there?

[z0tya]

Thanks Guys!

Good hints! I totally forgot to approach the problem from the hardware side, because of there is no problem with 1.1.27. But your replies remind me what will maybe wrong with wastegate piping. There is manual boost controller with a t pice between wg and turbo outlet, and the pipe from the t to the manual valve about 1 m length.
The turbo is changed from oe t3 to a t3/4, not a garrett but not cheapest chinese, between them in price and quality. I think 1 bar boost at 3000 rpm not bad. the logs are in 4th gear.
Map is connected in the intake plenum, as original, no other branch form this pipe. Go to the ecu directly.
The mixture maybe would be richer in this transient range (after the misfire rpm range the car pull normally.
The engine is standard with original plenum, standard cams, no other unusual. Similar this:


I will check yuor tips!

(but I have no answer why was good with 1.1.28    )

[GintsK]

May be missfire cause is overboost protection? In log file we see just part (just ~17Hz) what happens in ECU. Your log file show negative MAP spike. But it can be positive too in ECU.
In any case your MAP signal is terrible!
Isn't possible to relocate MAP sensor hose?

[z0tya]

Overboost set to 300, and the pw no went to zero (at first logs it was lower so this was the first what i switch off).
The map sensor place is original, all of cosworth use it. It must be good 




Can I try some damping in the map line? (filter and hole), maybe?

We will do some mod in the boost control piping , and make some new log next time...

[mattias]

You should not need any damping on this signal, and that will normally only cure problems with pulsing at idle and low load, your problem is at high rpm/load.

I'm 100% sure your problem is mechanical, keep it simple. Remove the boost control and just run a single signal line from the intake to the wastegate. No valve or bleed, just a straight line. If it keeps pulsing at high load still then you definately need to take the wastegate apart.

I have done one install on this engine before (4wd Sierra) but we used an exhaust manifold made by the owner of the car and a larget Holset HX40 turbo. It was as strong in the bottom-end as stock but pulled to 7000 rpm now, the "330hp" kit injectors were bottomed out instantly as it gave much more power than stock at same boost. If Cosworth made the engine with these components it would not have been easy to make it street legal

[z0tya]

Right Mattias! I will make a test with minimal wg piping.

Thx

[z0tya]

We made some test with minimal piping to wg actuator, no bleed valve. Fw remain 1.1.27 and car have problems, but I think no software. The actuator only 0.8 bar type so need to tight to much, no remains enough rod lenght to open the wg correctly. Only 7-8mm so I think the control of the boost pressure is not too good.
Bigger problem is missfiring at lower rpms. I have an interesting log. There is a 1.7 bar boost at 4400rpm nice acceleration no hesitation no missfire, but after that in 5th gear at 1 bar but only 3000 rpm the car is stumlbling.
Maybe surge?
www.cosworth.hu/misc/sapphire/v3.3_n001244-2010.10.21-20.20.29.vemslog

[GintsK]

Very interesting case! especially at 109s. Turbine surge? Didn't you heard it? But how it can be misfire cause?
When I had turbine surge logs looked little different. My entry MAP line was clean except surging area. At surge area MAP fluctuations frequency was higher, but amplitude lower. And no misfires of course.

However can you try to relocate map hose to more quite place?

[z0tya]

I couldn't try relocate the map source yet, its difficult a bit, dissassembing the plenum, and making a new connection. And all off map source (fuel pressure, bov, crankcase ventillatation, map sensor) are there as seen on the manual. This is the original place, all of cosworths use it. (But it is true, on original engine the map hose only 20 cm lenght, now about 120cm, because the ecu in the dashboard.
We will try a cooler plug, and make a leak test on weekend.

[z0tya]

We run a test with colder sparkplugs (Champion qc59c), gap 0.65mm, then swapped ignition coil, spark cables but no positive result. I have log with two acceleration after each other, the first is okay ,the second is bad.
I don't understand.
 I have no idea what is wrong.


This is the log:
http://cosworth.hu/misc/sapphire/datalog201010301225.xls

[mattias]

Was this log made with old 1.1.27 firmware ?

I would sort out your problems at lower boost, if they stop at lower boost at least you know it's related to the load on the engine and electrical system.
The ignition advance is pretty low and you're running borderline too rich unless the ignition system is up to it - that could produce a lot of strange behaviour too. That's also a reason to lower the boost and tune it better there, to get an idea of the timing and fuel mixture that the engine wants. I know, you're saying 1.1.27 works and 1.1.74 does not.. I'm not sure it's that simple, it's hard to get all the facts over the internet.

[GintsK]

Average lambda around strong misfire area is 0.8. Not so rich. But Mattias is right about retarded timing. Sometimes it cause misfires. Is it proofed as reasonable timing for this engine? Can you try little more advanced?

[z0tya]

Mattias: Yes it was made with 1.1.27. When we started to upgrade fw to 1.1.78 with the new turbo, I tought that is a firmware problem because of the new map samling strategy, the went back to 1.1.27 and found that this problem is apperared with 1.1.27 too. So the topic subject is misunderstable but I can't make a new topic for it.
The situation is this: The car was mapped with an old t3 turbo up to 1.5-1.7 bar boost and worked fine, but the turbo was small. So we went to a t3/4, and think we try newer fw with mapping for the new turbo. Then happened a misfire story, what i thinked because of the new fw. So we went back to 1.1.27 and tried test. The result is above.
The has strange missfire (surge??) at about 3000rpm when the boost lower, but fine above this rpm with much higher boost.
When you see the first pic there was leaner mixture, but mattias said "approach those boost levels with a richer mixture and lean it out"
What map area seems to retarted for you?