initial running. very lumpy can-lambda keeps cutting out when trying to autotune

[fiestacosworth]

hi. sierra cosworth 4 cylinder engine. spec

siemens deka 875cc injectors, 7.1 compression ratio, bd14 cams inlet and exhaust, link monsoon ecu. t35 turbo, mac boost valve. all original ford sensors. Honda k20 coil on plug coils. 36-1 trigger wheel and standard cosworth phase sensor. car starts and runs. runs very bad. lumpy etc . and lambda keeps cutting out when trying to auto tune as soon as you touch the accelerator.

I have attached a log file of the car running and also with a small amount of throttle. and I have also attached the base map. if someone could give me a point in the right direction that would be great! need this car drivable "in and out the garage" asap!!

thanks. any more info needed please ask.

fiesta cosworth log.llg

fiesta cosworth base map.pclr

[Woody]

Hi fiestacosworth, just had a look at your log file  , i am no expert but i have done a few Cozzies recently , i did notice you have very inconsistent rpm and you have set the filtering to 4 in trig set up for both Trig 1 and trig 2 it may be filtering to high thats stopping it , also have you checked VR sensor polarity on both sensors , this is critical. Also have you checked the trigger offset with a timing light yet?. It might be worth setting up fueling with modeled rather than traditional . 

The one i am doing at the moment is using std cozzie pulley with the 4 teeth , i set it up with the RS500 trigger set up and it works mint , also using 630 Siemans Dekka injectors, i can ping you a start up map from that if it helps , but i am running slightly more CR than you.

 

HTH cheers Keith

[fiestacosworth]

Hi Keith. Thanks for the response. I had the filter set to 3 on both triggers. The trigger error count is just going up and up and up. I thought it may have been interference. But made no difference. Have checked the polarity and they are correct also. I had the loom “new loom” running on omex management. Run worse than this tho!! When swapping the loom over double checked all of that. Have also checked with a timing light and it is running at exactly what I have the timing set too. It’s strange it’s like it will run fine for a few seconds then jump all over the place making the rpm limit fault kick in. 

I have not tried the modelled fuelling only traditional. So can try that. 

Mom running a toothed belt drive kit from turbo sport that has a 36-1 trigger wheel. But if it’s that then there may be no point in running it.

this is my 1st experience with aftermarket ecus. “Gave up with the omex was awful” but I am not a novice around cars. I am a vehicle technician for Audi by trade. But unless I have spent too much time with the car and am over looking something. But I am very confused. It literally runs like a bag of crap. 

I have checked cam timing with dti gauge and that’s spot on. Don’t want to waste my time taking it too a dyno when it doesn’t even idle right. As could be something that I need to take away and repair / change before it can be mapped. 

Getting frustrating 

[integrale8v]

Have you modified the Phase sensor "trigger wheel"  This has 2 teeth stock.   

 There must be only 1 tooth on this with your 36-1 wheel and cam sync setup.

[cj]

I think this might take a couple iterations to get right. There are a few seemingly separate issues here.

First up the triggering - at idle your trigger 1 error count doesnt change, so it doesnt look like a tooth count problem or it would increase once per rev. I'd guess its either a really noisy signal or your triggering thresholds are set too close to the noise floor and when the noise creeps up, the filtering level isnt enough and the triggering errors come in. Can you take a trigger scope with it at idle? Until we see this its hard to tell if its interference or just signal level threshold calibtraion.

Second, its pretty rich - about 0.76 lambda most of the time so its nearly at the limit of too much fuel to run, if you're sticking with traditional model for now, try dropping your master pulse width until your idle afr is up to 0.85 or leaner - or at least until it sounds happy at idle. You may need to increase your cranking fuel encrichment if it wont start after doing this.

Third, your alternator doesnt look to be hooked up, you're seeing 11.4 ish volts at all times.

Forth, if you look at the status messages coming from the lambda sensor, your log starts with it at normal temp of just under 800C, but an error of "heated too long", just before it drops the lamba1 signal it reports a temp of 0C, then it looks to reset, begin the heating+diag process again. It never really recovers - the temp flicks up for a couple seconds at ~680 which looks right, but then 10 seconds later it says "RE over voltage" and says the temp is ~20C which I dont believe with the engine still running. I'm not 100% sure whats going on there but I'd start by checking your wiring between the sensor and the can lamba controller and the various power feeds to both are correct. I dont see any CAN errors at all so I dont think its the ECU>CAN Lambda side that is the problem.

[fiestacosworth]

Thanks for the responses. My phase still has original 2 teeth on it. Nobody told me I had to remove a tooth. If so what tooth do I need to remove ?  Is this why the rpm is all over the place ? And why the trigger count jumps up when I rev it ? 

I know it’s running rich. Can smell it and have leaned it up but it jumps all over the place. 

I also forgot to mention the alternator stopped charging. I know this won’t help at all. I did have a batt charger hooked up when I logged it. But voltage too low. And I guess this is why the lambda voltage fault is coming up ? 

Link monsoon doesn’t support trigger scope ? Apparently that’s what it says when I click on it ? 

I had no idea that the phase trigger needed modifiying tho. Any more info on this ? Thanks 

12 hours ago, integrale8v said:

Have you modified the Phase sensor "trigger wheel"  This has 2 teeth stock.   

 There must be only 1 tooth on this with your 36-1 wheel and cam sync setup.

 

[ClintBHP]

There is no need to modify the cam trigger works perfectly with 2 teeth.

[integrale8v]

2 hours ago, ClintBHP said:

There is no need to modify the cam trigger works perfectly with 2 teeth.

He is running 36-1 multitooth / missing then he needs to have only 1 home tooth.

 

If he had been running with the stock 4 tooth wheel it would be okay in the "Ford Cosworth rs500" mode.

[Simon]

with a 36-1 crank the cam trigger will 100% need the second tooth removed. It does not mater which one is removed but it is best to use the tooth that is furthest away from the gap on the crank as it is not good to have a sync and the gap happening at the same time.

 

Regards

 

Simon

[fiestacosworth]

I spoke to Clint at bhp and he said just set the triggers to ford duratec as that uses a 36-1 trigger and 2 point cam ? Rather than using the rs500 triggers. I tried this and it run better. Not sure what’s best 

[Simon]

Our Multi tooth missing trigger pattern expects a single tooth on either the cam or crank. If there are two teeth it will try to reset the firing sequence at both teeth.

You must only have a single tooth sync for that trigger mode to work.

The Ford Duratec VTC is expecting 5 teeth on the cam not 2 so I would not expect it to be happy either.

 

[fiestacosworth]

So that would explain the misfire/spit I get every 3 - 5 seconds on idle ? 

Ok so either put the standard pulley back on the crank. Or chop a tooth off the phase? 

Anything else you can see in my log and map that might be messing things up simon ? And any idea why my lambda is cutting out like that ? 

[Adamw]

Your misfire is definitely due to the trigger error, in the log you can see there is a 100% ignition cut being applied every time there is a spike in the RPM trace.

It would be nice to keep the 36-1 provided it is a reasonably well fitted trigger system and not something just hacked together and flimsy.  So assuming that the 36-1 is well made then the best option will be to remove one of the "phase" teeth.

I would say the Lambda is dropping out due to your battery voltage issue.   Get the charging system working and report back.

[fiestacosworth]

Thanks for the replies. I will remove one of the teeth on the phase. I will turn the engine over by hand until the gap on the crank pulley is lined up with the sensor and then remove the phase tooth that is furthest away from the phase sensor ? Or just remove either one ? Is that correct ? And will also sort the changing issue out too and then perform another log and see how it goes. 

[integrale8v]

Edit: remove the phase tooth that is closest the phase sensor.  And report back when you have tested some more.

[fiestacosworth]

Ok so removed a tooth from a spare dizzy I had and fitted it to try the way you guys are saying. I also sorted the charging issue. Although I forgot to hook it up for the log!! DOHH!! However still reading 12.5v etc. So logged it and it runs very lumpy and then for no reason on its own just went really lean and cut out please see the attached log and advise. Thanks! 

new log fiesta cosworth.llg

[Woody]

How have you got the map sensor wired and piped up or which one are you using , the log shows between 6-7psi positive pressure , have you calibrated the map sensor in the software. 

Can you do another log with injector opening times and lambda ?

 

 

[fiestacosworth]

Map sensor is built into the ecu. Monsoon ecu. And it’s plumbed from the plenum straight to the ecu. 

I set up log and seletced select all.. I thought that would have put everything in. Will have to do another one tomorrow now. 

[Woody]

Ok,  that is the way to do it and the ecu works really well with a cozzie , can you watch the map reading, ignition on engine off and see what its reading , if the axis etc are all ok it should read 100kpa ish untill the engine starts then drop to 35 kpa ish. Also good way to check map sensor is with a vac pump , pull vac and pump pressure through both ranges and see if it matches the software, use the tuning page if you have guages configured or the runtime window..

[integrale8v]

How is your trigger filters adjusted?  If they are same as in the map you posted.  Set them lower.

  The MAP in the log seems okay for me, between 45 and 50 when idling.

[Adamw]

I'm not sure what woody is on about, your MAP is set up fine and appears to be reading correctly with about 50KPa at idle. Lambda and PW is in the log too.  The main problem I see is you still have trigger errors.  Perhaps the crank sensor is wired backwards?

[fiestacosworth]

It’s strange because the time I run it before I logged it the trigger count stayed at zero and didn’t go up. Then the time I logged it it turned over a bit error count went up then it started. 

[Woody]

have i downloaded the right file i cant see none of that?

[ClintBHP]

Ok to state again there is no need to remove the second tooth at all, I have loads of cars running 36-1 and 2 tooth cam, spinning to 10k RPM with no errors, its exactly what all WRC Escorts ran.

You can't set this up on custom trigger as it will error on the second cam pulse, but the Duratec setting seems to not reset the IRQ for the trigger 2 until the missing tooth is seen again so in effect only counts the first cam tooth after the missing tooth is seen.

You are more than welcome to grab support direct with me, instead of playing guessing games with people who have probably never set up a YB engine on a LInk ECU (no disrespect to anyone trying to help)

Direct email: [email protected]  - 01474 850666

Attached a picture of a car that is 280bhp without a turbo running exactly this trigger setup !

[Adamw]

4 hours ago, ClintBHP said:

You can't set this up on custom trigger as it will error on the second cam pulse

He has removed the second tooth as advised by us.

 

4 hours ago, ClintBHP said:

I have loads of cars running 36-1 and 2 tooth cam, spinning to 10k RPM with no errors, its exactly what all WRC Escorts ran

That is irrelevant unless these WRC escorts used a Link G4+ ECU.

 

4 hours ago, ClintBHP said:

but the Duratec setting seems to not reset the IRQ for the trigger 2 until the missing tooth is seen again so in effect only counts the first cam tooth after the missing tooth is seen

You should not be making suggestions like this when you have no idea how our trigger decoding works.  This user had already tried duratec mode with massive errors.  You are correct in that the trigger 2 teeth are only observed during certain windows of crankshaft rotation but you dont know where these are and since the user fitted his own trigger wheel you also dont know where his gap is in relation to TDC or the phase teeth.  You may have got lucky on the one you fitted but it is very unlikely to work reliably for every 36-1 YB install.