Alpha-N
The long story in short: I bought a kit from UK called MNR Vortx to build a Lotus 7 like car. The car is a tubular frame based with plastic covers font engine and rear drive. The drivetrain are follows: the Duratec HE 2.0 L engine (Ford Mondeo), Ford Sierra type 9 gearbox, custom made propshaft and Ford Sierra diff.
The engine is a pretty standards one with a little modification. Due to the rear drive the engine mount was changed from cross to along that cause a change of the oil sump. https://www.raceline.co.uk/products/part_section.asp?categoryID=1%20&SectionID=22
The engine was refurbished new rings, gasket, the head and valves were cleaned, all sensors are original Ford. The standard flywheel was changed to a lightweight one (approx. 5 kg)
In order to achieve around 200 HP the only change I made was the throttle body. I used the caterham type roller throttle body http://teamterrificracing.net/TitanNorthAmerica/RollerBarrel.html, with a larger injectors 237 g/min approx. 347 cc/min 14,5 ohm.
Because of the throttle body the EGR port was blanked off. The engine loom is a custom made done by Lorez garage.
In Hungary the car can not be registered so I need to take it back to UK where it can be done via special IVA test. During the test the emission is one of the crucial point, the engine must pass the original emission level.
CO 0.5%
CO2 14.5-16.0%
HC 100ppm
O2 0.1-0.5%
CO content at increased idle speed 0.3%
Increased idle speed for CO test 2800-3100rpm
Lambda 0.98-1.02
Setup are as follows:
Duratec HE 2.0 l
Vems V3 1. 2. 38 firmware
Vems tune 1. 5. 57
Titan roller barrel with Bosch 347 cc/min Injector 14,5 ohm
Adjustable fuel pressure regulator set to 3.5 bar
- Idle is controlled by PWM valve and 4 ID4mm small tubes connected behind the throttle body
- this hopefully equilises the charge between the 4 cylinders at idle
- but the 4 throttles are (theoretically) carefully tuned for equal charge;
- but VE is still very low at idle (15 or so), lower injector opening (max 0.7msec at 14V) will help to have slightly higher VE at idle
- MAP might not show much vacuum at idle, but could still be interesting...
- watch individual power at cruise and idle, maybe shows something... Some cyl might show symptom, swap with other cyl...
- add more fuel to 1 cyl at a time, as an experiment...
- perhaps unequal air/fuel between the 4 cyl; or some ign problem....
Sensors are as follows:
Crankshaft sensor standard Ford
Camshaft sensor standard Ford
TPS sensor standard Ford
Water temperature sensor OEM Opel Astra
Knocking sensor not in used
The control is RPM vs TPS with barometric compensation.
Engine is up and running with a basic configuration focus on the hot idle now. The throttle body was balanced so all four cylinders got the same amount of air. The biggest problem I’m facing with is that once the engine is at the normal temperature the idle is stable but I can not achieve the stable lambda.
- Modifying the VE table eg from 28 to 29 unit cause abnormally large lambda value changes.
- NOTE: if req_fuel: 9.8=>5.8 msec then 1.69x VE values (147 => 249).
- Also check fuel calculations view, perhaps some suboptimal injector opening settings makes it hard to tune
- After the second try we realize that the ECU controls the radiator fan, and when the fan switched on the lambda moved away and returned back after couple of minutes, but still stable lambda is difficult to achive. The log mode was on so I enclose the log I gather during the setup.
Thanks,
Peter
Quick review
- Basically adding an idle motor air feed on alpha-n without iac2llad compensation is a no-go
- for stable idle, just disable idle air altogether, block off and adjust the main throttle slightly that should stop the hunting
- Use minimal ignition based idle control ramp to keep rpm stable (warm)
Response
Do you mean to switch off the idle PWM valve completely? Using the valve is to have a cold start, in that kind of weather like today it wont start without additional air.
Ok i'll try it.
Response2
I modified accordingly and now lambda changed smoothly but the idle control was much worse. Any electronic changes like radiator fan switch on or headlight switch on caused a huge swing on lambda.
See log files.
Finally I set it correctly and the car was sent to UK for registration. The engine was up and running till yesterday where the car can not be started. Vemtune was connected and it was clear that there is no rpm at all. it 0 during the start. The sensor was checked and it's between 430-560 ohms.
Hopefully they will check at the and of the cable as well.
Any help would be appreciated.
No spark
The picture little bit changed and we got some info that rpm works fine, tps sensor also but still no spark at all.
Thanks,
Response DB:
If rpm is showing normally in VT, but spark is lacking; i would direct the search towards coil pack power:
- perhaps a fuse blown
- or some power relay not engaging?
In any case it is possible to trigger spark for testing in VT via Base Setup->Ignition Output Visual dialog, set Testting mode to ON; Spark for each output can be triggered by pressing the Test buttons (per output)
Best regards, Dave
Thanks for the reply,
Working towards today found out that there was a spark but not so strong and another strange thing appeared.
When we started the process the pulse with was around 34- 35ms but as the time went by it moved to 110 than 130 and when we finished the test it was 134 ms but nothing has changed as we still can not start the engine.
See picture. also enclosed the config file.
Response DB:
Your logfile directory only contained a recent config (17/10/2019) but no vemslog, please upload a vemslog and triggerlog of engine cranking 8-10 seconds without INJ and IGN fuse so i can review the basics. We'll work from there.
Best regards, Dave
Now after 4 new plugs the engine started but the cold start was very bad after the engine got warm up it was fine but HC was still high. I try to get a logfile today as everything is done in UK so I got the information remotely.
I uploaded 2 new log files and it seems that It's a way too much fuel at the cold start. They tried twice the morning I as they said it was very rough at the start but when it got warm than it was ok. Now they wanted to start again but did not start. Took the plugs out were totally wet. Dried and cleaned them and It was fine.
Any help would be appreciated
Rgds,
Peter
Response DB:
I have reviewed both logfiles: v3.3_u006146-2019.10.21-12.37.04.vemslog and v3.3_u006146-2019.10.21-12.37.55.vemslog, no bad startup behaviour present there.
From visual review of ecu data (the log files were not long enough for the wideband sensor to heat up so cannot judge any mixture related) i think your cranking enrichment is set a bit too high.
To help startup with current cranking mapping (and change for the better):
- use *flood clear* function if the engine is flooded (or appears), press full throttle during cranking than slowly release throttle; see: http://vems.hu/vt/help/v3/v3_cranking.html
- reduce cranking enrichment curve in all cells in sub 80degC area's (half it), so 44degC becomes 150->125% etc.
- enable ego closed loop control, by setting ego step size to 1 (from zero), see: http://vems.hu/vt/help/v3/v3_ego_control.html; you will never pass any HC or CO testing without it enabled.
Best regards, Dave
I've changed according to your suggestion. It's much better now but still too much movement on Lambda till it gets warm. At fast idle 2500 pm lambda still high. New log file was uploaded. Thanks for your help
Response DB:
I have reviewed your logfiles: v3.3_u006146-2019.11.15-19.51.26.vemslog
Engine startup is now good, but you have set ego minimum temperature for enable is 84degC, that means ego control is not active during warmup at all. In the end of the log (around 15min) you can see that when ego control enables (at coolant over 84degC) it tracks lamba target perfectly.
To resolve cold start lambda tracking, please lower ego control-> minimum temperature: from current to 84degC to something like 10 degC.
I hope this solves your issues,
Best regards, Dave
I've changed from 84deg C to 10 degC but the lambda gets too high till the engine gets around 82-83 C then It's fine. Any help would be appreciated.
I tried to solve the idle lambda issue and check everything and switched the trigger log on, and found that the signal was missing randomly but the engine run stable but it was heard the ignition was missing. The cam sensor cable was checked and worked fine.
Trigger logs are uploaded.
I've change the baud rate but no difference and cable signal was check by oscilloscope and it's fine.
Any help would be appreciated.
Steps:
- split to MembersPage/MnrVortx/TriggerInputs (required for support, because this page has too much unrelated info)
- ECU serialnr
- ECU sectrig input pin (EC36/13 ?), and input type VR or HALL (HALL ?) ? Since inputs are tested before ECU is sent out, this is always the first suspect.
- voltage measured with known resistor, eg. 1k pulldown to GND ? (1.35V ???)