Hi everybody,
The car is running, and I have made my first meter on road !!
The engine idling like a charm ,and acceleration at mid-load seems good ïŠ.
Now it’ s time for me to make fine adjustement….
My first question is why can’t I lean the idle air/fuel ratio lower than 12.25 ?
On my VE fuel table the values are near to zero,an exemple :
When idling,at +/- 900 rpm,the table value cell is « 2 » and mid-load about « 45 ».
And when I look on table from other guy,this values are much more high than mine
Where did I make a mistake ??
Can anybody help on the acceleration and cold engine enrichment % table ?
Thanks…..
Please publish a zip file with .vemscfg or .vemslog.
Are you running Speed-Density fuel algorithm?
My table use boost pressure /rpm...
Michele.
Quote from: rossi46numberone on February 24, 2011, 04:02:30 PM
My table use boost pressure /rpm...
Michele.
No, under Base setup select Engine setup. What is the number on req fuel field? and VE table control strategy is what?
fix the problem :D
it was the injectors trim value ,set to 100%.
decrease it to 75 % and my table look pretty good to do some fine adjustement!!!
At the same time the EGO correction issue solved and correction turn now around 97-99% at idle....
My req fuel is set to 5 ,Dwell :2.75.
Have tried the car on road pushing to 140 kpa and the car run fine!
Now I'm on the right way,thanks a lot guys for your precious advice and suggestion ::).
See you on next step:exhaust gaz temp and barometric correction.
I have also injector trim @ 100%, req fuel is 7,8 and VE table is @idle around 80.
So something else is right now richening a lot, because your req fuel should be something like 6,8 (you had 440cc injectors?).
The best thing is to post us the config file and we can help you a lot better.
Quote from: rossi46numberone on February 25, 2011, 04:59:50 PM
fix the problem :D
it was the injectors trim value ,set to 100%.
decrease it to 75 % and my table look pretty good to do some fine adjustement!!!
At the same time the EGO correction issue solved and correction turn now around 97-99% at idle....
My req fuel is set to 5 ,Dwell :2.75.
Have tried the car on road pushing to 140 kpa and the car run fine!
Now I'm on the right way,thanks a lot guys for your precious advice and suggestion ::).
See you on next step:exhaust gaz temp and barometric correction.
I would say thats a band aid fix that's not addressing the real problem. You shouldn't have to touch the injector trims at all, it all indicates that your req_fuel is too high. Try putting the injector trims back to 100% and increasing req_fuel value by 25%
ok,here is my config link file:
http://www.box.net/shared/vc3vxe3yvb
thanks....
Yes I run 440 cc injectors@3 bar ,with 3.5 bar they probably give +/- 500 cc.
I get rquired fuel at 6.2.
440cc injectors compensated for 3.5bar gives me 475cc, not 500cc.
EGO Correction step size is abnormally huge at 30%, should be at least 1% and no more than 5% in most cases.
Unless you are using ITB's with a turbo, MAP/RPM Correction table should all be set to zero.
What do you mean when you say ITB's?
Thanks for analysing my confiq and for these remark.
I'll try modiying this place et give you a feedback asap....
Michele.
What is the purpose EGO Correction step size?.
Quote from: rossi46numberone on February 25, 2011, 08:14:20 PM
What do you mean when you say ITB's?
Independent Throttle Bodies
ITB - Individual Throttle Bodies - One throttle body for each cylinder - 4 cylinder engine, 4 throttle bodies.
EGO Step size is the percentage of the active VE value, per step of EGO correction. Basicaly, with a step size of 30% any EGO correction with cause a change in fueling of 30% meaning huge swings in fueling. Look at it this way, if your VE table is out meaning there is an error in fueling of 10%, any EGO correction will result in an actual change of 30%, putting the fueling 20% beyond where it really needs to be.
Setting EGO step size to 1% as is suggested in the F1 dialogs means that EGO correction will only change in steps of 1% requiring 10 steps to get to the desired 10% correction, where as with 30% step size, 1 step would result in 30% change in fueling. 30% is just so wrong.
Actually I have just reallised. MAP/ RPM correction table should all be 100%. Everything for fueling is biased around 100% so 100% MAP/RPM correction is actually zero correction with 75% being subrtactive correction of 25% and 125% being an addative correction of 25%
This also stands for the injector trim you have set to 75%. This is not strictly correct, and they should all be set to 100% therefor, an actual injector trim of zero. Don't mess with settings that do not need correcting. while you may have achieved the result you are looking for, its not actually correcting the initial problem.
Is engine tuned? There is total failure with injector voltage compensation. It is approx 5 times too high for any injector. If tuning was done exactly at 13.2V - then no problem. If you had 1V more or less it is roughly 0.8ms error.
VE table looks weird at low load area.
I have been thinking about this MAP/ RPM correction table. Default config has the whole table set for 100%, yet Mattias has the whole table set for 0%
So a question to Mattias, if you are reading this, does all zeros dissable this function? and is the end result the same as it would be with all 100's? The reason I ask is that I was under the impression that all correction factors were zero'd at 100%
Quote from: Sprocket on February 26, 2011, 08:18:42 PM
I have been thinking about this MAP/ RPM correction table. Default config has the whole table set for 100%, yet Mattias has the whole table set for 0%
So a question to Mattias, if you are reading this, does all zeros dissable this function? and is the end result the same as it would be with all 100's? The reason I ask is that I was under the impression that all correction factors were zero'd at 100%
that table is only for apha/n.
I also think that is only for Alpha N,because I've changing the value and it does'nt change anything....
Quote from: MWfire on February 26, 2011, 11:45:32 PM
Quote from: Sprocket on February 26, 2011, 08:18:42 PM
I have been thinking about this MAP/ RPM correction table. Default config has the whole table set for 100%, yet Mattias has the whole table set for 0%
So a question to Mattias, if you are reading this, does all zeros dissable this function? and is the end result the same as it would be with all 100's? The reason I ask is that I was under the impression that all correction factors were zero'd at 100%
that table is only for apha/n.
I understand this, but your reply does not answer my question.
I've tried to set my req fuel to 6.2 and set the inj trim to 100% and the engine run too much rich...
If I want to keep the trim at 100% I have to decrease the reqfuel near to zero,is it suitable?
I don't understand why the trim are tunable and in opposite we cannot change value( 75% for example)....
At this time with this setup trim(75%) with 5 to req-fuel,with map/corr at 0,I can:
-Have a good Air/fuel ratio(even better in the futur with fine adjustment)
-Ego correction near to 90-100%
-the car run good to 150 kpa ,no mis fire,good feedback to pedal gaz
-MAT CORRECTION working
-Lambda target working
-the car also works LPG fuel!
Who can explain me that....
My question now is what will be the result at high load ,like 250~260 kpa?
I forgot,my battery calibration set now to "188" like the factory ecu recommendation....
At fuel calibration there is always sequence to follow. One of very first things is determine injector lag characteristics, calculate Required fuel. And tune everything else just after that. Otherwise you potentially can get all kinds of instability.
188 is too less. This value means what PW in microseconds need to add at 7V comparing with 13.2V. But in reality used not for so low voltage, but for stability in common 11...14V range. Ballpark is somewhere in 600...900us range.
Injector trim is designed for use like one of final adjustments, when it is clear that some cylinder eats more or less air than other. Usually separate EGT sensors for each cylinder is used for this. It is not for basic adjustments.
Gintsk,you're right ....
In fact when I set my Battery calib. I hesitated to introduce the recommended Vems factory value or the digital volt meter volt value.
Engine running,when I put "188",there is about ~0.7v of difference between the DVM and the vemstune battery indicator.
To align the value I need to increase it to "195-194",that way the difference is near to zero and the global air fuel ratio decrease.
But even doing this mod I can't set my trim to 100 %.
I'll post tomorrow my last config in case of you'll find something wrong,because after reading your last message I finally understand why is it so important to keep the trim at 100%
Thanks a lot ;)
Michele.
All sorts of settings were off or wrong in this config file. Always use the "verify config" button to check and fix all settings that don't make sense.
Press F1 in all dialogs to get instructions on proper settings.
You can always drag'n'drop working configs over your dialogs to copy just the settings pertaining to that dialog from the particular config file.
Please publish a more recent .vemscfg or .vemslog (preferrably zipped) and I might provide feedback on that, it seems many corrections to some settings might have been made since your last published file.
Michele, I wrote not about Battery calibration, but about Injector voltage compensation. You had somewhat like 4000usec in your setup. Purpose of this value is to make mixture not dependent from voltage changes.
Reqired fuel is not so important. It is recommended to use calculated value - then most of default settings will be close to desired. But engine is still tunable if this value is 0.8...2 times calculated.
After several tries,I finally run the engine at idle (+/_ lambda 1)with trim @100%.
The ego correction wasn't so bad.
Here the config file:
http://www.box.net/shared/922edbf5cf
Please don't pay attention to the rest of fuel map,I still do the road test.
Can you take a look to cranking value and give me your opinion...
thanks
Michele.
Quote from: rossi46numberone on February 28, 2011, 01:59:22 AM
After several tries,I finally run the engine at idle (+/_ lambda 1)with trim @100%.
The ego correction wasn't so bad.
Here the config file:
http://www.box.net/shared/922edbf5cf
Please don't pay attention to the rest of fuel map,I still do the road test.
Can you take a look to cranking value and give me your opinion...
thanks
Michele.
Hi.
Your Trigger reference tooth table and ignition output is configured to be runned as fully sequencial but your Secondary trigger is disabled.
Are you sure your Injection table isnt reversed?
Peer review is always nice, I went out of my way and e replaced and changed a lot of strange settings and uploaded a new config file for you to try here :
http://www.vems.hu/files/MembersPage/MattiasSandgren/vemscfg/vw-test-edit-vemscfg.zip
Injector voltage compensation was very off (4 ms is A LOT). Secondary trigger settings were strange (disabled). Ref trigger tooth table was wrong. Miscellaneous settings were disabled to help you start up (overrun fuelcut, idle control, etc). I loaded a decent VE table from a 4 valve/cyl engine. Unless you have way radical cams it should start nicely if your req fuel is set according to calculations and the correct injector voltage compensation is used (fixed that too).
After loading the config - before you do anything else you need to go through your ignition output table (injectors not powered) and verify the ignition order and make sure cyl 1 ignites at the correct spot (strobe light), and maybe others unless you're sure of the ignition order. I changed both your ref tooth table and ign out table, should/could work as it is but you never know.
(good to know : there are configlets/dropdown menu options that loads good default trigger settings and other configuration parameters, check bottom of the primary trigger dialog).
Hi Mattias,
Thanks for your reply and config.
I've tried it,and the engine won't start and regarding your setting i've decrease the inj voltage compensation and it seems to work well,and my fuel table begin to look right,the value are more logical but sure far from optimal .
Strange coincidence,today morning before receiving your message I have checked the cyl 1 ignites at the correct spot.
And it does,but I had to reduce my trigger point from 84° to 78~79°.
Now my digital strobe light corresponds to the VEMS timing gauge indicator.
I have also increase the Dwell time .
Is it possible that it is not obliged/necessary to work with the second trigger sensor??I think that other guys run this kind of engine only with VR sensor,I'm wrong....
Here is the last config test:
http://www.box.net/shared/sbmht5erc6
Michele.
Michele, we can't help if you try to build your tune with broken basics. Please reread Mattias previous post and try to follow it.
Nice to see a Wiki page, good to have one spot to find consistent info on your setup. Put a link in your forum signature for easy access.
Your boost solenoid doesn't work because you have not activated any output in the injector table for it. Check the visual output dialog, the last row says "boost solenoid" as that is what you selected and it has no output activated. This is the least of your problems. Use my config, forget about your old one and find the ignition order.
You can run without cam sync, but then you must use "dual out" ignition mode to fire the ignition outputs in pairs = fake wasted spark.
Ok mattias,
I'll do my best to find to right firing order....
One question about the priming pulse:
In the megatune soft it apear that is an important think to follow for starting the car in all conditions and if I'm not wrong, it's tunable like a "table" ,example:
-40°c = 11ms
-20°c = ...ms
0°c = ...ms
10°c = ...ms
70°c = 3.1ms
And in Vemstune at warm engine it seems that it can usually be zero or near to zero.....
Do I miss anything?
Prime pulse is very secondary thing. In addition for founding best values for table you need a laboratory, big refrigerator and thousands of cold starts. VEMS algorithm for prime pulse (pulse during power-up) is greatly simplified: it use just two points: pulse at cold (-40C, but I am unsure) and scale down at warm. So put somewhat like 20ms and 10%. 20ms - -40C 2ms - warm engine.
In fact,at warm engine and waiting a few minute between each start attemp,he does not start @ first key, 95% of the time @ the second attemp(always without touching the throttle).
Do I have to play with the Crank VE or cranking enrichement for best result,I'm scare to flow the engine end kill the battery/starter with too many cranking. :-\
Is there other things to take in count or a precise method for tune the starting sequence in all conditions with Vemstune
Ps:if I try immediately after,it start like a charm...
Yes: try to rise cranking VE by 10...20% above value in VE map 100kpa/idle (if tuned).
Quote from: rossi46numberone on March 05, 2011, 03:38:45 PM
In fact,at warm engine and waiting a few minute between each start attemp,he does not start @ first key, 95% of the time @ the second attemp(always without touching the throttle).
Do I have to play with the Crank VE or cranking enrichement for best result,I'm scare to flow the engine end kill the battery/starter with too many cranking. :-\
Is there other things to take in count or a precise method for tune the starting sequence in all conditions with Vemstune
Ps:if I try immediately after,it start like a charm...
It´s because you´re trying to run the engine fully sequencial, but without a camsync signal.
You have to use "Dual-Out"-mode for your ignitioncoils.
The reference tooth table you should use is:
0
30
0
30
Wire cylinder 1 ignitioncoil to Pin 34
Wire cylinder 3 ignitioncoil to Pin 24
Wire cylinder 4 ignitioncoil to Pin 36
Wire cylinder 2 ignitioncoil to Pin 10
And use "Dual-out"-mode.
And at "Ignition Outputs visual" you are going to use these settings:
0 - Ignition output 2+3 (7A) Grounding, EC36 pin 34, 36
1 - Ignition output 6+7 (7A) Grounding, EC36 pin 24, 10
2 - Ignition output 2+3 (7A) Grounding, EC36 pin 34, 36
3 - Ignition output 6+7 (7A) Grounding, EC36 pin 24, 10
You maybe have to switch places in order to make the correct pair fire.
That will probably solve your problems.
Ok i'll try tomorrow and give a feed back asap...
Thanks a lot guys for your precious help!
Regards,Michele.
Sooo... what happened here? ::)
i have been helping a little and i believe he is up and running on dual out.. still waiting for his 1 window cam wheel to come.
now he is trying to sort his stock tach in dash to work from vems output..
did you get your tach working with the info i sent you?
Hi,
Some news.....
Today the car is running fine.
Finally I decided to keep the original ECU in place and using the Vems for timing,injection and boost.
Why?
Because there is somes operations that the Bosch ECU can control nicely like increasing Rpm when sterring or when turn on the A/C,cold idle speed control,tacho,OBD,etc....
What about Vems:
For ignition I used the wasted spark mode,I'll probably swap to fully seq because it seems that the coils don't like this.
For injection I have choose the "simplified" injector setting and it's work like a charm and more easy to tune.The car is running on Bi-fuel with the swapping Table/config activated,able me to tune each kind of fuel properly.
The engine idling at Lambda 1 and cruising at 1.05 with no missfiring or issue.
I also keep the LPG ecu(Brc system) and make all the stuff working :Gas gauge,transition fuel superposition,gas temp control,specific open LOW_Z injector curve,Tank solenopid valve security,P1-map delta pressure,etc....
Michele