Help with calibration + log + dyno results

[fizban]

Hello all,
I am fairly new to VEMS tuning. I use fw 1.1.64 and VEMS tune.
The car is Peugeot 306 with GTI6 engine and a turbo (GT28RS).
I've managed to tune it ok for cruise, but at full throttle @ the dyno I am seeing strange things like power drops and torque changes. I used mild advance on full throttle since i didnt have a dyno to tune with at start.
Attached are Dyno results and a log file (was not taken on the dyno but on the road).
I've been told that my MAT/TPS calibration is wrong and that i have too much Accel Enrich.
I would appreciate any help with the tuning.

Log:
https://docs.google.com/leaf?id=0B6l1SB-S-teyY2RiMzU4N2UtYjIwNi00N2NjLTkwNTUtZDQ4ZDI0NmQxOWIz&hl=en_US

Dyno: (280HP@6100 RPM, 33 kgm and 9 psi of boost)

[GintsK]

Ouch!

Seems you are in trouble.
First. IIRC no firmware in 1.1.6x range was not good. I do not remeberer what problem (bug) it was, but simply better not use it!

Second. You are in trouble from very base of tune. Push "validate" button in Vemstune. Trigger settings are wrong. Means ignition calculations are off.  WBO2 calibration are wrong.

Tables are scaled just up to 120Kpa? You blow more than 0.2bar for sure!

Injectors fires all in parallel once per cycle? Why? Worst scenario!

IMHO you have to start again from beginnings- this tune worth nothing in my eyes.
Put somewhat like 1.1.95, rewire injectors, correct wrong settings, calibrate sensors. Rescale tables for your planed boost range, put default MAT correction...
Migration to newer firmware is very easy with Vemstune - just follow instructions. With newest firmwares you will get better support. Simply very few people ran 1.1.6x. Most of us jumps from 1.1.53/54 to 1.1.70 or later f/w.

Good luck!
Gints

 

[fizban]

Thanks for the reply.
I am new to VEMS, but not that new...
You really think the car could output 280hp with the problems you describe without blowing up? 

The firmware was provided by DP_Engineering - who also assembled my PnP VEMS unit and gave a base map.

Obviously you see something else from what i see in the same log file. Maybe compatibility issues?
Here is the VE table. This is how i see it in my vemstune.
Maybe there is something i need to do before exporting the log file?

[GintsK]

For some reason for me tables looks so:
http://www.4shared.com/photo/YMVwLFRQ/VE_online.html
Had you something logged it that file? I see just 13 seconds stalling engine.

[fizban]

Yes, it looks different from mine.
I used http://www.dp-engineering.nl/random/VEMS_PnP/VT110120.exe
This is the version given to me by dp.
Maybe this is why it looks different.
The log contains a run with 2nd gear to 6800rpm. This is the limit of my current 440cc injectors (97% DC). obviously i don't run that kind of RPM normally.

[GintsK]

Then try to share separately your config. And convert log file to csv (under tools) and share it.

Or install last official VT. It shouldn't hurt as long as you have .ini for your firmware. Developers made really huge work to improve VT during this year.

BTW your VE table screenshot looks more like fuel pump flow insufficiency than injectors. 440 should give more than 300hp at 100% DC. You have 280.. on crank?

[fizban]

Config file:
https://docs.google.com/leaf?id=0B6l1SB-S-teyMGI2ODk0YzEtN2FiZS00OWI4LWFiM2YtNmM2YWRiYmNlZTBl&hl=en_US

Log:
https://docs.google.com/leaf?id=0B6l1SB-S-teyYzM0YmRlZGMtMDAxMC00ZjdjLThmYTYtYjY2YTNiYmUyNTRh&hl=en_US

280 from the crank, I think. It was tested with dyna pack dyno and corrected to crank (unless i am missing something)
How do you analyse that is a fuel pump issue according to the VE? This is very interesting. I was wondering why i get high duty cycle with this power level.

[GintsK]

Regarding table. From experience lower lines in true VE table always are more/less parallel with high KPa lines. If top right corner of map goes up it means compensation of fuel pressure drop.
Also there is no reason for volumetric efficiency go sky high from 4000 to 7000rpm.

Now config looks hardly different to me.
Anyway validation check gives two significant errors. WBO2 value are off. So it can be that sensor readings are not reliable. Look at data supplied with unit.
And you have partial fuel cut enabled: simply change cut or resume rpms so that difference is not 1000 or 1100.

Did you use boost control? Values there looks weird. Nowadays Boost PID control works perfectly. No reason to use alternate control strategy as main.

I would still recomend configure Lambda table different - within operating range.
Also your MAT/TPS correction table is dangerous. Look how looks default one.

Injector wiring/firing are crazy! Why you do so?

Acceleration... hard to say how VEMS calculate it when injection divider 4 is used. Seems it multiply by 4. But as you see no hard reaction visible on Lambda signal. You heve too much AE at slow tps movements. And Rpm curve should be configured opposite: more PW at low rpms and low or even no AE  at high rpms.

Does "dyno" and logfile comes from same run? You had tyre or clutch slip during WOT. It disturbs road-dyno measurements for sure.

Gints

[fizban]

Many thanks for your help, it is much appreciated!  I don't have anyone here even remotely familiar with VEMS.
I changed the fuel cut to 2900rpm and resume to 1700 to avoid partial fuel cut.
I have another WB sensor connected to  gauge (AEM) and i use it to compare values. I did the WBO2 calibration (with sensor held out in the open air). When the AEM gauge shows 14.5:1, the VEMS gauge (i think its s VW sensor) shows 1 lambda. So i guess i run a little bit richer than expected, but on the other hand, maybe the AEM is off scale. Who knows?

I would very much like to see default tables. Can you direct me the relevant locations?
I did not touch injector wiring/firing - it was provided by DP.
I've changed AE (rpm) to:
800 30
1500 25
3000 15
6000 5

The car feels crispier upon accel, but i get a short spike of lean mixture when pushing the throttle. Is this natural or should i increase dotrate vs. scale?

I did not use any boost controller. I plan to use one in the future to have 3 settings (off - 0.6 bar/low - 0.9 bar and high - 1.2). I dont think that VEMS provide several settings (but i might be wrong, of course)

the dyno and the log file are not from the same run. Dyno was performed with dyna pack (no tires) on 4th gear and the log is on 3rd (if i remember correctly). The car does not slip at 3rd gear so i probably did it on 3rd (on the road)

 PS. I tried to tune the car up to 176kpa. beyond that i did not dare to go because i am not confident yet and also limited by injector size. so all the data above 176kpa can be ignored for now. I added a lot of fuel there in case i have boost spikes.
Also , i have a fuel pressure gauge and i see slow fluctuations at hard accel between 4 to 4.5 bar

[GintsK]

There are two values related for ECU hardware calibration. Your nernst target is off for sure. So we can presume also pump zero as unchecked. Second one is more important. Please check! First one just describe target voltage for narrow-band part of sensor. Normally it is 0.45V. Now you have something higher.

Default tables comes together with each Vemstune. It is possible to import separate tables easily. Just push [import] and browse for file.

Sorry I haven't clue what is close to right values for your funky injection strategy. Mine would be somewhat like 100 60 30 15 for your rpms. Try more/less and keep that shape. Also take attention to AE TPS sensitivity graph.

What is your fuel base pressure? What is planed boost? Sum both. Set this sum in fuel rail (e.g. by attaching spare wheel pressure to FPR feedback nipple) and measure fuel flow in return. You want to see somewhat like 2.3L/min. Your 176Kpa line in VE table has same disturbance.

Is it simple engine with no camshaft angle or manifold geometry controls? Do you run OEM cams?
How dyno operator comment results? Sorry I misunderstood from where dyno graph comes.

BTW you can add potentiometer to VEMS or simpler kind - switch with 3 voltage levels - and get whatever boost levels you want! Within w-gate capabilities of course. Check "anytrim" feature! VEMS boostcontroller is really excellent!

Gints

[fizban]

I do not wish to play with hardware calibration done by DP, as they build it. I will consult them as well.
I've seen the default tables, thanks!
I will also re create the AE table to get better results.
The pressure valve is rated 3.5 bar. Maybe i should just replace the fuel pump to a more flowing one (it was planned anyway, just in a later stage).
The engine is simple. No geometry change or camshaft angle changes. It is 16v and the best head Peugeot had to offer for the 306. I run OEM cams.
The dyno operator said that something is wrong with ignition and that after 6000rpm there might be something wrong or Catalytic converter limitation
I will worry about boost control once i am sure the car is tuned well enough to push up the boost.

[GintsK]

So you had misfires during dyno run?

[fizban]

No knocks were registered, i think he based the ignition problem after looking at my ignition maps (which is mild since till then i did not have dyno to tune ignition). i think i will schedule a dyno session to finish with ignition, but i cant to that before ensuring my config and sensors are ok.

[GintsK]

Misfire is non successive ignition event. It is not knock.
Do you feel those torque waves on street?

[andreNL]

I would:
Check and calibrate all sensors
http://www.vems.hu/wiki/index.php?page=GenBoard%2FManual%2FWideBandHardwareTest

Turn off accelerateenrichments and fuelcut,
Check all base settings
Now drive the car with loggin on and ve analyser
Use lambda table for fueling
If fueling ok turn log on go wot from 1500 till 3500
Retard ignition 1deg do it again if wbo leaner its better
Add fuel if to lean in lambda table
Tune for drivebility not mathematics