Hi all!
First let me just say that I'm a recent user of this forum and in tuning Vems ECU's.
I have experience in tuning Megasquirt ECU's.
I'm tuning a friend's Citroen boosted Saxo VTS with Vems ECU that he has bought from DP Engineering.
The ECU is a V3 and it has came with the firmware version 1.1.64.
I have managed to tune the idle, not perfect but acceptable, and the low kpas of the VE table and the car was drivable.
Then I decided to upgrade the firmware, because I read that the versions 1.1.6x wasn't that good.
I upgraded to version 1.1.70 and uploaded the config that I was using in version 1.1.64 and then the problems started.
Now the idle is very bad and the rpm very high, even if I lower the ignition to 5 degrees, close the IAC valve and with the AFR showing 15.5, the engine stil revs at +/- 2000 rpm and I can't lower it.
Also the car was very smooth when I hit the throttle and now is very weird and sometimes backfires a litle bit.
Another thing is that I think the coolant gauge is not showing the right temperature.
Don't understand what happend. I just upgraded the firmware.
Can someone explain what I'm doing wrong or what's happening?
Thanks in advance.
I would upgrade to at least 1.1.96 using the latest VemsTune from 2012-11-02.
Then get it running and/or even if you don't, publish the config and/or vemslog (log contains config) and a triggerlog (Tools menu at the bottom) using the built-in reporting tool in VemsTune (see Help menu).
Nobody uses these old firmwares anymore (1.1.70, etc..) and more recent firmwares provide more information in the datalogs to help solve issues.
Ok I will do that.
Thanks.
And to avoid problems at the start, I would disable things like :
- EGO correction (set enlean/enrich to 0%)
- IAC PID control (set integral increase/decrease to 0)
- Ignition advance control (has a disable option)
- Overrun fuel cut (set cut/resume to 25500 rpm)
That's a good start to figuring out the basic fuel and ignition issues.
Here are the logs.
The first log is with all the things enable.
The second is with all the things disable.
http://www.sendspace.com/file/g91fnj (http://www.sendspace.com/file/g91fnj)
http://www.sendspace.com/file/k9qz2v (http://www.sendspace.com/file/k9qz2v)
I downloaded the first one.
For some reason ignition table is from TPS and with wrong scaling.
Also VE table is strange. It never looks like this. I mean high boost/rpm corner. If VE is high at those rpms it also represents at lover loads.
My VT crashes when I want to look lambda target table. Try to share .vemsconfig file.
Can you comment in more details what engine setup is used. Camshaft. Injectors. Turbo.
Gints
The car have a T4 turbo with A/R 78, 630cc siemens injectors, inlet and exhaust manifold from DP Engineering.
Also have mechanical camshafts but don't know exactly wich ones.
Later today I will share the config file.
If you download the second log that I share, you will see that in that log with all the things disable (EGO, IAC PID, Ignition advance, overun fuel cut) the car is idling fine and was fine for a little while and all of a sudden, even though I have done nothing, the rpm droped and the the engine stalls.
Now he can't maintain idle.
I have enabled IAC PID and Ignition advance but is the same, the engine can't maintain idle and stalls.
Seems that this don't do nothing... don't know why...
For instance, if I turn ignition advance on I should see ignition gauge moving like crazy to try compensate, at least that's what's happen on Megasquirts.
Here on VEMS the advance is always on the value that is defined on the ignition table.
Also seems that the stepper motor doesn't do what I want...
Here is the config file that I'm using.
http://www.sendspace.com/file/bz9fcu (http://www.sendspace.com/file/bz9fcu)
Why do you say that ignition table is from TPS?
I've been checking and ignition table is RPM/MAP.
It is 1.1.64 config. Can you share also 1.1.96 one? Here ignition is OK. But something is wrong with 1.1.96.
Is it 1600cc engine?
I do not like MAP value at idle. For turbo engine idling at 1600rpm 56kPa. 75kPa@950rpm. Huge valve overlap? Can you check cam timing to be sure?
Idle is rough always?
You can check stepper function by dismounting it. Then under tools->iac lock you can try to move it. More steps - wider opening.
Yes it's a 1600cc 16 valves engine.
The camshafts are the Piper PK5. You can check the details of the camshafts here http://www.pipercams.co.uk/pipercams/www/product.php?pid=PK5 (http://www.pipercams.co.uk/pipercams/www/product.php?pid=PK5).
So sorry. ???
The camshafts are the newman PH5 and the details are here http://www.newman-cams.com/pdf/citroen.pdf (http://www.newman-cams.com/pdf/citroen.pdf)
Quote from: Bcup on December 04, 2012, 05:29:42 PM
So sorry. ???
The camshafts are the newman PH5 and the details are here http://www.newman-cams.com/pdf/citroen.pdf (http://www.newman-cams.com/pdf/citroen.pdf)
Ouch! I am not familiar with particular engine. But IMHO those cams are overkill for
any 16V turbo engine! 290deg with
104/106deg max lift angles and
3.4/3.7mm opening (!) at TDC...
Last values means lack of idle (without ITBs). During idling even moderate vacuum in intake plenum creates backflow from exhaust manifold trough combustion chamber to intake. And engine breathes own fumes. And big valve overlap is not for turbo! Because same effect takes place as exhaust pressure rises up and over intake pressure value. I would try to retime exhaust cam closer to 115deg from TDC. Look at this page with timing values (under specs) and descriptions:
http://www.camshaftshop.com/products.php?categoryid=75
Stepper I think is almost useless. Ok it can give some help. But most effect here for idle stability you can expect from very aggressive ignition idle aid. Like +/- 12deg with 45deg/256rpm. And I think ~1500rpm is lowest idle what you can hope on.
1600cc. Your required fuel is calculated wrong (hit F1). I suggest to rescale it together with VE table. Some of us use even twice lower r.fuel as calculated. It gives better resolution for VE map. Similar as MS.
What about 1.1.96 config?
Car is purely for dragracing? Or?
Gints
This cams are for a non turbo car but the owner decided to keep them with the turbo installed.
The car is for dragracing and trackday, I think, not for road.
Today I will share the 1.1.96 config and I will try your values in the ignition idle control to see if it gets better.
Here is the config 1.1.96 that is in the ECU right now.
http://www.sendspace.com/file/087znd (http://www.sendspace.com/file/087znd)
Quote from: Bcup on December 03, 2012, 11:48:20 AM
Hi all!
First let me just say that I'm a recent user of this forum and in tuning Vems ECU's.
I have experience in tuning Megasquirt ECU's.
I'm tuning a friend's Citroen boosted Saxo VTS with Vems ECU that he has bought from DP Engineering.
The ECU is a V3 and it has came with the firmware version 1.1.64.
I have managed to tune the idle, not perfect but acceptable, and the low kpas of the VE table and the car was drivable.
Then I decided to upgrade the firmware, because I read that the versions 1.1.6x wasn't that good.
I upgraded to version 1.1.70 and uploaded the config that I was using in version 1.1.64 and then the problems started.
Now the idle is very bad and the rpm very high, even if I lower the ignition to 5 degrees, close the IAC valve and with the AFR showing 15.5, the engine stil revs at +/- 2000 rpm and I can't lower it.
Also the car was very smooth when I hit the throttle and now is very weird and sometimes backfires a litle bit.
Another thing is that I think the coolant gauge is not showing the right temperature.
Don't understand what happend. I just upgraded the firmware.
Can someone explain what I'm doing wrong or what's happening?
Thanks in advance.
I think I've managed to watch your 1.1.96 config. I had broken VT installation on my job PC.
I can't find somewhat wrong comparing to your 1.1.6x one.
But anyway - there is some issues:
As I mention - required fuel. Some enrichment values is scaling of required fuel: cranking, warmup, acceleration...
I suggest to check ignition timing. Because 1.1.6x was not officially supported f/w.
My suggestion is also switch to "simplified" injector lag compensation and write there some reasonable values. Better - find real ones.
EGO step size should be 1%. Cycles - usually 4...10. Speed limit 40...64. You can use closed loop also under boost if you have no undersized exhaust (pressure disturbs signal). You will find faster O2 sampling in 1.2.0 - migration here is easy comparing to your 1.1.6x->1.1.96 case.
MAP sensor range is 413 and offset 3...4.
Check wide band calibration.
Coolant and MAT temperatures. Here you can use patch files (tools->firmware->patch upload) or use own curve. Second option give more flexibility.
VE table is not tuned. And probably it will be very hard task with those cams. May be better to choose Alpha-N strategy with MAP compensation under boost. Best - change the cams. With correct required fuel you will not see there values higher than 120. If higher required - means defect: pump, pressure regulator, injecors 100% duty...
Igniton table generally is ok. load range is sufficient? Dragrace require much more boost!
MAP/TPS table is dangerous. In real life it is not so sloppy. You can import one from default configs and start from there.
Acceleration dtps map is terribly wrong. Small movement should mean little additional fuel.
Low rpms require more RPM acceleration values, high RPMs - less.
Stepper. Try in open loop (PID zeroed) - does stepper gives enough flow to change something. Most of correction will be upon to ignition.
Boost control. Do you use it?
Speed sensor?
SD card?
Gints
Quote from: GintsK on December 05, 2012, 02:33:01 PM
Quote from: Bcup on December 03, 2012, 11:48:20 AM
Hi all!
First let me just say that I'm a recent user of this forum and in tuning Vems ECU's.
I have experience in tuning Megasquirt ECU's.
I'm tuning a friend's Citroen boosted Saxo VTS with Vems ECU that he has bought from DP Engineering.
The ECU is a V3 and it has came with the firmware version 1.1.64.
I have managed to tune the idle, not perfect but acceptable, and the low kpas of the VE table and the car was drivable.
Then I decided to upgrade the firmware, because I read that the versions 1.1.6x wasn't that good.
I upgraded to version 1.1.70 and uploaded the config that I was using in version 1.1.64 and then the problems started.
Now the idle is very bad and the rpm very high, even if I lower the ignition to 5 degrees, close the IAC valve and with the AFR showing 15.5, the engine stil revs at +/- 2000 rpm and I can't lower it.
Also the car was very smooth when I hit the throttle and now is very weird and sometimes backfires a litle bit.
Another thing is that I think the coolant gauge is not showing the right temperature.
Don't understand what happend. I just upgraded the firmware.
Can someone explain what I'm doing wrong or what's happening?
Thanks in advance.
I think I've managed to watch your 1.1.96 config. I had broken VT installation on my job PC.
I can't find somewhat wrong comparing to your 1.1.6x one.
But anyway - there is some issues:
As I mention - required fuel. Some enrichment values is scaling of required fuel: cranking, warmup, acceleration...
I suggest to check ignition timing. Because 1.1.6x was not officially supported f/w.
My suggestion is also switch to "simplified" injector lag compensation and write there some reasonable values. Better - find real ones.
EGO step size should be 1%. Cycles - usually 4...10. Speed limit 40...64. You can use closed loop also under boost if you have no undersized exhaust (pressure disturbs signal). You will find faster O2 sampling in 1.2.0 - migration here is easy comparing to your 1.1.6x->1.1.96 case.
MAP sensor range is 413 and offset 3...4.
Check wide band calibration.
Coolant and MAT temperatures. Here you can use patch files (tools->firmware->patch upload) or use own curve. Second option give more flexibility.
VE table is not tuned. And probably it will be very hard task with those cams. May be better to choose Alpha-N strategy with MAP compensation under boost. Best - change the cams. With correct required fuel you will not see there values higher than 120. If higher required - means defect: pump, pressure regulator, injecors 100% duty...
Igniton table generally is ok. load range is sufficient? Dragrace require much more boost!
MAP/TPS table is dangerous. In real life it is not so sloppy. You can import one from default configs and start from there.
Acceleration dtps map is terribly wrong. Small movement should mean little additional fuel.
Low rpms require more RPM acceleration values, high RPMs - less.
Stepper. Try in open loop (PID zeroed) - does stepper gives enough flow to change something. Most of correction will be upon to ignition.
Boost control. Do you use it?
Speed sensor?
SD card?
Gints
Yes I use boost control and speed sensor.
SD Card no.
I will update to version 1.2.0.
Required fuel - Going to lower to 4.
Ignition timing - Why do you say to check that?
Coolant and MAT temperatures - I have patched them when I update the firmware but they have so many options that I don't really know what to choose much less making my own curve. In MS I have specific values to calibrate this sensors.
Load range - The owner wants to make 250 kpas of boost
The rest I will try what you've said.
Thanks.
Quote from: Bcup on December 05, 2012, 03:29:44 PM
Yes I use boost control and speed sensor.
SD Card no.
I will update to version 1.2.0.
Required fuel - Going to lower to 4.
Ignition timing - Why do you say to check that?
Coolant and MAT temperatures - I have patched them when I update the firmware but they have so many options that I don't really know what to choose much less making my own curve. In MS I have specific values to calibrate this sensors.
Load range - The owner wants to make 250 kpas of boost
The rest I will try what you've said.
Thanks.
Then you have to fill Boost control windows with reasonable values. Also here you can start with PID zeroed.
What w-gate type?
Did you multiplied the VE map after req.f change?
Ignition - because you had non-official firmware before. And engine running changes after upgrade.
Temperature curves is bit painful thing for VEMS, yes. But usually curves is close to BOSCH default. For VAG 1.8T air sensor I often use 1900/256 patch.
Ignition map goes only up to 220kpa. You need more than expected boost level to add some safety.
Quote from: GintsK on December 05, 2012, 04:08:15 PM
Then you have to fill Boost control windows with reasonable values. Also here you can start with PID zeroed.
What w-gate type?
Did you multiplied the VE map after req.f change?
Ignition - because you had non-official firmware before. And engine running changes after upgrade.
Temperature curves is bit painful thing for VEMS, yes. But usually curves is close to BOSCH default. For VAG 1.8T air sensor I often use 1900/256 patch.
Ignition map goes only up to 220kpa. You need more than expected boost level to add some safety.
It's a dual port external wastegate.
I still haven´t made the change in required fuel.
Going to do it tonight. When I made the change I have to incresase the VE table.
The MSD ignition now only sparks the first and fourth cylinder.
I have checked connections, swaped cables and the MSD coil and still the same.
Can this be hapenning because of some worng parameter of the ECU?
The ignition outputs maybe?!?
Does triggerlog shows all four spark events?
Under ignition outputs you can find test mode and visual represantation of connector/outputs.
Hi GintsK.
I have tryed test mode but it wasn't working, don't know why. ???
I've turn it on and then push the test button on the output I want to test but nothing.
I haven't try triggerlog because like I said I have no experience ins Vems ECU's and with Vemstune and I didn't know that is possible to see the spark events in that.
But I will try it. :D
Thnaks.
Gintsk can you check the trigger log for me, please, to see if it shows all 4 spark events.
I don't understand it. ???
http://www.sendspace.com/file/f824cg (http://www.sendspace.com/file/f824cg)
Thanks.
Yes, it shows. 2 events per rev = 4 events per cycle. Correct.
What type of MSD?
I have the MSD coil with pn8241.
So it's nothing with the parameters of the ecu or with the outputs?
I do not see anything wrong. Dwell is shorter. But anyway...
So you can check the outputs by inverting them. With coil not connected!!!
At this state best way is t check outputs with some significant current. E.g 60W bulb.
During firmware change did you disconnected the coil? It could be damaged if was connected. Sometimes, especially for older firmwares outputs went to random state.
Yes I have disconnected the coil when I updated the firmware.
The car went to the owner, so he could check what's going on and it seems it was the msd coil that's broken.
Tomorrow I will receive the car to start tuning again.
Thanks GintsK for all your help until now.
I have the car again and now it's much better.
I have followed the advices that GintsK suggested.
The idle is now at 1200/1300 rpm with PID disable and ignition control disable.
I enable ignition control and use the values that GintsK suggested "+/- 12deg with 45deg/256rpm" but nothing happens. It stays the same.
Anyone know why this don't work?
GintsK you also said to "fill Boost control windows with reasonable values". But what are reasonable values. Can you give me some values please?
Also would like to know how to change from lambda table to AFR table because I'm accustomed to see AFR values and I can't change it.
Thanks.
Nothing happens - probably your TPS value is above idle treshold. No?
For boost: boost target (it has no influence with PID zeroed), bias DC table (low values at beginning). If you want to use separate boost for each gear - it is possible to define separate tables for each gear. Both tables.
You can go with PID from begging. Try P/I/D 100/20/70 as start point. Integral limit 100. TPS weight 200. Fill also TPS table - very useful for cars with limited traction. - It will scale down target boost and also DC together with TPS weight value.
Boost by gear require working and calibarted speed sensor values.
For AFR - go to File->Firmware parameters and change USEAFR147 to 1. Restart VT. But remember AFR is technically valid only for gasoline. But Lambda is universal value for any fuel. AFR causes confusion once you start to use any other fuels.
Gints
Quote from: GintsK on December 12, 2012, 10:27:59 AM
Nothing happens - probably your TPS value is above idle treshold. No?
But the IAC light comes on when the car is idling.
If the TPS was above idle threshold the IAC light come on?
.vemslog please.
Here is the log with ignition control enable.
http://www.sendspace.com/file/hgbaz8 (http://www.sendspace.com/file/hgbaz8)
Launch control is activated all the time :)
Why you say that? ???
And that's why ignition control don't work at idle?
Launch control deactivate idle. It is not designed for idling. Usually pushbutton is used for it and just for short period before race start.
Probably channel is wrong. Hit backspace to see input channel screen.
GintsK what do you think of this log now. ;D
http://www.sendspace.com/file/w7439u (http://www.sendspace.com/file/w7439u)
:o OK... ;D
Did you also re-timed the cams, or? Your MAP values seems much lower.
Now you could try to implement stepper in your orchestra.
Gints
No I haven't re-time the cams.
I think I will let the idle with ignition control only.
It's very good now.
Just one more question.
I saw in some place of vemstune an option to close the stepper when in boost.
Can you tell me where it is?
I can't find it now... >:(
Saturday I'm going to tune the boost part of the table in the RR.
IACV closing is under Idle Settings. But I don't think this is a good idea with stepper. Sometimes steppers loose their reference position "learned" at startup. During tuning is wiser to block stepper tube. Or simply connect input side to pressure pipe.
Did not retimed?! OK.... did you measured compression?
If you can achieve 55Kpa at idle - it means very big lobe to TDC angles. Or simply more "normal" cams than you wrote.
No I didn't measure compression.
The owner of the car said to me that the cams that are installed in the car are the ones that I said.
I doubt that the owner have changed it.
Thanks for all your help GintsK.
We went to the RR this saturday to tune.
The car is with 340BHP and I haven't been able to tune any more because the screews from the flywheel has loosen up so we have to stop the car.
In January we will go again to the RR to finish, I hope. ;D