VEMS Community Forum

Technical => Documents => Topic started by: gunni on October 13, 2012, 11:54:45 AM

Title: Missfire detection, cylinder identification
Post by: gunni on October 13, 2012, 11:54:45 AM
So I was on the dyno the other day with a M50B25 turbo, it was producing 360hp@1bar boost which is far below what such an engine should be making at that boost with a 56mm compressor.

Thankfully it has a VEMS ecu and using it I can identify which cylinder is dropping. In this case there was actually two problems, the number two injector connector was fully on but wasn´t maintaining perfect connection so I was getting a drop on cyl 2

After that got resolved it showed dropping out at higher boost on number 6. The injector was good and it had spark so the spark can only be weak under load. The owner had a second batch of coils so they where all replaced and problem solved and it made 450hp/[email protected] boost with safe ignition.

The drop image shows this engine but I didn´t log anything after that so I used one from a 2.8 running a 56mm compressor as well.

You could also see a cylinder which has lowered in compression vs the other ones as it would have a longer time delay.



Dropped cyl identified you can see that the time value for number 6 cyl (Individual power 5) is much lower then the others, meaning it produced alot less power then the others. it´s being time sensitive in terms of when a certain crank angle should arrive after the combustion and you can see the time delay 176us compared to the other ones.

(http://myndasafn.bmwkraftur.is/d/102290-1/droppedcyl.jpg)

Here below is the non dropped 2.8 engine you can see that the cylinders are all much much closer in terms of time. And at the end of the trace you can see how they all become equal when there is no combustion and only non fuelled overrun (i.e equal constant load between cylinders)

(http://myndasafn.bmwkraftur.is/d/102287-1/nodropped.jpg/bmi_orig_img/nodropped.jpg)

This is a very handy thing to have when diagnosing missfire issues or power loss.
Title: Re: Missfire detection, cylinder identification
Post by: mattias on October 13, 2012, 07:02:32 PM
I haven't checked with the latest firmware (1.2.6?) and VemsTune (2012-07-30) that it actually works on an engine with more than 6 cylinders. That would be great  as I deal with V8s on a pretty regular basis as well. It would be nice if the "human-computer" interface was a bit improved as well, to make it easier to spot the correlation between which cylinder you're looking at and which "individual power" value it belongs to.

It's good that you mention this "tool", it's valueable help when things aren't working as they're supposed to. I can add that you can spot wrong polarity on VR trigger systems, one pair of cylinders will be a little off compared to the rest.

Title: Re: Missfire detection, cylinder identification
Post by: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.

Title: Re: Missfire detection, cylinder identification
Post by: audis2turbo on October 16, 2012, 05:00:42 PM
Hello,
what say us these Value Range for the Individual Power ?
When the Value is negative the Spark it is earlier than time to Sparkevent???

It is Knocking?

Thank you
Marcus
Title: Re: Missfire detection, cylinder identification
Post by: Sprocket on October 16, 2012, 07:00:50 PM
individual cylinder power is shown in bar graph form in the 'injector calibration' gauge group.
Title: Re: Missfire detection, cylinder identification
Post by: audis2turbo on October 16, 2012, 07:22:44 PM
Sorry Sprocket,
i don`t understand these values.
For what is it good for?
When any Cylinder the Spark comes later or earlier???   In time : xxx  uS
Title: Re: Missfire detection, cylinder identification
Post by: Sprocket on October 16, 2012, 07:58:14 PM
Quote from: audis2turbo on October 16, 2012, 07:22:44 PM
Sorry Sprocket,
i don`t understand these values.
For what is it good for?
When any Cylinder the Spark comes later or earlier???   In time : xxx  uS

It is an indication of combustion efficiency.

I'm not sure how it is calculated, but it is derived from the acceleration of the crankshaft during each individual cylinder combustion stroke.

Therefore, if you have dramatic differences in values between cylinders, combustion is inconsistent.

As can be seen in the gragh above individual power5 is clearly not performing, and was found to be miss firing. Basically, no combustion with no crank acceleration as a direct result (most likely decelleration). Individual power3 shows poor performance compared with the other good cylinders. This could be anything such as poor combustion, lean mixture, poor cylinder sealing, poor spark, basically anything that would atribute to poor combustion, and thus lower crankshaft acceleration.

worth noting at this point, I am not sure how these individual power values are related to the ignition chanels, whether directly linked to the ignition output configuration (Individual power0 is ignition chanel 0 ?? ) or refference tooth table (individual power0 is first event in refference table ?? )

At the moment it doesn't matter what the actual value means, only if there is a difference, indicating an issue on the cylinder.

Knock I suppose could be detected, but then you would need to be able to understand what you are looking at. At a recent seminar at Swansee Metropolatin University, we played around with an engine on the dyno with in cylinder pressure sensing. Short of pushing the engine into detonation, the ignition was over advanced to demonstrate what happens. The cylinder pressures became erratic and un predictable, ether happening earlier or later, with a noticable loss of power. I can only assume this would translate into varying crankshaft acceleration per cylinder ( the graphs would be more peaky, than a straight line). Taking the extra advance back out made combustion more stable and repeatable with peak cylinder pressures around that magic 14 or 15 degrees ATDC.

Bottom line. It is a comparison of cylinder performance, nothing more. the same could be shown with EGT sensors  in each exhaust port, or in cylinder pressure sensors in each cylinder.
Title: Re: Missfire detection, cylinder identification
Post by: Sprocket on October 16, 2012, 07:58:53 PM
Quote from: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.

dont forget cylinder 0 ;)
Title: Re: Missfire detection, cylinder identification
Post by: audis2turbo on October 16, 2012, 08:21:06 PM
I thank you for this good description!
This I have got on well with my poor English skills ;-)
Here i find the Outputs from Ingnition Out to Cylinder Individual Outputs : http://www.vems.hu/vt/help/v3/datalog/individual_power.html
And I'll time my whole logs to check :-)

Maybe someone can say something about the values in the comming days

Greets Marcus

Title: Re: Missfire detection, cylinder identification
Post by: mattias on October 16, 2012, 09:21:01 PM
Quote from: Sprocket on October 16, 2012, 07:58:53 PM
Quote from: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.

dont forget cylinder 0 ;)
It actually only goes from "individual power" 0 to 6.
Title: Re: Missfire detection, cylinder identification
Post by: Sprocket on October 17, 2012, 10:14:40 PM
Quote from: mattias on October 16, 2012, 09:21:01 PM
Quote from: Sprocket on October 16, 2012, 07:58:53 PM
Quote from: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.

dont forget cylinder 0 ;)
It actually only goes from "individual power" 0 to 6.

Oh yeh, bollox, I never noticed that ::)

4 is all I need ;D

Edit* actually no ??? there are 7 cylinders in the Injector calibration gauge group. A mistake perhaps.
Title: Re: Missfire detection, cylinder identification
Post by: GintsK on October 18, 2012, 06:58:07 AM
As far I knew it is because of some software or hardware limitations.
Title: Re: Missfire detection, cylinder identification
Post by: VemsedE36 on July 31, 2013, 06:12:20 PM
Quote from: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.
Gunni, have you solved that issue? What was that issue?
Title: Re: Missfire detection, cylinder identification
Post by: gunni on July 31, 2013, 06:56:41 PM
I´m not involved in solving it, I believe the developers are aware of it.
Title: Re: Missfire detection, cylinder identification
Post by: VemsedE36 on July 31, 2013, 07:02:28 PM
I have -400uS in cylinder 3+4   ???  :o
Title: Re: Missfire detection, cylinder identification
Post by: mattias on July 31, 2013, 08:28:25 PM
Quote from: gunni on October 13, 2012, 10:37:55 PM
It only seems to go up to cyl7.
Dave (NL) has solved this issue, it'll have to wait until 1.2.15+ firmware and new VemsTune is released.
Title: Re: Missfire detection, cylinder identification
Post by: Erikk on August 19, 2013, 08:58:25 AM
Quote from: VemsedE36 on July 31, 2013, 07:02:28 PM
I have -400uS in cylinder 3+4   ???  :o

The vr-sensor polarity is switched. As a result, you get a tooth retardation ( 6deg ) on two cylinders.
If you are using the vr-sensor from bmw m50, then connect yellow to primary trigger input and black to sensorground.

Black as signal is for m20 and m30 engines.


Title: Re: Missfire detection, cylinder identification
Post by: GintsK on August 19, 2013, 11:32:04 AM
Erikk - on some wheels first tooth after missing is little disturbed anyway (e.g. 1.8T IIRC)(Clio 2L 16V wheel by default). First trigger tooth defined some teeth later. Cylinder power shows significant time difference. Two questions:
1) why cylinder power do not use First Trigger Tooth + reff tooth table: that signals are evenly spaced?
2) in this particular case - I hope ignition events are based on given trigger tooth + reff tooth table and not disturbed by one not so good signal?

Gints

Title: Re: Missfire detection, cylinder identification
Post by: Erikk on August 20, 2013, 06:12:17 AM
Quote from: GintsK on August 19, 2013, 11:32:04 AM
Erikk - on some wheels first tooth after missing is little disturbed anyway (e.g. 1.8T IIRC)(Clio 2L 16V wheel by default). First trigger tooth defined some teeth later.
Is this the same 60-2 wheel that we install behind the flywheel on Audi I5 with hallsensor?
It has a tooth thats smaller than the rest, so it needs to be triggered from falling edge, else it will to the same as runninig bmw with wrong polarity.




Title: Re: Missfire detection, cylinder identification
Post by: GintsK on August 20, 2013, 11:24:27 AM
No, not that wheel...
But it does not matter. Questions is more about VEMS in general than about specific wheels.
Title: Re: Missfire detection, cylinder identification
Post by: Matus on August 26, 2016, 09:06:17 AM
I had cylinder that wasn't running from the beginning but I didn't realize that as you know it is not that obvious especially when engine is not tuned yet. I found out it coincidentally when was checking plugs .. one piston was shiny and the rest was black.

So this is nice and and really useful topic for me :D

This would be proper reading for V8.

    Ign_out0 ->individual power2
    Ign_out1 ->individual power3
    Ign_out2 ->individual power4
    Ign_out3 ->individual power5
    Ign_out4 ->individual power6
    Ign_out5 ->individual power7
    Ign_out6 ->individual power0
    Ign_out7 ->individual power1

As I understood value shown in time value is just difference but now major question is which values are differentiating.
Does it matters about aliment in the log ? for example.. lets say that I put together accidentally two bad cylinders together so difference might be small ?

Or does it means that value represent acceleration and "difference" is just optical result ?  ( I will bet on this one however I would like to be sure because nothing is worst than bad data interpretation   :D )

Title: Re: Missfire detection, cylinder identification
Post by: Mads Lund on August 29, 2016, 09:19:43 PM
Regarding individual power measurement, never use 1'th tooth as trigger tooth. I made a post som time ago on this topic. In short, the missing tooth will offset the voltage of the VR sensor slightly, compared to the smaller gaps between the othr teeth's. This will affect the analog signal a littel, and postpont the trigger event on the first tooth a few microsec, enough to make individual power useless.
So never use 1'th tooth as reference, use second or higher. On mine v6 duratec, did I have to ofset the complete cycle, and have cylinder 2 as the "no 1" cylinder, worked but anoying. So at the end did I rotate the trigger wheel 40 deg to solve the issue.