No spark sometimes.

[ingus]

Hi. i have wierd problem with starting when it is hot. From data log i can see that ecu sends Dwel to coil ,but it dos not fire it! I can start engine with no problem when it is cold or after i shich engine off and wait about 10-20minutes(even if engine is hot ) Sometimes when starter runs and i let off ignition key engine starts(sometimes i need to pres accelerator to help engine start)

In data log what i attached in the beginning you can see that it starts no problem. The same 2 attempt to start was successful. In 3,4,5,6 attempt there was no spark. Spark start to arrived when i let of ignition key. In 7 and 8 attempt i started engine with help of accelerator pedal.
This problem reks my head for 2 weeks now
I'm running MS2 V2.87. 325 turbo BMW
Advance Ignition Options:
Trigger wheel teeth - 60
Missing teeth - 2
skip teeth - 20
Delay teeth - 15
No dual spark
Start IAT retard - 800
Full IAT retard - 1500
Signal delay Parameters
FET output delay - 0
VR input Delay - 0

BASE IGNITION SETTINGS
General ignition
Triger offset - -5,0deg
Skip pulses - 3
Prediction settings:
last interval
Tach signal masking:
time mask - 0.0
percentage mask% - 50
Next pulse tolerance:
cranking% - 80
afterstart- 70
normal running - 25
Ignition Input Capture- Falling edge
Cranking Trigger - Trigger rise
Coil charging Scheme - Standart coil charging
Spark output - Going High(inverted)

Dwell Settings:
Maximum Dwell durration - 2,8
Maximum spark durration - 2.0
Acceleration compensation - 0,6
Battery voltage compensation
1V - -4.0
2V - -2.0
3V - 0.0
4V - 2.0
5V - 4.0
durration at v 1 - 2.4
2 - 0.9
3 - 0.0
4 - -0.5
5 - -0.9

I checked ignition timing with Timing light ,adjusted offset in idle to mach readings with Megatune. Readings at 2000rpm was ok and in higer revs was little bit off. I dont think My Dwell is part of timing or maybe i'm wrong?
Realy i dont know where to start solving this problem!!!

There short video of my project car: http://www.youtube.com/watch?v=yp5c-cIPOrs

[Matt Cramer]

What exactly are you using as an ignition module?

[ingus]

ECU is controling coil. High current driver is installed in ecu.

[ingus]

I don't have seperate ignition drvier.

One more thing. In manuals i read that led (closest to DB9) have to flash when it is reciving rpm signal. My led don't flash its stays on as soon i swich ignition on and starting the car. It gose off after wile if engine is not started,but ignition is stil on.

[Matt Cramer]

Is your high current driver a VB921 or a BIP373?

The LED can sometimes blink so much it looks like it's on solid. Don't worry about that one.

[ingus]

I'v VB921 high corrent driver chip installed.

[TheMonkey]

Ingus:

very cool project.

the one thing that struck something particular with your problems is the hot start. i had something similar with ignition problem on a MS2 3.0 board. do you have access to a scope to take a look at the voltage signal sent from VR circuit to the ECU? i had a square wave signal that should have been from 0v to 5.0v, but high side only went to 3.x, and i put an additional transistor in that brought the high signal up to 5.0v which fixed the problem.

just a thought.... check output of VR signal circuit to MS2 chip.

Scott.

[ingus]

Thanks. I like old school
I don't have scope available around my area. I'm taking Pikascope courses next month and than i will buy my own scope. Cause i thin i will need it in future a lot. All this engine management tuning sh#&~ thing is very interesting. And now i've 2 costumers project cars waiting for boost Al i need now is alot and alot

In my engine i've inductive type VR sensor what can produce up to 40v in hi revs. As far i undrestand 0-5V is in hall or optikal distribitur pickup only.
Correct me if i'm wrong. i'm only student

[TheMonkey]

Ingus-

the VR sensor sends varying voltages in a sine wave INTO the VR circuit on the board. the VR circuit on the board then converts that AC signal into a square wave pattern that should be 0v low and 5v high consistently and sends it to the chip. this is the signal that i'm talking about. it is VROUT on the board.

if you don't have a scope to see the signal, you can take a look with just looking at DC voltage. the JimStim gives about a 50% duty cycle on the signal, so the voltage reading from a DVOM can be about doubled to estimate the amplitude. if it's 2.3v or higher, that's good. if it's < 2.0v, i would be suspect of it.

[ingus]

Hi. i just checked VRout put with dvom. I used JimStim. My readings was 1.8V only. What type tranzistor you used and what is the circuit lat out?
Thanks.

[TheMonkey]

i think this is good news for you. i'd bet this fixes your problem....

i wish there was an answer why some of us have low voltage coming out of VR circuit, but this fixed mine and a couple others.

mod is called the 'Krisr Mod'. I went through a very exhaustive process to diagnose and fix problem. my full thread can be read here: viewtopic.php?f=89&t=30997

get a 1k resistor, 330ohm resistor, and a 2n222 NPN transistor. in the proto area of the board, wire them like this:

disconnect jumper from VROUT to TSEL, then:
VROut through 1k resistor to base of 2n2222
emitter of 2n2222 to ground
collector of 2n2222 through 330 ohm to 5v
collector of 2n2222 to TSEL

this will also invert your signal one more time before it gets to TSEL, so change your input capture from either 'falling edge' to 'rising edge' OR the other way around. the VR flips the signal going into TSEL, but this flips it back and puts it in phase. this is just like microsquirt and sequencer because they have the additional transistor in the circuit already.

good luck, this should get you a full 5v swing.

[ingus]

Hi. I soldered circuit in the bord and i got 2.45V. When i put ECU back in the car and try to start from cold it did start perfect like before this mod. Engine run fine, but i stil have that problem. Aftrer i swiched engine off and try to start nothin happened. From datalog i cen see that Dwell go up to 7.4 than bit down to6.8(7.0-6.4 before mod) and jusk keep going like that and not firing the coil. After about 10 min it start strait away, Dwell droped after few engine revs and fired the coill!!!

[TheMonkey]

Ingus-

Sorry to hear the problem persists.

Changing the VR circuit to deliver a full 5v is a much more stable signal, so that is good to have.

I think it would be most helpful to take a look at the signal to the coil on a scope to determine that the signal is exactly what the ECU *thinks* it's sending.

remember, dwell will vary a bit based on voltage and accel. there is a table to set that. maybe set it to a straight line with a fixed dwell time while you diagnose this.

... Dwell droped after few engine revs and fired the coill!!!

sorry, i didn't understand this part... are you saying that the coil is now dead? is it possible that you had a bad coil to begin with? are you sure that your coil and dwell times are compatible? i think i would probably insert a shunt resistor in line with the power to the coil and measure the voltage drop across that to look at both dwell times and saturation at the same time.

first determine if ECU is sending expected signal and figure which side of that your problem is occurring.

just a thought, but maybe a noisy TPS signal affecting the accel compensation. try unplugging TPS?

other thing i would try is an ignition box. maybe borrow an MSD box and see if that is it.