Dwell max dur & max sprk dur; clarification neded.

[moteleb]

Hi tuners.

I recently made a setup on a VW turboed BEV engine using 4 independent ignition coil with internal powerstage just like this http://c1552172.r72.cf0.rackcdn.com/202580_x600.jpg

I running it waste spark mode by plugging 1-4 & 2-3 drive together...

I've set Maximum dwell at 2.5ms & Spark duration at 2ms and after 2 month of success (normal drive) I went to a race track to push it a bit more. Knocks occurs at 4500 rpm or more and finally broke pistons 1 & 4 at the same time.

After rebuilding the engine I tried it on a close road again with same problem. I decided to try several thing to solve issue by retarding max load/rpm degree, running fixed advance (timing light tested etc)...

At the end I just reduce dwell time & spark duration to 1 and 0.5ms, the problem appear to be gone and it match better with timing light and timing map
I use a standard 60-2 trigger wheet that first tooth is exactly 84° from sensor at tdc.

Now; why 2.5ms dwell with 2ms spark duration cause problem? Help Youston.

Best regards.

Moteleb

[ashford]

dwell time is maximum dwell. spark duration is minimum time between coil charging.

if dwell was fixed and rpm exceeded time of dwell durung firing events the coil would be charging and never fire.

minimum spark is basically "there must be this much time" between the firing and the beginning of the next charge, it has priority over dwell time so if your dwell time is 4 ms and the time between firing events on the coil is 4 ms and your spark duration is 1ms the coil will be dwelled 3 ms. or if time between firing events is 3 ms and spark duration is 1ms actuall dwell is 2ms

[moteleb]

Sure but I have a 4cyl wasted spark (each coil should spark every 360° ?). If yes, at 4500rpm I have 13milliseconds firing events ?
I calculate for 7000rpm=8.5ms every event (2.5dwell+2spark duration) look conservative ?

I use a 60-2 trigger Wheel, Is MS2/extra really calculate spark event every 360°?

[ashford]

Sure but I have a 4cyl wasted spark (each coil should spark every 360° ?). If yes, at 4500rpm I have 13milliseconds firing events ?
I calculate for 7000rpm=8.5ms every event (2.5dwell+2spark duration) look conservative ?

I use a 60-2 trigger Wheel, Is MS2/extra really calculate spark event every 360°?

its quite possible your coils don't like that much dwell, for instance certain ls coils self protect/discharge after 4ms when hot basically advancing the timing. you might be running into the same problem if your coils have built in ignitors

[moteleb]

Make sense, I need to use normal spark output instead reversed. I also scrap another ing coil, This coil pack have built in ignitor just like you said.

Now I have to find how exactly this coil should be dwelled... Maybe someone on this forum can check his VW bev drives from stock ecm.

[moteleb]

Hi tuners,

After scope testing and several tries, my conclusion is; theese coils from vw bev engine have internal protective circuit that make dangerous spark advance If dwell is higher than something 5ms. You'll not see it with a timing light but over 4000 rpm big spark advance make engine pigning...

I also remark the dwell gauge exactly match the spark output (Checked with scopemeter) Make sure you have around 2.5 ms on this gauge.

Dwell settings sould be like this

Standard dwell
max dwell 1.4ms (Dont put 2.5 because will give more)
sprk dur 0.5.

Actually running dwell should be close 2.5 ms all time.

Regards

[16v dude]

Sorry to hear about the pistons. Have you ever thought of Knock control? Do you have a knocksensor installed on the engine and is it hooked up to the MS? MegaLogviewer can datalog knocks and MS itself has setting that you can change so that you can protect your motor from knocks.

[jsmcortina]

Standard dwell
max dwell 1.4ms (Dont put 2.5 because will give more)

That depends on your reported battery voltage and the correction curve.
It sounds like perhaps your Megasquirt is reporting a lower voltage and boosting the dwell to compensate. However, if the actual voltage at the coil is higher then the dwell will be incorrect.

James

[moteleb]

I just comeback playing with values.

I set overdwell & underdwell to "on" To make sure dwell stay fixed

by trying different values I've got this

If RPM < 4500= no Knock regardless dwell value or timing
If RPM > 4500= Knock If dwell is 2.5 or more (even if timing was set very low like 5°)

If Dwell is low as 1.5ms everything works great regardless RPM but Something around 5000rpm, power decrease like an engine running too rich. AFR is 11.5-12:1 Sometimes it run on 3 cylinder (like fouled plug) and comesback 4cyl after 2-3 seconds (sorry for my very bad English)

I dont use any Knock control system module or wathever that affect timing or dwell. Knocks are clearly audible from Inside the car "krrrlick!".

Another question: Normally that coilpack has 6wires; One for voltage, One for ground; 4 other ones are normally driven independently & probably sequentially by the stock ECM. Since I plug 1-4 & 2-3 together (parallel), it is possible that's not good?

Best regards

[ashford]

try closing the gap on the spark plugs

[sleeper]

Hi tuners.

I recently made a setup on a VW turboed BEV engine using 4 independent ignition coil with internal powerstage just like this http://c1552172.r72.cf0.rackcdn.com/202580_x600.jpg

I running it waste spark mode by plugging 1-4 & 2-3 drive together...

I've set Maximum dwell at 2.5ms & Spark duration at 2ms and after 2 month of success (normal drive) I went to a race track to push it a bit more. Knocks occurs at 4500 rpm or more and finally broke pistons 1 & 4 at the same time.

After rebuilding the engine I tried it on a close road again with same problem. I decided to try several thing to solve issue by retarding max load/rpm degree, running fixed advance (timing light tested etc)...

At the end I just reduce dwell time & spark duration to 1 and 0.5ms, the problem appear to be gone and it match better with timing light and timing map
I use a standard 60-2 trigger wheet that first tooth is exactly 84° from sensor at tdc.

Now; why 2.5ms dwell with 2ms spark duration cause problem? Help Youston.

Best regards.

Moteleb

I'm running boosted VW engine with it's stock 60-2 crank wheel too but I find some strange things in your config.
Most stock 60-2 VW/Audi crank wheels have their tooth #1 angle at 78 degrees (yours is at 84) I'm also running 78 deg tooth #1 angle and it's aligns perfectly when checked with timing light
Other thing, you are basically running 4 coils as 2 coils so you really don't need that pack if you don't intend to fire each of the 4 coils individually somewhere in the future
I recommend switching to coil from 1.6-2L Golf mk4 or Polo. It also has internal ignitors and it's a typical wasted spark coil for 4 cyl engine (2 coils in one pack)
I'm running this coil on my setup (3ms dwell and 1ms spark duration) and it's perfect ! Coil is VW OEM# 032905106B and it looks like this



If you decide to go for this coil, here is the link to my old thread when I was first installing this coil. See my last post in the thread since it contains directions how to modify MS board to make this coil work with Megasquirt

http://msextra.com/forums/viewtopic.php?f=101&t=49837

[moteleb]

I juste comeback to take some picture of tach signal youre correct. Tooth #1 is really at 78° instead 84° (I used 84° to match the timing light)

THis signal look's bad I gess. I'll try with another scopemeter just to make sure... Harness & tach sig wire is closely sparkplug wire. It is possible that is the cause?

[moteleb]

My older scopemeter show better sine wave. I"m now set tooth #1 to 78° actual timing soult be 15° but I have almost 30°! I can't wait to solve that issue.

[sleeper]

How did you configure you MS board for tach input signal ?
Most VW engines (BEV engine too, I believe) run Hall type crankshaft sensor, not VR as most people think
You should configure your MS board for hall input (step 26 in MSextra assembly manual) and ignition input capture should be set at "Falling edge"

I would also change that coilpack you are running to something like I suggested or similar. Just to exclude the coilpack as a possible cause
If you have licenses TS, you could make some logs with tooth logger, just to make sure you have nice and clean signal

Here is my config, I'm running 2.0 16v ABF engine. Yours should be configured identically I believe

inition.png

[moteleb]

My settings are exactly the same as youre.

Each year I make 2 car mean with megasquirt. Almost 10 success stories, every car run well (but not mine) I also made several 1.8T that work great with coil on plug. When I try something new, I use my car for R&D

you have to know this
G28 (crank sensor) is variable reluctor type sensor third wire is a shield ground.
G40 (Cam sensor) Is hall effect sensor



Finally found a link from diy autotune that said "put a 10K or 20k resistor on tach signal for 60-2 trigger wheel from bosch motronic"

http://www.diyautotune.com/tech_article ... _index.htm

[moteleb]

Still have lack of power at 4500rpm or more. I gess Ignition power is not enough to burn mixture at high rpm/boost. When I reduce the boost, I can go higher rpm...
like I said previously, too much dwell cause too much timing. probably theese coils have current limit regarding rpm value

I decide to destroy an old coilpack to get information about his construction.

I've found Integrated circuit like this
http://www.st.com/web/en/resource/techn ... 29307.pdf‎

pin #8 have a condensator but dont know value, i'll try to get it.

Hopefully this thread will help Future VW squirters.
Regards

[sleeper]

Is it backfiring past 4500 rpm or just loses power ? If it's backfiring it could be your spark plugs.
Also, some people reported power loss at high rpms caused by badly adjusted spark hardware latency.
You can check this with timing light and locked timing. If it starts retarding as rpm's rise then you need to adjust your spark hardware latency to keep timing solid all the way to the redline

BTW, I'm running 1,2 bar of boost all the way to 7000 Rpm with no problems and with coil pack I mentioned in my earlier post.

[moteleb]

I loose power then run on 3 cylinder until comeback after 3-4seconds of deceleration (AFR is 12.5:1 during acceletration)

To match the timing light I absolutely need to put 84° as the #1 tooth. look's good at all rpm range with 2ms maximum dwell. If I set more dwell, dangerous spark advance occur. I doublecheck this at fixed timing, I remark a slight retard with rpm something like 3° less at maximum rpm. With 3ms dwell timing mark disapear fast where is too much.

I really gess internal powerstage from BEV coil make something wrong if we intend to give to much dwell.

I'll try a coil from aeg engine (like the one you shown to me, I'm sure this unit is made to work wasted spark) And you have one like this working well.

I've also set a 16V turbo with 1.8T's coil on plug that running great since 3years at 1.7bar of boost.

[sleeper]

You should definitely go for AEG unit. Yes, it's the same as mine and it's a classic wasted spark unit
You will need to modify logic output of your MS board to run with 12v pull up (msextra guide gives a guide only for 5v pull up) as this unit will not fire with lower pull up.
Pinout of the unit and schematics how to modify the MS board can be found at the link to my thread i posted here earlier

[moteleb]

I just finnished to swap it. It run great at idle (Even if I use 5volts trigger) I've set max dwell to 2.5ms, Coil & powerstage stays cold

I wait for the coil holder.... Tonight is the road test at 5000 or more to see the difference.

After some reflexion, BEV coil is made to run in sequential mode. Because I made it waste spark , internal powerstage think the rpm is twice

By calculation; each internal powerstage (each intelligent coil driver) are made for something close to 80hz where dwell is internally tweaked depending rpm(frequency)

VW BEV coil pack woul be great only for sequential ignition (not good for waste spark)