Dwell control apparently NOT working....

T3Bunny · Post by **T3Bunny** » Thu May 04, 2006 2:42 am

Okay, I am at a loss guys. I am gonna restate some of the details I stated in Dwell setup questions. as we still have not managed to figure this issue out.

Again we are running a V3.0 with MS-1 and code MsnSe code 027d. We are trying to get direct coil control on a 4cyl VW 8valve motor. The trigger setup is a hall, modded the normal way for VW's with the signal inverted. We do this by powering the hall with 5v out of pin 24, then bring it back in on pin 25 and jumper that over to XG1. I have worked on directly or helped setup a couple dozen cars this way.

The ignition output signal is taken off the led 17/D14. Again modded per normal same as every other setup I have done. Now stuff starts to differ from the V2.2 as this is the first V3.0 I have personaly setup this way.

I used a jumper over to the IGBTIN. The setup info I was using didn't mention using a 330ohm resistor. So I didn't at first. Also I was going by the build manual and had R57 installed. We set the output in MT for inverted to directly drive the coil. Car ran like crap. At about 3.0ms you can see the spark dissapaering (with a timing light) at about 2800rpm and the car sounds like its hitting a rev limiter. Went up and down the numbers a few times. At 6ms running dwell, we get a clean spark singal according to a timing gun, but misfires, of course at 7ms its worse. Understandably. Since then we have tried every other combo. R57 has been removed, and we have put the 330om resistor inline to the IGBTIN pad.

But we can NOT find a location between that works. Either its dropping out the spark, or we are getting misfires. I have built a couple base .msq's for these cars and the V3.0 ECU's with the default settings left as is, and never had a problem. All fired up and ran fine. The only differance I can think of is the default setting for minimum discharge is .5ms. When we had issues, we swapped it to the .1ms and didn't fuzt with that setting any further. So most of the testing was done at .1ms.

SO, any pointers on a direction to go? I have poured over all the info I can find. It seems to me that I am missing something. I suspect its a hardware issue at this point, but we have tried every combo mentioned.

We have been struggling with this for two weeks now and not getting anywhere. The only thing that comes to mind is referances in other EFI systems of a minimum coil resistance. And the fact that I don't understand how or IF that would effect the minimum discharge time of the coil.

Finaly the fallback question. I don't have many options left now unless I have missed something stupid. If we swap over to a smart Bosch ignition module, I need to change the setting back to non inverted. Will this cause an issue with retaining the high current ignition driver setup? I know if a coil is attched it SMOKES the VB921 nicely...

But would flipping it cause issues when connected to the Bosch Module?

I am going to try and reproduce these issues here on my own setup here if I don't get any further, but I am running a v2.2 and its going to require LOTS of harness butchery to do it. I am also going to get some datalogs, but I can't think of anything it would tell me that I don't already know...

Post by **jsmcortina** » Thu May 04, 2006 5:06 am

Any reason to use 027d ? That's now obsolete.

James

T3Bunny · Post by **T3Bunny** » Thu May 04, 2006 7:39 am

Paul_VR6 wrote:What coil are you using?

We have tried to use both a Jaocbs coil, and when that was giving us issues we swapped back over to the stock 8v CIS coil.

jsmcortina wrote:Any reason to use 027d ? That's now obsolete.

James

Mainly simplicity and reliability. We know it works and have never had an issue with it, YET. I have had some troubles with newer codes on the other hand. For my own cars and setups, I will usally keep upgrading to newer code. But I have a bit of practice getting them smoothed out and communicating.

If there is a good reason to beleive its a code issue, I will happily upgrade to a newer recomended code (or the newest one). From a remote setup standpoint though, I kinda just stuck to the same code for simplicity.

If there is a reason to suspect the code, we will give that a try later today. At this point, I think I will try that ANYWAYS.

woh · Post by **woh** » Thu May 04, 2006 7:55 am

This connection for the output at bottom of page is known to work.
http://home.comcast.net/~whaussmann/MS_ ... ign3_a.htm

Is there a ballast resistor for the coil or is it part of the coil?
Do you have access to a scope?

Measure the current using R43 (measure the voltage at R43 and divide by .01) It should be several amps when the coil is on.

Measure the voltage at the "-" terminal of the coil. It should be near 0 volts when the coil is on.

You can measure these values with a DMM by disconnecting igbtIN and connecting it to +5V momentarily with a 330 ohm resistor and testing with a DMM. Wait 5 min between tests to let the VB921 cool. This does risk blowing the VB921 especially if you do not have a ballast resistor or momentarily is too long.

T3Bunny · Post by **T3Bunny** » Thu May 04, 2006 8:20 am

woh wrote:This connection for the output at bottom of page is known to work.

Okay, no we were using a variation on that. I used the same setup we used on the V2.2 when modding the LEDs for a spark output. I am starting to comprehend the differances that could cause the issues though. BUT (in my defense lol!) some of the other VW guys had already gone down the road and I was just following in their footsteps.

My scope got fried and is out for repairs at the moment. I actualy would be VERY interested in seeing good and bad scope patterns though and an explanation. I am still in the learning stages of scope usage.

I honestly don't know on the ballast resistor. There is NOT a visable one sitting on the side of the stock coil. As for the Jacobs coil, its encased. So where would I measure and what sort of measurments are we looking for?

I am off to go reproduce this setup on my car in the driveway... It should make my comprehension of the going ons a lot easier!

woh · Post by **woh** » Thu May 04, 2006 8:39 am

What is a Jacobs coil?

Measure the coil primary resistance (from + lead to - lead) it will be ~1.5ohm if it is a 6V coil and ~3ohm if it is a 12V coil or a coil with a ballast resistor internally.

The ballast resistor might be in the wiring. So connect the coil the normal way, and short the - lead to ground. Measure the + lead. If the + lead is less than 10V (ignition on, with good battery, usually 7-9V) there is likely a ballast resistor in the wiring.

woh · Post by **woh** » Thu May 04, 2006 9:25 am

Here is a picture of a scope trace of the coil current.

The top trace is the coil current. It builds up to a max of almost 4amp. The dwell time needed to reach 4amp is about 5ms. This is longer than most because it has a ballast resistor. I chose to leave the ballast resistor and original wiring as initially installed in the car. This is simpler and a bit safer.

The bottom trace is the tach signal. I think at IRQ1 (I'm not sure, it was taken several months ago)

T3Bunny · Post by **T3Bunny** » Thu May 04, 2006 10:53 am

Well its safe to bet, there is NO ballast resistor in the wiring....

I made the whole engine managment harness for this car.

Well the mods we have done are definatly questionable. So I think we will start with redoing those and see what happens.

Sice I do have a KV tester.... Anyone know what output I should be seeing on the wires? I am thinking that should give me about the same idea as scoping the coil. Just not the pretty graph.

woh · Post by **woh** » Thu May 04, 2006 11:51 am

The voltage you see at the secondary of the coil is determined by the gap. It will not tell you if things are working correctly. It is the energy of the spark that is important. There are two questions we should try to answer.
1. Is the current through the coil correct and
2. Does the driver, VB921, pull the "-" terminal of the coil all the way to ground. And is the "+" of the coil 12V when that happens?

Another issue might be the coil??

T3Bunny · Post by **T3Bunny** » Sat May 06, 2006 11:14 pm

Well we have tried two differant coils.

At this point I am suspecting either the mods we used are wrong, or there is a defect on the board itself. I am finishing up a small jumper harness now, so I can test out a V3.0 on my own car. I should be further along with a real answer by the end of the weekend.

T3Bunny · Post by **T3Bunny** » Mon May 08, 2006 10:17 am

I am no longer sure it IS an igniion issue. I had him setup everything with a 7pin module and still having the same exact issue. The datalogs are not as much help as I would have liked, due to only a NB fitted at the moment... But we are definatly seeing the motor go lean when it misfires.

And his boost gauge disagrees with the readings in his datalogs. I am suspecting the misfire is actualy a lean misfire as opposed to too much dwell.

As for the ECU mods, I used CdnDub's directions off the Tex. I am going to go back over them though. I am currently hacking into my own harness to go ahead and test everything out on a running car. Currently my car runs a V2.2 with a 7-pin IM. And I didn't make this first harness plug and play for either ECU like I have done since then... I think my car is going to get a new harness this week!