Automobiles Forum banner
1 - 8 of 8 Posts

·
Registered
Joined
·
15 Posts
Discussion Starter · #1 ·
Here's an interesting one for you guru's. How much trigger delay is there on a Bosch 0-227-100-211 ICM between when the ECU triggers ( input side ) to when the output stage triggers the coil. I already know that the 211 does not control coil dwell.

Here's the scenario. I recently converted my 1998 Audi A4 1.8T ( AEB ) from the ICM triggered " Dumb " Audi/Beru 3 wire coils to an MSD DIS-4 system. I used the IGN-CD coils from DiyAutoTune and bypassed the ICM. I triggered the MSD box directly from the ECU triggers.

The results. MSD fired perfectly...but MASSIVE over-advance and detonation!! Knock sensors maxed out, timing pull maxed out and I had to reduce boost to 5lbs ( MBC ) with 94 Octane Chevron and Octane booster just to make it drivable. Detonation became worse as RPM climbed...even with boost limited to 5 psi.

After some deliberation, I deduced that the ICM must have some sort of delay inherent to its design and the ECU timing map was adjusted to compensate for that.

Putting the ICM back into the circuit solved all of the problems. MSD still triggered properly, but all detonation was now gone, and I could run 20 lbs boost with no issue. Minimal timing pull and all knock sensor values now normal.

So the question is. Does anyone know just how much delay is inherent to the Bosch 211 ICM? As a educated guess... ( 40 years of working on cars ) I would estimate that the engine was running with about 15 to 20 degrees too much advance. Difficult to get a timing light on these things ( no real ignition timing marks on crank pulley ).
 

·
Registered
Joined
·
2,486 Posts
IMHO delay is probably not causing what you're seeing. A module like that might have 5 microseconds of delay, but for comparison 24 microseconds is one degree at 7000 RPM.

What's much more likely is a signal-inversion problem. Because of the changes, the coil is now being turned on when it should be turned off, and vice-versa.

So here's the scenario: in normal operation, the ECU turns on power to the coil so it can start charging a few milliseconds before the firing point... call that time "T1". Then a few milliseconds later when the spark should occur, the ECU turns off power to the coil and it makes a spark... call that time "T2".

Now suppose the signal to the coil driver is somehow inverted - meaning the "on" and "off" are reversed. That means at T1 when the coil is supposed to be turned on, it will instead be turned off and generate a spark. Since T1 comes a couple milliseconds before the actual firing point, what you get is way too much advance.
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #3 ·
Thanks for the explanation. :cheers:

I'm just glad it's finally fixed. Went through a frustrating time late at night, when the engine wouldn't fire with the MSD hooked up to the ICM. Then I had one of those " Eureka " moments at 3:00 AM. I'd gotten the trigger wiring mixed up and they were 180 degrees out of phase. S**t happens when you're tired :x

This is my second attempt at building a higher output ignition system for my 1.8T. First attempt with LS2 coils failed. Reason being is that my early model ECU ( Bosch ME-5.92 ) uses a fast saturation coil. Dwell time is a maximum of 2ms and runs as low as 1.4ms. Later model ECU's ( Bosch ME 7.0 ) have a 3ms dwell time with a minimum of around 2.4ms. This is enough to make the LS2 coils work ( they are optimized for 5ms, but still work very well at 3ms ), but unfortunately 2ms is just too short. ( The dwell time on ME5.92 was not something that was easy to find at first go round )

Of course the ME-5.92 has to be a PITA ECU. The ME-7.0 is a joy. You can read the maps and flash then no problem, and a ton of info is available. Maestro, Unisettings, and NefMoto Flash ( freeware ) can easily tweak the dwell tables ( or anything else ) on an ME7.0. The ME 5.92 ( which is cannot be swapped easily to an ME7.0 ) is flashable only. You can't read the chip.... unless you de-solder it from the board and read it through a programmer. What a PITA. I've got a couple of chipped ECU's that use a separate daughter-board and a standard AMD EEprom. Problem is that those are encrypted. :curses:

Was turning out to be more of a headache than I had planned, then I came across an MSD-DIS4 for a good price ( $199 ) on E-Bay. Changed out the coils LS2 coils for the IGN-CD ones and got this up and running..after a few hiccups.

Any one want to buy some slightly used LS2 coils? :cheers:
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #4 ·
Well, rewired the DIS-4 though the Bosch ICM. Almost works perfect. Engine pulls hard all the way to redline with 20 lbs Boost and .032" spark gap. I have a feeling I can go higher.

One slight problem. Get a severe misfire under very light throttle, particularly after coming off of a high boost situation.

Have narrowed it down to the DIS-4 requiring a minimum of 2.5ms STATIC dwell time ( works at 2.0ms ). Bosch ECU uses a dynamic dwell table, varying dwell by several variables including coolant temperature, engine boost and RPM.

At low speeds and no load, the ECU doesn't supply enough dwell time to send the DIS-4 a clean signal. After high boost and returning to light throttle, ECU goes into a coil cooling strategy with reduced dwell time. I believe it is dropping into the 1.3 ms range. So more head aches. :frustrated:
 

·
Administrator
Joined
·
201,871 Posts
Are you working with some canned AEB file, or tuning it for this setup? The dwell maps shouldn't be hard to bump up/flatten if that's all you need.
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #6 ·
newt said:
Are you working with some canned AEB file, or tuning it for this setup? The dwell maps shouldn't be hard to bump up/flatten if that's all you need.
I do have all the flashing cables ( MPPS v13 ) and Software ( TunerPro, WinOLS Pro ) and even have the hardware for a bench flash unit.

Problem is that I'm working on a ME5.9 ECU with an APR Stage 1 file. So facing the APR encryption and the lovely problem of the ME5.92 which cannot be read unless you de-solder the chip from the PCB. If I had an ME7, which is a readable and flashable, I could easily read the Bin file, alter the dwell maps and be done. On an ME5.9 it's not so easy and not many people work on them any more.

I may have found a better way. I'm looking at a Kenne Bell " Boost - A- Spark ". This system is widely used by tne Forced Induction Mustang, DSM and Honda Forced Induction crowd and gets very good reviews. It works on the principle of boosting the supply voltage to the coils when under boost. Comes with a standard Hobbs switch set at 4 psi ( I'd use a 10 psi Hobbs switch ) and the amount of voltage increase is adjustable. Since it only operates when on boost ( In my case over 10 psi ) the coils are not subjected to a constant increase in voltage.

http://www.kennebell.net/KBWebsite/Acce ... aspark.htm" onclick="window.open(this.href);return false;
 

·
Administrator
Joined
·
201,871 Posts
Ok, so it's just a canned stage 1 file. The APR files for those AEB's are loaded onto EMCS hardware modules.

Which leads me to my next question..... why are you going through this much trouble to increase spark energy on a stage 1 file?

These hardware stick-ons seem like the absolute last thing you would need, especially if you have the flashing tools you claim to have at your disposal, just grab a second ECU off ebay and start tuning it properly. Unless I'm missing something?!
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #8 ·
newt said:
Ok, so it's just a canned stage 1 file. The APR files for those AEB's are loaded onto EMCS hardware modules.

Which leads me to my next question..... why are you going through this much trouble to increase spark energy on a stage 1 file?

These hardware stick-ons seem like the absolute last thing you would need, especially if you have the flashing tools you claim to have at your disposal, just grab a second ECU off ebay and start tuning it properly. Unless I'm missing something?!
Thanks for the replies. I'll try and explain further. I've been researching the peculiarities of the ME5.9 ECU and re-flash capabilities for about a year now and keep discovering new quirks of this ECU.

Reason for upgrading ignition. Because I have a BT setup waiting to go on and I wanted to upgrade ignition and test before hand. And with my current setup I've added a slight amount of pre-load to the stock KO3. With my other supporting mods ( And 94 Octane pump gas ) I can spike to 20 psi in 3rd gear and hold 17 psi all the way through 6,000 tapering to 14-15 psi at 7,000 RPM just on a Stage 1 file. The stock ignition setup is inadequate with misfires over 16 psi ( NGK BKR7E at .026" to .028" ). Yeah I can gap the plugs down to .024" and solve the issue, but I consider that a band-aid fix as do others. I Autocross this car and Track Day it on occasion.

As to why I don't just flash a file?

1: As mentioned the ME 5.9 is a PITA to flash. Chip is not readable when installed on PCB. Not many people work on these any more and I wanted to go a different route by improving the hardware. Unfortunately it didn't work out. It's been an interesting journey though.

2: Same problem with APR. APR module does have EMCS, I even have the APE file for the Stage 1 from APR. But I've been unable to read it and get past the Encryption. And forget about APR giving you a custom flash...they just don't want to bother. Been there..done that.

3: I also have a GIAC 440 cc file ECU and a spare Stock 557P ECU. Still have the issue of having to de-solder the Flash chip, either buy or find someone with a Willem programmer that takes a PS 44 chip ( AM29F200BB ) and then air-solder the thing back on. IF the LS2 coil conversion or MSD conversion had worked it would have been a cheaper and easier solution. THe GIAC has a separate MB and I have spare AM29F200BB chips ( empty ). Unfortunately a buddy of mine, while trying to help out, broke 3 legs off of the GIAC Chip. Can't read the file now, until I repair chip ( if I can even do that ). And I still have to find a proper 44 pin flash programmer such as a Willem. Just getting to be too much bother..

4: I originated an LS2 coil conversion thread on the Vortex for the 1.8T engines. The LS2 coils work fantastic...on ME7 ECU's, just not ME5.9 though ( as I found out after the fact ). In fact I got the original idea of using LS2 coils from the AAN conversions done here on MotorGeeks. DSM crowd and RX8 crowd have proven it's a worthwhile conversion.

5: Got a good deal on an MSD DIS-4, and thought that would be a solution after the LS2 fail. But no..the short dwell of the ME5.9 brick-walled that idea as well. BTW...thank you MSD for not documenting the minimum dwell requirements for your products. Their Tech line is useless and won't even bother answering your questions if you do not have a product that is not a current Production model ( My DIS-4 is an early model ). I finally found the information about the 2.0 to 2.5ms minimum well requirement on a HalTech DSM forum :frustrated:

6: This adventure has cost me a bit of time and money, but I can still recoup some of my money by selling off my LS2 coil conversion on the Vortex and I'll have no problem selling the DIS-4 locally.

7: I may still give the LS2 or MSD hardware another go, if I can find someone local to give me a hand. I haven't exhausted that avenue yet...but personal issues with family and health have limited my time to work on car and play with the ECU's. I believe that Tony from NefMoto lives near me...but the family and health issues just haven't allowed me time to run around a lot. I've been collecting parts ( Flashing equipment etc ) for quite a while now, but am relatively inexperienced at flashing. I really need some help locally...but circumstances have limited that avenue.

8: The Kenne Bell Boost A Spark is a proven sound system and I just may give that a go to see if it works for the ME5.9 group.

Whew... sorry about the long novel. But I thought I may as well tell the whole story from the beginning.
 
1 - 8 of 8 Posts
Top