Well, thanks to morerevsm3, I was finally able to open the ECU of my car. Just a brief history of this car...it was an import from Japan and I acquired it at 63,000 kms last october 2006. Both the Ping Sensor (all 3 of them) and the O2 Sensors were replaced in the car, and it is completely stock...exhaust and intake box as it came form the factory. Baseline dyno showed it to have 220 peak at the wheels (will get the dyno brand)...and AFR shows to be rich up the top.

I have had problems of the car knocking, because we only have 96 octane gas here. After Ping Sensors were replaced and spark plugs changed and all, the car had knocking from 1,000 to 2,000 RPMs but above that, it runs very well.

Here are some pictures of the ECU...

This is the Ribbon side of the ECU morerevsm3 was talking about...its about 1cm from the bottom part of the ECU and it had to be lifted to about 2.5 cm while the other side is stuck at the same position. It was really scary because a new ECU costs a lot of money, and we did not want to break this one!

http://i65.photobucket.com/albums/h229/gibsonyu/IMG_1038.jpg

This is the other side of the ECU that needed some pushing to unlock the upper part of the ECU. It took 2 of us to remove the whole thing...someone lifting the back, and the other one poking this side to try to push out the top part of the ECU...

http://i65.photobucket.com/albums/h229/gibsonyu/IMG_1039.jpg

Finally, we were able to loosen it up, and open the ECU to expose the Eprom where all the parameters of the car is stored. Time to prep the car for dyno tuning.

http://i65.photobucket.com/albums/h229/gibsonyu/IMG_1042.jpg

The chip is the one with the white sticker on the left side of the opened ECU.

The car is now being prepped, because we need to install a wideband O2 sensor, no farther than 8 inches from the collector of the headers.

Looking forward to start the tuning the car in the dyno in 2 days!

glad u finally cracked that open, goodluck with the project keep us updated ! ;)

I have a question....I've always assumed that the single VANOS in my car is an infinately variable cam timing system that advanced or retards the cam timing to maximise torque in the low end and high end.

Now, I have also heard from someone that the VANOS is like and on/off switch...like a bigger cam/smaller cam thing? Like VTEC of hondas? I am not sure about this, but can anyone shed light on this matter?

US spec VANOS is on/off switch..
ours is continuously variable
it's not like VTEC, it only changes the cam timing, to change the lift/duration of the cam they'd have to have different cam lobes in the cyl head like VTEC equipped cars have,
a friend has a modded type-r the stock "vtec point" is around 6000rpm .. it's very noticeable when it kicks in.. so i believe VTEC is on/off, however they can lower the VTEC point by replacing the ECU with K-Pro, honda crowd has it going easy for them!

btw i've seen a late model porsche 911 cyl head once and it appeared to have variable cam timing as well.. except it had like this rail under the chains between the two cams.. and there was this thing under the rail that looked like it pushes the rail up and down effectively changing the chain length.. (i guess ???) to change timing ??

can anybody clarify ?
and does our VANOS work in the same manner ?

sorry if i'm thread jacking but in my defense you brought the subject up :P

ITs cool, this is what the thread is for...to be able to talk in detail all these technical things so we all understand.

Yeah Honda people has it easy...I used to have a civic with a swapped K20A Type R engine in it. Got Hondata K-Pro, and the car was so easy to tune! I just plug it into my lapton via USB and i can tune it already!

Yeah that is what I suspected. The guy who told me this has a US version 3.0L engine. will find out more once the car is strapped on the dyno, and the tuning has commenced. If it is infinately variable...where does it get its inputs? I assume if it is like the VTC of modern honda engines (k-series), then it would be from engine load, or throttle position. Can anyone confirm this?

I think porsches, starting from 996 models have some kind of variable cam timing as well. Many cars actually have variable cam timing...but each one has its own way of doing this...but I think the concepts are all thesame.

I would like to call on the 3.0L tuning expert morerevsm3 :) How did you tune your VANOS?

On the Hondata K-Pro...there were 6 positions, 0, 10, 20, 30, 40, 50 degrees. One owuld have to tune the entire table to 0, then run it on the dyno, then 10, then run it on the dyno, and so on...overlay all the graphs, and you would see which rev range each position would be beneficial in, and that is what you use as your starting point. A very tedious task...

vanos chip came with alpha n chip, I just installed it

do u know of anything similar for the S50B32 ??
since the B32 has 1 ECU for both EFI and VANOS...unlike the B30 which has two different control units..

btw, a while ago i cracked open the B32 ECU out of curiosity..it had no swappable chip/e-prom (as you had told me).Had "intel" logos everywhere
here's a pic: (sorry for the quality took it with the cellphone in a hurry)

http://img215.imageshack.us/img215/9713/image017ff1.jpg

exacc, i think it gets the inputs from the cam position sensor
hense why the B32 has an exhaust cam sensor added along with the exhaust cam vanos.. (Seems logical to me)

i guess it uses rpm and TPS as well ?
i wish we could figure out exactly how the dreaded vanos works .. i'm sure i'd figure out my problem quite easily..
btw man how much did the ping sensors cost im thinking of changing mine as well.. cuz its bogging down below 4500rpm on WOT

btw isnt it weird how the B32 ecu is so tiny compared to the B30 ecu with its two layers of circuitboards and many microprocessors, and the B32 ECU contains the VANOS controls all into 1 unit ..

exacc, i think it gets the inputs from the cam position sensor
hense why the B32 has an exhaust cam sensor added along with the exhaust cam vanos.. (Seems logical to me)

i guess it uses rpm and TPS as well ?
i wish we could figure out exactly how the dreaded vanos works .. i'm sure i'd figure out my problem quite easily..
btw man how much did the ping sensors cost im thinking of changing mine as well.. cuz its bogging down below 4500rpm on WOT

it uses oil under pressure to rotate cams in relation to sprockets based on throttle position and rpm

exacc, i think it gets the inputs from the cam position sensor
hense why the B32 has an exhaust cam sensor added along with the exhaust cam vanos.. (Seems logical to me)

i guess it uses rpm and TPS as well ?
i wish we could figure out exactly how the dreaded vanos works .. i'm sure i'd figure out my problem quite easily..
btw man how much did the ping sensors cost im thinking of changing mine as well.. cuz its bogging down below 4500rpm on WOT

The Cam position sensor tell the ECU where the cam is, wether it is advanced or not. I think morerevsm3 is correct in that the ECU decides where to phase the cam, based on a combination of Throttle position and RPM...just like the hondata k-pro does it. I haven't confirmed this, but this is what makes sense to me anyway.

Because you have to think about it logically too...the VANOS will adjust the cam to give you more or less power, depending on your needs. How will the car know that you need more power, and in turn advance the cams? It is basically by the input of your foot.

The reason why they added cam position sensor on the B32, is because they need to get an input the confirms the position of the cam on the exhaust side, as it is variable as well.

ok well why did the car become slow when i removed the intake cam sensor connector.. and how did it go back to normal after reconnecting it ? if it's only there to confirm operation ...

ok well why did the car become slow when i removed the intake cam sensor connector.. and how did it go back to normal after reconnecting it ? if it's only there to confirm operation ...

It probably was not able to confirm the position of the cam, so the VANOS operation probably defaulted to a pre-designated output or something to protect the engine...kinda like "safe-mode" in windows.

So by removing the Cam sensor, the ECU is not sure if it is doing the right thing to the cam or not.

Some updates on this project...

We tried to find header gaskets, so when we remove the gasket to install a wideband O2 bung, we can reinstall it and ensure a good seal. unfortunately, it is hard to find parts here so we were not able to find one. We ended up installing a bung 10 inches from the collector of the exhaust (Pics to come) so we don't have to dismount the headers.

Changed the sparkplugs to Denso Iridium (one step colder) plugs...

We also discovered that the supplier cannot crack the code for the VANOS ECU, so we are not touching that anymore, and just focusing on the Ignition timing, and the fuel delivery.

The last time I had the car on the dyno, it was showing an AFR of a rich 11.6:1 (or .79 lambda) on the higher end (5,000 to 7,200) of the spectrum (peak power of about 220 WHP)...something that needs to be corrected to produce better power. Don't know if it is normal for this car! But the last time it was on that dyno, the ping sensors were not working properly, and the O2 sensors were not good either...so it should be better this time around.

Tunig will start tomorrow, and I will try to take as much pics as possible...

cool,
11.6:1 is too rich, N/A cars get away with ~-13:1
mine is at 12-12.5:1 @ the high end
btw no biggies about the headers.. i've installed my wbo2 sensor there as well (after the 3 join into 1 pipe)

cool,
11.6:1 is too rich, N/A cars get away with ~-13:1
mine is at 12-12.5:1 @ the high end
btw no biggies about the headers.. i've installed my wbo2 sensor there as well (after the 3 join into 1 pipe)

I know...its is really high...I need to get it to 12.9 to 13...that is the best power output for N/A engine though (text book wise) but my mechanci says that some cars get beeter output with different AFR.

I just popped open 3 S50B30 DME's, as described to check on the chips inside them.

All 3 vanos controllers had hand written dates and VNC written on them.....

One of the DME's had a 28 pin chip with no sticker on it, no typed numbers on it... just a clear window.

Does anyone know from opening a "virgin" DME if they ever came that way....or, is this likely an aftermarket chip of some kind?

The third DME has what I hope is a tuned chip, but its a 28 pin chip in a 32 pin socket. I'm told that as long as the extra pins (all 4) are on the non-flat side, all is good, and it should run.

These are a PIA to open, you have to really, really stretch that back ribbon, and release the side clips in the plug end to accomodate that angle... then, pull it back. Not easy or fun.

Alex.

Just started with the tuning process today...I haven't dynoed the car since I replaced the knock sensors (ping sensors) and the O2 sensors. It turns out that the car runs a lot better with these replaced. The car had 220 WHP on a Dynopack before the sensors...now, the first run shows 235 WHP (something more expected with this engine) and a better A/F ratio. The Dynopack estimated the flywheel to be at around 282.1 HP, and that with a 102 degrees f ambient temp. I am quiate happy, as my car is all stock, and is a 7/94 build...so I think I have a reasonably good condition engine.

Too bad, the software that was sent to my tuner is not correct, and the ignition timing can't be changes...we e-mailed requesting for the correct program, and hopefully be able to tune ignition timing tomorrow. Worked on proper enrichment today, and lost a bit of HP running the car a little more rich to be on the safe side. I am hoping I would get more power imrpovements tomorrow when we do the timing!

I'm not surprised it pings in the 1-2k RPM range - BMW is typically most aggressive in this range compared to other RPM ranges. They don't even add much timing in from this low RPM to the 2-3k RPM cruising range, whereas I've seen most engines like at least a solid 4-8 degrees more in this range vs. down by idle.

Running rich up top isn't that bad for a high compression NA engine. I doubt you'll find much power by going from mid-high 11's to mid-high 12's AFR, but the combustion chamber temps will climb. I haven't tuned one though, so I haven't seen firsthand, just guessing on what I've seen with AFR changes relating to power.

Got my tuning results back from the shop...it is well worth the money! Will post up pictures...but before the tuning, we were able to get a best base run of 238 to the wheels on a Dynapack....(not dynopack like i said before). running it on 3rd gear in an axle dyno. We calibrated the gear ratio by setting the car to run on 3000 RPMS...which resulted in a 5.45 overall gear ratio.

After the tuning, note that this car is completely stock...stock exhaust, stock filter, stock everything! We were able to get 255.6 to the wheels, 17 WHP at the peak...and what is impressive is that the car had gains all throughout the rev range.

The AFR was actually adjusted to be a little more lean and the timing on the higher RPM advanced a bit to sqeeze out the power on this engine. The AFR was se tto an average of 13.37:1 or .91 lambda.

I have to say I am quite happy with the car, it now has an estimated 306.7 HP on the flywheel...my very first car with over 300 HP...of course, I know it could just be the dyno and it is not really to the wheels...but I am still happy!

I certinaly hope my combustion chamber temp will not go up as that much....does anyone have a rough estimate on how much AFR will climb relative to temp change?

I know that as temp goes down and the oxygen is more dense, the AFR will be more lean. Is 13.3:1 too aggressive for daily driving and track use? I think this AFR was taken with temp of about 100F....I live in a tropical country.

The first image is the chart showing the estimate HP to the flywheels...i guess it is just a computer estimate...but the Green is the best base run we made, and the red is after the tuning. Note that there were no mods to the engine and the engine is completely stock with stock exhaust and stock air intake box.http://i65.photobucket.com/albums/h229/gibsonyu/Scan3.jpg

The second is the actualy Wheel HP...same graph, the computer just basically changed he number on the left side...255.6 to the wheels isn't bad for a stock S50B30 that was built on 1994 :redspothttp://i65.photobucket.com/albums/h229/gibsonyu/Scan2.jpg

The final image is the AFR expressed in lambda...1 lambda = 14.7:1 AFR...basically, if you multiply the lambda to 14.7, then you get the AFR. The green is the stock AFR that was taken on the base run, this is after I replaced the O2 sensors and the ping sensors.http://i65.photobucket.com/albums/h229/gibsonyu/Scan1.jpg

You have to love the torque curve of these BMWs....90% of the torque is available from 3,500 to almost redline!

Is it good if a AFR goes up and goes down and once again goes up and down...?

When I did a dyno of my M3 the curves of torque and power were much smoother. If you want to see it, I’ll post.

BTW, nice amount of WHP.
Have a nice time!

Actually, that is what I asked also, but if you look at the units on the left side, the variance is only.01 lambda, which is basically 1% of 14.7..its really not that big. Most dynos, the units on the left side would be running between 12 and 14...which would seem smoother because you would need to move 13.7% to touch the top and bottom of the graph.