E46 330i Automatic - M54B30 Turbo

[Butters Stoch]

It worked.

However, how much I had to open it up is a bit shocking.

I wouldn't call the mounting plate and rubber boot "bolt in". (I've cut off the boot...but I'am considering zip-tying the top of it back on to partially cover the hole in the transmission tunnel)

This is exactly what I meant a few months ago about being
Unsure about the shifter. For the amount of money it is, I think there is a bit of improvements it needs. Yourself and I are very competent. The install isnt the issue. Product needs improving. But, it surely shifts like a bastard.

[PEI330Ci]

This is exactly what I meant a few months ago about being
Unsure about the shifter. For the amount of money it is, I think there is a bit of improvements it needs. Yourself and I are very competent. The install isnt the issue. Product needs improving. But, it surely shifts like a bastard.

It works now...not happy about the drama to get it done though. I thought I might have had a bad trans, and was was going to have to pull it out....

The next issue that's absolutely kicking my butt is a fuel leak somewhere with the surge tank. I'll end up with fuel all over the compartment it sits in, but when I pull off the cover and run both fuel pumps (Lift and main) I can't seem to find where the issue is. Sometimes nothing leaks.....other times it just pours out. I'm so pissed off about this situation that I'm considering ditching the surge tank, removing the A1000 pump, and having the OEM fuel pump feed an A750 I have sitting on the shelf and run them in series. I'd leave all the existing lines in place and just plug everything together with back-to-back fittings.

[PEI330Ci]

And as far as my wheel-hop issue goes, I'm a dumb ass.

At some point I did have the MCS 2 way (remote reservoir) dampers set half way through their adjustment range, but when I just checked all the dampers, everything was set to zero.

So I've had minimum compression and rebound damping on both the front and rear. The handling can only get better from here....jeepers!!!!

[Matutino]

On a completely separate note, the Steed Speed Twinscroll is holding 10 PSI of boost on a 10 psi WG spring (Using an MVR). Spool isn't as quick as I was hoping, but I think this is probably a function of how I have the VANOS tuned and very conservative ignition timing. Once I have some time to sort through all the data, I'll post some graphs here.

With a gt30 you should be spoiling even at idle!
Seriously, what are you seeing spoolwise

[PEI330Ci]

With a gt30 you should be spoiling even at idle!
Seriously, what are you seeing spoolwise

I can hear it spooling pretty quick, but the power just isn't there until 4k RPM.

As previously mentioned, I have just massively upgraded my data logging ability....but I also have a firmware upgrade that I'm working on where I have to move the entire tune over to a new software package that I've been working on. I'm reluctant to spend too much time fiddling with the tune on the old firmware.

Just the same, I might lock the cams in 10 degree increments and do some pulls to see how manifold pressure (and engine load) is affected.

The graphing abilities of Motec i2 Pro is staggering....just looking at how to easily export screenshots.

[PEI330Ci]

Shifter clearancing:

[PEI330Ci]

Downpipe tacked up:




Finish welded with WG recirculation:






WG pipe needed some extra space; ground down some webbing on the transmission:




Exhaust installed:






Sensors and Meth:




Engine all buttoned up with both Meth and Nitrous plumbed:

[PEI330Ci]

I'm leaning really heavily towards removing the surge tank and A1000 pump. The amount of fuel that I've "spilled" with this setup is alarming, and still don't have an automated lift pump strategy that I trust. (I've been manually filling the surge tank)

I've got a pump on the driver's side of the OEM fuel tank to move fuel to the passenger's side pump, so that should eliminate the fuel starvation issue for the most part at lower fuel levels. I would use the OEM fuel pump (always on) to feed an A750 fuel pump that would be run with PWM control based on engine load.(Load based speed control)

[jakermac]

Why does your lift pump need any sort of on/off strategy? All the systems I've seen and worked on were simply on all the time, and the overflow off the top of the surge tank returned to the main fuel tank.

[PEI330Ci]

Why does your lift pump need any sort of on/off strategy? All the systems I've seen and worked on were simply on all the time, and the overflow off the top of the surge tank returned to the main fuel tank.

That's exactly how it's plumbed.

Initially I was seeing the surge tank being super pressurized, and I was afraid it was going to break open. (It's aluminum sheet metal) I added internal bracing to the tank, and a level sensor to create a fill strategy based on the fuel level. The tank still had a lot of pressure build up, which could be seen at the FPR with the main pump off. I then re-plumbed the return line to the OEM tank using a bulkhead fitting instead of the plastic nipple and the "suction jet" system. This released the pressure build up completely with the lift pump being on, but when the tank was 100% full I had fuel leaking out around the seal for the ATL level sensor. I pulled the surge tank again, removed the level sensor, and welded a thick piece of material over the hole. Everything should have been mint at this point, but with both pumps on, there was still fuel pooling inside the surge tank compartment.

Yesterday I had the surge tank compartment with probably 1/2L of fuel spilled in it. I pulled the cover off, dried everything up, then ran both pumps for 5 minutes with no sign of any leaks. Today I've been driving the car with only the main pump on, and filling the surge tank manually for 30 seconds every time I start the car. I've left the cover off the compartment, and haven't seen any leaks so far.....

I might pull the surge tank off this week, fill it was water, and then pressurize it with air to see WTF is going on. None of it makes sense at this point.....

[PEI330Ci]

What data traces are you guys interested in for the TS/GTX30 combo?

What I'm looking at right now:

Engine Speed
Intake Manifold pressure
Exhaust Manifold pressure
Vehicle Speed

[PEI330Ci]

[wazzu70]

Turbo speed
Intake air temp (T1)
Exhaust manifold temp (T3)

Also if you have for boost controller:
Target demand
Controller feedback

[PEI330Ci]

Turbo speed
Intake air temp (T1)
Exhaust manifold temp (T3)

Also if you have for boost controller:
Target demand
Controller feedback

Turbo Speed sensor is installed, but the signal hasn't been calibrated yet.

The Inlet Manifold Temp is calibrated, but the other 4 pre-engine air temp sensors, and 6 post engine (EGT) are not yet....so while the data looks interesting, I don't trust it yet.

Boost control is turned off on the Motec, I'm just running off the WG spring. (10 PSI) As you can see, it hits 10 PSI then creeps up to about 12 PSI above 6000 RPM. I've got a hard limit set at 6500 RPM currently until I make sure everything is running OK, then I'll change it to 7000 RPM. (Soft limit is about 200 RPM before)

Also, while I'm tempted to turn on the Boost control and run 15 PSI....I haven't set up all the safety features to keep the engine alive. V1.03 firmware, which I'm currently running, uses a static pressure target for fuel pressure. V1.04 firmware, which I've spent the last year transferring over all the features I wrote in M1 Build from 2014 onwards, uses fuel injector differential pressure as the safety measure input. (You have a somewhat constant value that is the sum of fuel pressure and manifold pressure....which means in my current state of tune, I have a 3.00 BAR line across the graph....unless I lose fuel pressure, and then it drops below 3.00 BAR) Then there's the knock control system, which currently I'm setting up the individual cylinder trims based on frequency 4 different frequencies, and I'm going to have to initiate knock to set the threshold. (Scary)

[wazzu70]

Sounds good Adam. Take your time and get it set up right. This is a fun and interesting build to watch!!

[PEI330Ci]

Went for a drive to get the vehicle inspected and registered.

Floored it in 5th gear @2400 RPM:

2450 RPM = 0 PSI
2650 RPM = 5 PSI
3000 RPM = 10 PSI

So it spools....I just don't feel like the car is making any power under 4000 RPM.

The knock sensors are showing lots of noise at full load....but nothing alarming. I'm going to work on analyzing the data a bit more, and then adjust the gains before trying to instigate knock at lower loads....

I got offered some high octane race fuel....but I think I'm going to work on sorting out the tune on pump fuel before turning anything up.

[NOTORIOUS VR]

The knock sensors are showing lots of noise at full load....but nothing alarming. .

What data are you looking at? volts or is your knock setup properly tuned/filtered?

If it's volts, that doesn't mean a thing.

[PEI330Ci]

What data are you looking at? volts or is your knock setup properly tuned/filtered?

If it's volts, that doesn't mean a thing.

It's in percent, and there is a bandpass filter applied to 4 channels for each cylinder.

Have a look at the online Webinars from Motec for more details. (There are 3 of them for M1 including practical demonstrations)

[Matutino]

Went for a drive to get the vehicle inspected and registered.

Floored it in 5th gear @2400 RPM:

2450 RPM = 0 PSI
2650 RPM = 5 PSI
3000 RPM = 10 PSI

So it spools....I just don't feel like the car is making any power under 4000 RPM.

The knock sensors are showing lots of noise at full load....but nothing alarming. I'm going to work on analyzing the data a bit more, and then adjust the gains before trying to instigate knock at lower loads....

I got offered some high octane race fuel....but I think I'm going to work on sorting out the tune on pump fuel before turning anything up.

What exhaust housing are you using?
On my gt3586r 1.06 divided I see 10psi at 3500 rpm

[NOTORIOUS VR]

It's in percent, and there is a bandpass filter applied to 4 channels for each cylinder.

Have a look at the online Webinars from Motec for more details. (There are 3 of them for M1 including practical demonstrations)

I understand the basics for knock sensor setup, that isn't much different from other ECU's, although the Motec does have a nice way of filtering and setting up individual cyl levels that some ECU's lack.

Have you done the cyl. level setup as well?

[Matutino]

Went for a drive to get the vehicle inspected and registered.

Floored it in 5th gear @2400 RPM:

2450 RPM = 0 PSI
2650 RPM = 5 PSI
3000 RPM = 10 PSI

So it spools....I just don't feel like the car is making any power under 4000 RPM.

The knock sensors are showing lots of noise at full load....but nothing alarming. I'm going to work on analyzing the data a bit more, and then adjust the gains before trying to instigate knock at lower loads....

I got offered some high octane race fuel....but I think I'm going to work on sorting out the tune on pump fuel before turning anything up.

I just saw you are using 1.06. Your turbo is spoiling less than much bigger turbos currently running this manifold but it completely makes sense when comparing against my turbo running the same exhaust housing.

[PEI330Ci]

I understand the basics for knock sensor setup, that isn't much different from other ECU's, although the Motec does have a nice way of filtering and setting up individual cyl levels that some ECU's lack.

Have you done the cyl. level setup as well?

The level setup is taking some time for me to understand.

At idle and under light load, all 6 cylinders across all 4 frequency bands are pretty tight. When you add load, there is a fair amount of randomness to the signal.

I've tried using a variety of histograms and scatter plots to get a sense of the signals, but I don't have a picture that I'm comfortable with. Looking at the averages, there is noticeable differences between the cylinders. Trimming the gains so that the averages match, you then have low level variance, and high level randomness. (The high values skew the data)

What I'm looking to do next is to export data into Excel, use only the values where manifold pressure is over 8 PSI, then generate an average value for all 4 frequencies per cylinder. This should give me the baseline to start inducing knock to set the trigger thresholds.

One thing I will say is that the engine isn't nearly as consistent as I would like it to be. Both the inlet and exhaust VANOS are NOT following the aim target, which is affecting load, and the fuel model isn't properly tuned yet so actual lambda doesn't follow target either. All of this is sure to affect when knock occurs, so while it will be useful to have a knock trigger set for big events, I won't be able to get too granular on the knock model until after I get engine airflow and fuelling sorted.

I just saw you are using 1.06. Your turbo is spoiling less than much bigger turbos currently running this manifold but it completely makes sense when comparing against my turbo running the same exhaust housing.

I was expecting more power down low, but I suspect part of my issue is in the tune. VANOS for example is moving around more than it should...which can't be helping....and I haven't tuned it at all.....

[NOTORIOUS VR]

One thing I will say is that the engine isn't nearly as consistent as I would like it to be. Both the inlet and exhaust VANOS are NOT following the aim target, which is affecting load, and the fuel model isn't properly tuned yet so actual lambda doesn't follow target either. All of this is sure to affect when knock occurs, so while it will be useful to have a knock trigger set for big events, I won't be able to get too granular on the knock model until after I get engine airflow and fuelling sorted.

I will be really honest with you, it sounds like you're trying to do a lot all at once.

If I may make a suggestion, leave the valve timing alone for now. Fully variable systems can only be tuned correctly on the dyno anyway. So create a basic map, leave the exhaust cam off, and have just a simple on off point for the intake cam for now mimicking the M5x single vanos for now. That will VASTLY reduce VE changes and issues and let you at least dial in your VE table better until you're ready to start calibrating VVT PID values and verifying that the duty cycle set points are correct to get your camshafts where you want to be.

I wouldn't worry too much about knock at this stage, while I don't remember what CR you're running and if you have aftermarket/forged pistons or not at 10 psi with a proper turbo CR chosen it will be difficult to make any large mistakes or damage to your motor even if your timing and fueling is slightly off.

Get your fueling sorted with a decent timing table plugged in and then worry about the rest.

[pbonsalb]

I'm leaning really heavily towards removing the surge tank and A1000 pump. The amount of fuel that I've "spilled" with this setup is alarming, and still don't have an automated lift pump strategy that I trust. (I've been manually filling the surge tank)

I've got a pump on the driver's side of the OEM fuel tank to move fuel to the passenger's side pump, so that should eliminate the fuel starvation issue for the most part at lower fuel levels. I would use the OEM fuel pump (always on) to feed an A750 fuel pump that would be run with PWM control based on engine load.(Load based speed control)

I'd like to see someone try the Holley Hydramat. Expensive, but seems like it should work for those having pickup issues at the pump sump.

[PEI330Ci]

I will be really honest with you, it sounds like you're trying to do a lot all at once.

If I may make a suggestion, leave the valve timing alone for now. Fully variable systems can only be tuned correctly on the dyno anyway. So create a basic map, leave the exhaust cam off, and have just a simple on off point for the intake cam for now mimicking the M5x single vanos for now. That will VASTLY reduce VE changes and issues and let you at least dial in your VE table better until you're ready to start calibrating VVT PID values and verifying that the duty cycle set points are correct to get your camshafts where you want to be.

I wouldn't worry too much about knock at this stage, while I don't remember what CR you're running and if you have aftermarket/forged pistons or not at 10 psi with a proper turbo CR chosen it will be difficult to make any large mistakes or damage to your motor even if your timing and fueling is slightly off.

Get your fueling sorted with a decent timing table plugged in and then worry about the rest.

I am trying to do a lot!

I locked the cam timing.

Engine is 8.5:1, 272/256 cams, highly ported head,

Fuel is very close to target under load, it's just the transitions and lift-off that needs work.

The timing map I'm currently using is from YOU!!! LOL

I'd like to see someone try the Holley Hydramat. Expensive, but seems like it should work for those having pickup issues at the pump sump.

I pulled the surge tank this morning, and plumbed in an Aeromotive A750 pump in series with the OEM fuel pump. No more fuel leaks, or smell. :P