N54 swap?

[amcink]

Holy Sh*t!!!!!!!!!

Photos of the engine bay and any modification done.

[SLEEPYDUB]

Bump this up. I have a complete 335i i may be putting into my e30 and would like some insight

[nando]

if you have a complete car then it should be relatively easy - sorta. essentially you need all the modules from the 335i that interface with the engine - the DME, CAS, Cluster, etc. Then wire it up as if it were a 335i. you also need the key to start it of course.

Basically you're putting the entire 335i electronics into an E30 body. Not impossible, but quite a lot of work.

if there were an EWS delete for MSD80 then you would only need the DME, EKP and maybe the cooling fan modules, but as far as I know nobody has a workaround for EWS4. I have a potential workaround for EWS3 (MSV70/MSS70) but that doesn't help the N54 guys at all. I don't know if any MSD80 modules used EWS3 (don't think so).

also in the project forum there's a guy doing an N54 swap into a 323i.

[berlow94]

A $2.5k investment in a Vipec or Emtron would control all of these things that you guys are talking about. Electronic wastegates, dual Vanos, valvetronic, e-throttle, etc. Both of those standalones will also integrate with just about all of the factory sensors as well.
Will be a much cheaper and more reliable solution than trying to make a factory DME happy with a different guage cluster and all of that extra work.

Also, keep in mind, a good majority of a factory DME's job is to keep the car emissions compliant. NOT to optimize performance or gas mileage. This isn't the case with a standlone setup.

*Also, keep in mind that god forbid your current engine swap doesn't work out for you or you simply want a change of scenery. Those ecu's will work with almost any motor you can possibly fit into that engine bay. It's a one time investment that will serve you for many years to come. The only cost will be having an experienced tuner calibrate everything.

[nando]

I disagree - have fun tuning it. it's not like N54 tuning with the stock DME is an unknown element. if you go standalone, you're basically completely on your own. Most N54 guys pay a shop to install a $300 JB4 and call it a day. you'd be giving up the best part of the N54, which is the aftermarket support.

I can almost guarantee the stock DME is a lot more capable than any affordable standalone. It does a lot more than you think it does - sure, emissions, blah blah - but that's the same tuning you'd be doing with a standalone computer, only you're not starting from scratch. It's not like the stock computer says "oh, you're making too much power - EMISSIONS!" and cuts the throttle or something. :p

Having to choose between fuel economy/emissions and power is false anyway. that's what a modern computer does - the whole point of having 3D maps and load controls. it's the same reason why map switching is stupid unless you're changing fuels all the time (oh, the stock DME does that too, BTW). The difference between a "performance" map and a "fuel economy" map is your foot.

also, N54s don't have valvetronic, but N55s do.

[the imitator]

Nando, lol

I'm in the field and will comment later but as I mentioned, you initially disagreed, you DO need the cluster as stated in MY thread.

There is a standalone option specifically for the n54 power plant, yes it costs a few grand, but it's a better ecu than the stock dme, and is capable of running the higher power direct injectors. If you call aem and other various standalones they even state they don't have a solution, and the closest option they have are meant for disel applications.

Atleast the way I'm skinning the pig it's a full wiring and module swap, with a couple modules being optional such as the Frm.

[nando]

yeah it turns out the CAS needs the a VSS input and some other info from the cluster or it freaks out. you could use a CAN emulator to trick it into thinking it was getting those CAN messages, or you could hack the firmware and remove that (which would be tough).

for MSV70 I might be able to do without the CAs altogether so I don't have to deal with the associated issues.

[berlow94]

I'm going to have to respectfully disagree Nando.

I am no tuner myself. But am good friends with Nick Pavloski from Pavlotech. He did the full calibration on my blueprinted S52 car (from scratch). He does all of Ratchet Head's race cars, and does contract work for many big name shops such as Turner and Vac Motorsport. Aside from loading things such as sensor calibrations and injector dead times, what else is needed to run a motor? Firing order, ign. timing, and fueling. In my experience, the most crucial thing being ign. timing, which can be figured out pretty easily using the stroke vs. rod length equation. I also strongly disagree with the hardware in a stock dme being better than a $2.5k standalone. Although it is no $8k Bosch motorsport system, it is by no means "entry level".

I'm happy to put you in touch with him if you'd like a more in depth explanation.

[DesertBMW]

you DO need the cluster as stated in MY thread.
.

True. Cluster has to be detected by the DME if not then DME will be in 'safe mode'. That is the case with my N55. There is a list of N55 trouble codes and a result, the list tells you what code will make the DME reduce power. I am sure same is similar for N54.

Aside from loading things such as sensor calibrations and injector dead times, what else is needed to run a motor?

N54 is very sophisticated engine. Getting aftermarket standalone ECU to run it properly would be frustrating mission.

Besides N54 has sick torque, what makes you think that your E30 would need even more power than what N54 is making? You don't even have enough room to mount some wide sticky tires back there anyway. Vladimir Palariev won Ukrainian drifting championship with E30 with bone stock N54 using bone stock MSd80.

[nando]

you can delete the trouble codes, the issue is the CAS not the DME. you only need the CAS because there's no software EWS4 delete like there was for MSS5x and similar computers.

berlow94, I think you missed my point. of course it can be tuned - I didn't say it was impossible. But the expertise of somebody like your friend is not free, and neither is the dyno time. Yes, tuning WOT for maximum power is pretty simple. But that's not what I'm talking about. It's the day to day drivability that is difficult and laborious to tune for.

Things like cold starts, hot starts. Hot starts on cold days. Cold starts on hot days. Humidity, altitude. Fuel economy (you can say it doesn't matter all you want, but it does), throttle tip in, responsiveness, smoothness of power delivery. It's hard enough to get all of that right on a 25 year old single cam n/a motor.

Now you're doing it on a highly advanced, dual vanos turbo engine (and valvetronic, if you're counting the N55 - and no, it's not just like controlling an E-throttle). Yeah you can slap an aftermarket computer, tune it on a dyno for a couple hours, and it will be fast - but you will never have the drivability and reliability of the stock computer unless you spent a lot of time (or money) tweaking it endlessly.

let me give you an example of something the stock N55 DME does that I seriously doubt a $2500 standalone can do. For one, the MAF isn't really needed on a valvetronic engine - it's mainly for emissions and a sanity check. the real load and air mass calculations are done by controlling the amount of valve lift. The DME actually knows exactly how much air will enter the engine at any given RPM and valve lift before it even measures it with the MAF. And the N55 valvetronic is even more advanced than the N52 - not only can it control the valve lift, it can actually control each valve independently, per cycle (not to the full travel of the valve, but enough to have an effect). That's just scratching the surface.

If you think a $2500 computer is going to be better than the one a multi billion dollar company designed with years of research and millions in R&D, then I guess this discussion is over. if you want to spend $4,000+ (unless you're not including installation tuning, and dyno time, again, not free) on a standalone that will be fast at WOT and somewhat tractable on a daily basis - fine by me! But I've been there, the endless tweaking trying to get every single possible condition right gets old after a while.

The one thing it might gain you is a somewhat simplified installation. You'll still have to wire it in though and it will likely be a custom harness. The wiring alone isn't going to be cheap, especially if you're paying somebody else to do it for you.

There was a time that people thought the S54 computer was too complicated and they were all going with standalones. I helped put together one once, and actually tried to convince them that the stock computer would be a better idea. Especially after working on that DME and harness, I was more convinced than ever. They always had problems with it too and I don't know if they ever got it running correctly. Had they used a stock computer it would have fired up and driven perfectly on day 1. They could have spent an hour re-tuning ignition and vanos, and left the rest alone because all the cold start routines etc. would already be perfected.

[the imitator]

picture of the unit made by SYVECS
You will have direct and port injection control likely for a total of twelve injectors. Those worrying about the DCT will finally have a real solution for full control of the transmission.

They expect the N54 unit will be a variant of the S6plus Syvecs ECU meaning the following should be supported:

Four More Outputs meaning running 12 Injectors becomes easier
- Six more Half Bridge Outputs
- Can use OEM Denso Lambdas or Bosch LSU4.9 sensors
- Logs Clutch A Pressure, Clutch B Pressure, Gearbox Main Pressure, 4WD %, Clutch Slip A and Clutch Slip B from the TCM
- Main Relay control on board
- Package (Canbus) Firmware can be updated in the field
- Lower cost than S8 Kit

It is still early and the ECU is still under development so we will see what the final product actually is as well as the final cost.

The Motec M182 works on the N55 simply because that engine uses a solenoid type injector, not piezoelectric like the N54.
Cant remember who, but someone machined an N54 head to accept N55 injectors.


All this is stupid talk regarding standalones its would make zero sense in almost every way to use such an item on an N5X in general;
Unless that of course you have a dedicated track car, and someone who knows how to use the extra features that a standalone offers, the practicality just isn't there.

[npavlo]

Hi guys,

Nick here from Pavlotech.

Thanks Andrew for the shout out.

First.
I wrote this quickly, and mostly to make sure people know what the aftermarket is capable of now. Maybe it needed to be a new thread....

Emtron cannot currently drive GDI injectors. It can run any port injected engine though...... I even did a simulation on the BMW Mclaren F1 engine, and one KV series ECUs can run it. I feel like that is the highest I/O count of any engine ever....

Anyways, there are some ECU that can drive GDI, but most of them are just "opening" the injector. When talking about Bosch solenoids (N55, VW, N20), there is a specific activation strategy. There are opening phases, hold phases, closing phases, polarity swaps, etc. There are multiple methods of opening the injector, and I haven't even gotten into multiple injections per stroke, and fuel pressure controls.

Instead of releasing support for GDI by just being able to open the injector (like some), Emtron wants to be able to have the user define all of these functions completely, similar to the way the ECU allows you to totally define current peak and hold injectors.

Anyways... Regarding aftermarket supporting other BMW engines, I have been involved in these headaches for over 10 years. I can say that FINALLY, there are a few ECUs that can control things properly.

Motorsport Vanos control (S50e, S54, S62) has a very specific control strategy.
Individual electric throttles have a very specific control strategy.

Most ECU say they support VVT and Ethrottle, but cannot be configured to drive these functions on those engines properly. For example(s),

Motorsport Vanos does not require constant regulation to "hold" a position. This is why it is output intensive (two outputs per cam). Some ECU control the solenoids using a "slave" strategy similar to controlling a twin winding idle speed valve. This over-regulates and can cause solenoid failure. Limitations in closed loop PID adjustments are usually the reason for this.

Individual electric throttles require heavy duty drivers to overcome multiple return springs. Most ECUs that support E-throttle do not have hardware large enough to support this.
The geometry of the linkage is NOT linear, since it was NEVER designed to be used with a cable throttle. This means the ECU (whether factory of aftermarket) needs to have a very specific compensation to the closed loop control depending on the target position. Most ECUs have limitations with PID as with the Vanos I mentioned above.
It also relaxes closed, which is unconventional. It also never opens ALL the way. In fact, the throttles are closed sometimes at WOT to change air velocity to provide more efficient cylinder filling.

I can say with confidence, the Emtron can handle ALL of those above complications, PLUS improve it. Another cool thing is I have the base calibrations set up so accurately, I can target cam position much better than the factory ECU, as well as the throttle kits. Not only that, but everything is adjustable. The calibrations are complete enough to provide start up, cold drivability, and 95% of hardware config normally. Not to mention, enable a slew of motorsport functions like launch control, motorsport traction, high speed logging, improved knock control, flex fuel support, anti lag, boost control, etc.

anyways...

Funny thing is when new generation engines came out, control of throttle and VANOS became more conventional. It is EASIER to run now. The only thing that's difficult on N54 is the injectors.

Valvetronic and the aftermarket? Don't expect support for that anytime soon (at least from Emtron). Besides the fact the motor will draw 40+ amps on the N52, and the N55+ are BLDC motors (both require special hardware, wiring, special connectors), there is no performance component to that technology, which is WHY the aftermarket parts exist in the first place - for performance. The system is designed to reduce pumping losses and allow a third dimension regarding valve timing control for emissions (valve phasing at medium lifts).

Regarding factory ECU, on N54 like you guys have mentioned, the ECU is heavily dependent on CAN integration. JBE gateway for diagnostics, CAS immobilizer, DSC info, EKP fuel pump control, etc. I assume eventually people will find ways to hack these dependancies out (or they have) like the MSS5X ECUs, but in the end you are LIMITED in what you can do with the factory ECU.

I disagree with the comments about "how flexible" it can be. A great example is how it is common to remove the MAF sensors on some of these engines and call that an "advantage". How could having the ECU have NO point of load reference be an advantage? Carburetors do a better job of determining load at that point. I drove a supercharged car that had belt slip and let me tell you how great that car drove. When people call that "Alpha N", that's far from the truth as well. There is a lot going on in the back end of the ECU to determine efficiency based on active substitute values, an intricate fuel model, referenced by yes.... the throttle position. Which on those engines the throttle doesn't even MOVE until a certain load. Weird how that works if it was true Alpha N.

You normally can't change fundamental functions like re-assign i/o, change the firing order, change PID, RE-WRITE how the ECU calculates efficiency, etc. If the system is based on air mass, and it was a normally aspirated car, how can it compensate for air density when the air all of a sudden becomes compressed? You can manipulate look up tables to obviously make it work, but the point I am trying to make is having proper control of the EFI system on modern engines is very important.

Back to the aftermarket, you can actually define your efficiency strategy. You can measure air mass directly. Use speed density. Compensate for fuel pressure differential. Enable wide range closed loop. Use back pressure as a source of engine load. Enable engine trips (your own limp modes based on engine temps, fuel pressure, mixture, etc), The list goes on and on. A properly functioning, GOOD, aftermarket ECU system is MUCH more capable of dialing in an engine calibration. If this wasn't true, then you would see more pro racers using factory ECUs.

Don't mis construe me.. It may seem like I am bashing the factory ECU. I am not trying to do that. I think it is really cool actually what can be done now with some of these ECUs that seemed to be totally locked down years ago. I just know how much better the aftermarket can work if setup PROPERLY. If your project is within the realm of capability of using the factory computer, go ahead. Especially if budget is a concern... but there are lots of guys who spend 40 grand on hardware, and expect to tune the car for less than $500. That just doesn't make sense.

If your project requires more accuracy (because you care about your hardware), more flexibility, or you just plain have a control complex (not a bad thing regarding EFI and performance), then there are options. They do cost money. So do pistons, wheels, paint, etc. People need to learn more about investing in proper electronics.

I don't post on forums much due to time (and somehow had a few min on NYE haha), but I will pop on when I can to see what happened..... :)

[nando]

thanks for the post! I actually do agree with pretty much everything you said. The thing is, you are obviously a very knowledgeable and competent tuner, and are using good hardware. There's so much crap out there with little to no support, and tuners who can't do more than a couple WOT pulls, that people often get stuck with a lesser solution than the factory DME could provide at a lower cost. Obviously, that's not always the case.

There's always tradeoffs of course - the one really nice thing about an aftermarket computer is the ability to change the tune in real time, and the datalogging is usually far better. with the stock computers we are limited by slow data rates and it takes several minutes to flash any changes because you have to write the whole tune instead of changing an individual table.

I assume you've opened up a valvetronic DME - the N52 has a HUGE capacitor inside which I presume is to soak up some of the valvetronic demands. 40 amps is a hell of a lot of current! It's too bad there isn't aftermarket support, but I think that will change eventually as more BMW engines change to valvetronic, including M cars.

[npavlo]

* edit
I didn't quote this properly... sorry. Read between your quote for my responses

Hey thanks. I try not to sound too biased. I hope mostly to provide information about the subject. I can't tell you how many conversations I have DAILY regarding factory ECU vs the aftermarket solutions I provide. It is all kind of a massive joke to me considering what I do.

thanks for the post! I actually do agree with pretty much everything you said. The thing is, you are obviously a very knowledgeable and competent tuner, and are using good hardware. There's so much crap out there with little to no support, and tuners who can't do more than a couple WOT pulls, that people often get stuck with a lesser solution than the factory DME could provide at a lower cost. Obviously, that's not always the case.

Right... I have dealt with the sub par hardware and support issue for 10 years. That's why I am managing the Emtron roll out here in the USA. We are very proud of how well we support this product so far. You can check out the website to read about some of the features/functions and what comes with some of the packages I put together for the BMWs.

There's always tradeoffs of course - the one really nice thing about an aftermarket computer is the ability to change the tune in real time, and the datalogging is usually far better. with the stock computers we are limited by slow data rates and it takes several minutes to flash any changes because you have to write the whole tune instead of changing an individual table.

Yup. Live tuning. Plus... say for the M3. Wide range lamdba, individual knock, on board logging (black box style), map switching, automatic map switching, table blending, motorsport traction, CAN integration so all the gauges/ABS/etc work, and more. :)

I assume you've opened up a valvetronic DME - the N52 has a HUGE capacitor inside which I presume is to soak up some of the valvetronic demands. 40 amps is a hell of a lot of current! It's too bad there isn't aftermarket support, but I think that will change eventually as more BMW engines change to valvetronic, including M cars.

Yes.... I have :). I know a lot about this, because like you said, I looked into supporting it. The factory header splits the input into two pins so the pins don't melt, and then the H-Bridge output is split into two pairs (4 pins) and then tied together in the engine harness. I can make something to run the motor, but there is a lot of other issues. One big one is the position sensors are super high resolution binary sensors. They are not analog inputs anymore, and they have redundancy.

Now... every BMW engine has Valvetronic too. They all use the fancy position sensor, but now the motor is a brushless DC motor (since N55). This is more efficient current consumption wise, but the control strategy is more complicated. Currently the aftermarket is just catching up on BLDC motor controls, and this is in the form of running fuel pumps mostly. Valvetronic can be done... the question is will developing the function be worth it? Like I said, anytime you are throttling the engine, the valve lift is at max. It is used for pumping losses regarding fuel economy, and emissions regarding the valve phasing concept at medium lifts. For N52 for example, I just crank them to max lift, and control engine load with the throttle (which it does anyways to a certain extent). It is a tough subject because my first thought is to run EVERYTHING, and then allow the user to switch off what they don't want to use (this is what I did with the E46M3, E39M5), but lately the truth is... the aftermarket ECU is intended for motorsport performance. So, that must be considered.

I have a ton of scope traces, diagrams, and other info. I didn't mean to hijack the thread (but I did... sorry), but feel free to ask any other questions. When I have a second I will reply.

[the imitator]

Hi guys,

Nick here from Pavlotech.

Thanks Andrew for the shout out.

First.
I wrote this quickly, and mostly to make sure people know what the aftermarket is capable of now. Maybe it needed to be a new thread....

Emtron cannot currently drive GDI injectors. It can run any port injected engine though...... I even did a simulation on the BMW Mclaren F1 engine, and one KV series ECUs can run it. I feel like that is the highest I/O count of any engine ever....

Anyways, there are some ECU that can drive GDI, but most of them are just "opening" the injector. When talking about Bosch solenoids (N55, VW, N20), there is a specific activation strategy. There are opening phases, hold phases, closing phases, polarity swaps, etc. There are multiple methods of opening the injector, and I haven't even gotten into multiple injections per stroke, and fuel pressure controls.

Instead of releasing support for GDI by just being able to open the injector (like some), Emtron wants to be able to have the user define all of these functions completely, similar to the way the ECU allows you to totally define current peak and hold injectors.

Anyways... Regarding aftermarket supporting other BMW engines, I have been involved in these headaches for over 10 years. I can say that FINALLY, there are a few ECUs that can control things properly.

Motorsport Vanos control (S50e, S54, S62) has a very specific control strategy.
Individual electric throttles have a very specific control strategy.

Most ECU say they support VVT and Ethrottle, but cannot be configured to drive these functions on those engines properly. For example(s),

Motorsport Vanos does not require constant regulation to "hold" a position. This is why it is output intensive (two outputs per cam). Some ECU control the solenoids using a "slave" strategy similar to controlling a twin winding idle speed valve. This over-regulates and can cause solenoid failure. Limitations in closed loop PID adjustments are usually the reason for this.

Individual electric throttles require heavy duty drivers to overcome multiple return springs. Most ECUs that support E-throttle do not have hardware large enough to support this.
The geometry of the linkage is NOT linear, since it was NEVER designed to be used with a cable throttle. This means the ECU (whether factory of aftermarket) needs to have a very specific compensation to the closed loop control depending on the target position. Most ECUs have limitations with PID as with the Vanos I mentioned above.
It also relaxes closed, which is unconventional. It also never opens ALL the way. In fact, the throttles are closed sometimes at WOT to change air velocity to provide more efficient cylinder filling.

I can say with confidence, the Emtron can handle ALL of those above complications, PLUS improve it. Another cool thing is I have the base calibrations set up so accurately, I can target cam position much better than the factory ECU, as well as the throttle kits. Not only that, but everything is adjustable. The calibrations are complete enough to provide start up, cold drivability, and 95% of hardware config normally. Not to mention, enable a slew of motorsport functions like launch control, motorsport traction, high speed logging, improved knock control, flex fuel support, anti lag, boost control, etc.

anyways...

Funny thing is when new generation engines came out, control of throttle and VANOS became more conventional. It is EASIER to run now. The only thing that's difficult on N54 is the injectors.

Valvetronic and the aftermarket? Don't expect support for that anytime soon (at least from Emtron). Besides the fact the motor will draw 40+ amps on the N52, and the N55+ are BLDC motors (both require special hardware, wiring, special connectors), there is no performance component to that technology, which is WHY the aftermarket parts exist in the first place - for performance. The system is designed to reduce pumping losses and allow a third dimension regarding valve timing control for emissions (valve phasing at medium lifts).

Regarding factory ECU, on N54 like you guys have mentioned, the ECU is heavily dependent on CAN integration. JBE gateway for diagnostics, CAS immobilizer, DSC info, EKP fuel pump control, etc. I assume eventually people will find ways to hack these dependancies out (or they have) like the MSS5X ECUs, but in the end you are LIMITED in what you can do with the factory ECU.

I disagree with the comments about "how flexible" it can be. A great example is how it is common to remove the MAF sensors on some of these engines and call that an "advantage". How could having the ECU have NO point of load reference be an advantage? Carburetors do a better job of determining load at that point. I drove a supercharged car that had belt slip and let me tell you how great that car drove. When people call that "Alpha N", that's far from the truth as well. There is a lot going on in the back end of the ECU to determine efficiency based on active substitute values, an intricate fuel model, referenced by yes.... the throttle position. Which on those engines the throttle doesn't even MOVE until a certain load. Weird how that works if it was true Alpha N.

You normally can't change fundamental functions like re-assign i/o, change the firing order, change PID, RE-WRITE how the ECU calculates efficiency, etc. If the system is based on air mass, and it was a normally aspirated car, how can it compensate for air density when the air all of a sudden becomes compressed? You can manipulate look up tables to obviously make it work, but the point I am trying to make is having proper control of the EFI system on modern engines is very important.

Back to the aftermarket, you can actually define your efficiency strategy. You can measure air mass directly. Use speed density. Compensate for fuel pressure differential. Enable wide range closed loop. Use back pressure as a source of engine load. Enable engine trips (your own limp modes based on engine temps, fuel pressure, mixture, etc), The list goes on and on. A properly functioning, GOOD, aftermarket ECU system is MUCH more capable of dialing in an engine calibration. If this wasn't true, then you would see more pro racers using factory ECUs.

Don't mis construe me.. It may seem like I am bashing the factory ECU. I am not trying to do that. I think it is really cool actually what can be done now with some of these ECUs that seemed to be totally locked down years ago. I just know how much better the aftermarket can work if setup PROPERLY. If your project is within the realm of capability of using the factory computer, go ahead. Especially if budget is a concern... but there are lots of guys who spend 40 grand on hardware, and expect to tune the car for less than $500. That just doesn't make sense.

If your project requires more accuracy (because you care about your hardware), more flexibility, or you just plain have a control complex (not a bad thing regarding EFI and performance), then there are options. They do cost money. So do pistons, wheels, paint, etc. People need to learn more about investing in proper electronics.

I don't post on forums much due to time (and somehow had a few min on NYE haha), but I will pop on when I can to see what happened..... :)

Great information, and thank you for validating, and debunking parts our conversation. This is the type of information that progresses forums, new swaps, new information, new products, good domino effect.