NOTORIOUS VR's E36 328is turbo build

[GG///M3]

There is no pnp VEMS for the M/S5x, I'll be doing a straight wire in using the OE harness though. I've decided I will never go back to the Siemens ECU so I will slim down the OE harness and use it as is.

Very awesome sir.

[NOTORIOUS VR]

Thank you sir

Put a pressure sensor in the "collector". It will be very helpful to refrence that off of IM pressure. This is ideal to keep things safe when cranking up the boost. I'm sure I'm not telling you anything new here..

While I think balancing the pressure ratio between the intake and exhaust is important on this build I won't bother. I don't plan on pushing this setup too far at this time. Just to have some fun while I debate and hopefully build a 3L next year some time and which at that point I will be going with a top mount setup and a larger turbo.

[wazzu70]

How did you mount the EGT sensor. Did you tap the manifold or weld the bung?

[Colby Colbs]

Gotcha. I figured it your going to moniter EGT, backpressure is just as important to keep things happy. I also figured since your right there, it would be very easy.

[NOTORIOUS VR]

Gotcha. I figured it your going to moniter EGT, backpressure is just as important to keep things happy. I also figured since your right there, it would be very easy.

True, you are correct but at this point the EGT probe is more there for "just in case" situations so the VEMS can add fuel to cool things down a bit if needed.

With that said, it's been a while since my last update. I've been working hard between my day job and tuning cars to get this motor in the car and I finally did it last night.

So after the wiring was completed, the only thing left to do was to install the engine accessories, new exhaust studs and nuts, connect all the turbo lines, install the blanket, RMS, clutch, oil pan w/ bung and just generally button up the motor and get it ready for the engine bay.

good as place as any to start would be to make sure the oil pump nut doesn't decide to remove itself



then of course the turbo needs cooling so welding on some -6 bungs...



then make and run some lines to connect it all...



oil feed...


all secured at the back


dvcee A/C relocation bracket and engine mount:



all accessories fitted and belts installed...


took the motor off the engine stand to get at and replace the RMS before installing the oil pan...


UUC flywheel.. notice the XX's that is how you know where to place the 2nd plate to keep the clutch kits balancing BTW... FW bolts are TQ'ed to 105Nm as per OEM as that is apparently what the UUC twin disc kit calls for.


the clutch all together, of course a friend as to finger fvck it.. PP bolts TQ's to OE spec of 25 ft-lb (this I am assuming since there was no TQ spec's delivered with the UUC clutch kit, and since the FW bolts are supposed to be TQ's to OE spec I figured the PP ones are to as well.


Then on to removing the OBD2 fuel system monstrosity, as I will be running an AEM external FPR with -6 lines from the OE hardlines to the regulator, fuel rail and back again...


all gone, just installed a new OE fuel filter...


then a few minutes later and some wiggling the motor was back in car


lots of room, but as you can see the A/C line is too short... I'm hoping DVCEE will answer my PM's about this.


back side..


oil pan (need to get a 45 deg AN fitting still...


And that's all she wrote for now.. Today I'm hoping to clean up the wiring mess everywhere up top and figure out the IAC hose situation and possibly get the IM on. After that it's just a matter of modifying the OBD1 fuel fail, plugging in the harness, and a few other small thing and I might just be ready to throw the front end on to start fabbing the 3" IC piping

[NOTORIOUS VR]

so pushing forward... spent some time cleaning up the tranny, rebuilding the shifter linkage with new parts to get rid of the slop from 228k km on the car..

new parts!


shifter assembly taken apart and cleaned:


old worn out parts:



back together again with new parts!



and filled the tranny with some fresh Motul 300 fluid to finish it off..


time to get to work on prepping the intake manifold to be permanently mounted to the motor in the car.

First task was getting the IAC from the OBD2 mani to be mounted on the OBD1 mani... first attempt:



the above IAC position didn't work out, it was hitting the AN fitting for turbo cooling on the side of the head (originally the TB coolant port) so had to change it up by making a bracket to center the IAC under the manifold. My buddy Robbie from RS Autosports (the shop where I'm using the hoist and doing this build) quickly welded up something that worked out perfectly!



OBD2 TB on the OBD1 manifold:


Old O2 sensor harness reused for the MAP sensor connection:


Intake mani complete:



Intake installed and driver side wiring cleaned up and connected:


VEMS ECU in it's new home:


Plugged in the ECU and test fit the cover...


Wiring from the VEMS to the inside of the car for serial com port, and misc inputs for PID boost control changes and flat shift/2-step limiter, etc.


Through firewall into the ECU box behind the airbag:


10-pin connector for easy access and modifications:


Inside harness w/ COM port finished. The aux inputs will be hooked up at a later time...


What's next?

Tranny needs to go back in, finish wideband O2 wiring to the VEMS, connect EGT, and main power wires to distribution block in engine bay. Figure out how much -6 hose and what fitting I require to finish the fueling system and get the front end on to mount the cooler and fabricate the IC piping.

I can almost hear the car starting

[mainuh325is]

Don't take this the wrong way, but I think your shift linkage is upside down. The curve is there to clear the guibo

[NOTORIOUS VR]

Don't take this the wrong way, but I think your shift linkage is upside down. The curve is there to clear the guibo

That is entirely possible.. I took it off without marking a reference point, so I did a quick search on the net and didn't find much... so I just put it on for now and decided to deal with it when I get the tranny on.

[Cyrix2k]

Looking good! If you want to try something cool for your VEMS, look into using this bluetooth adapter.
http://www.ebay.com/itm/160799222877...84.m1497.l2649

You have to program it once on a computer - after that, you can leave it plugged into the ECU and interface with it as a virtual serial port. I use one on my MS and it's amazing. Not only does it make it simple to log, you can alter tables from across the garage.

[wazzu70]

I don't see many builds that use the external MAP sensor. Any reason why you chose that versus a board mounted one?

Just curious.

[RobertFontaine]

How does vems compare to megasquirt?

[Eloff19]

Nick:

I am also using external map...I took Dmitry's advice/logic that the shortest vac lines may (or may not) make a difference for response. It does, however, definitely reduce the length of tubing which is a big plus in my book (potential reduction in the chance of a leak at the very least).

RobertFontaine:

It's better in my book (though I have not used new(er) ms3) simply due to the proper automotive AMP/weatherpak connector setup. The software is stable and the various output opportunities are great. That said, the two gripes I have are that I think it needs more high powered outputs and the documentation support is poor.

Notorious:

Build looks excellent sir! You Canadians are certainly doin' some work reppin' the VEMS 'round here.

[5mall5nail5]

The gain it makes up in having short tubing can also be detrimental due to electronic noise on the longer leads.

I have my map sensor shielded using teflon and screened cable.

[chikinhed]

I recommend you don't use the plastic clutch fork pivot, Rallyroad makes a stainless steel one that will put up with the abuse of an after market clutch.

[NOTORIOUS VR]

Looking good! If you want to try something cool for your VEMS, look into using this bluetooth adapter.

I will do that eventually, I did but a BT serial adapter already but it's far, far down on my list :P

I don't see many builds that use the external MAP sensor. Any reason why you chose that versus a board mounted one?

Just curious.

Actually I normally always build/order my ECU's with an internal MAP, the only reason why I did an external this time around was because I already had the external MAP sensor which I was going to use in combination with my VEMS Round Gauge in my Audi. Since that didn't happen and the gauge will now be used to display AIM data from the ECU I no longer needed it and was able to use it for my VEMS in the E36.

Nick:

I am also using external map...I took Dmitry's advice/logic that the shortest vac lines may (or may not) make a difference for response. It does, however, definitely reduce the length of tubing which is a big plus in my book (potential reduction in the chance of a leak at the very least).

I've personally never seen any issue running a longer hose to the ECU w/ an internal MAP

RobertFontaine:

It's better in my book (though I have not used new(er) ms3) simply due to the proper automotive AMP/weatherpak connector setup. The software is stable and the various output opportunities are great. That said, the two gripes I have are that I think it needs more high powered outputs and the documentation support is poor.

I agree on the poor documentation, but we do have a pretty good network if knowledgeable people to help. As for the high power outputs, in most cases (4 & 6 cyl you will have enough) , you have 8 IGN, 8 INJ and 4 stepper (push/pull) outputs, and then various low power outs as well.

Notorious:

Build looks excellent sir! You Canadians are certainly doin' some work reppin' the VEMS 'round here.

I try.. I've been dealing with VEMS for many years now and I like their products, I guess it's about time one of my own projects started to use them.

The gain it makes up in having short tubing can also be detrimental due to electronic noise on the longer leads.

I have my map sensor shielded using teflon and screened cable.

If you have noise issues, I'd say there is a problem in your ignition system. Generally most sensors won't need any shielding when it comes to the wiring except for VR sensors.

It basically comes down to preference more then anything IMO.

I recommend you don't use the plastic clutch fork pivot, Rallyroad makes a stainless steel one that will put up with the abuse of an after market clutch.

Well the good thing is it's not hard to do a tranny on this car as the new plastic one will have to do for now

[5mall5nail5]

I don't personally have noise issues. But, for all high rate sensors I shield them.

[chikinhed]

I have a Rallyroad one that I'm not using, I could chuck it in the mail to you for a nominal fee if you want. I'm in Vancouver. Lemmieno.

[NOTORIOUS VR]

I have a Rallyroad one that I'm not using, I could chuck it in the mail to you for a nominal fee if you want. I'm in Vancouver. Lemmieno.

PM'd

[Eloff19]

I agree on the poor documentation, but we do have a pretty good network if knowledgeable people to help. As for the high power outputs, in most cases (4 & 6 cyl you will have enough) , you have 8 IGN, 8 INJ and 4 stepper (push/pull) outputs, and then various low power outs as well.

Most definitely on the knowledgeable people front...but navigating that wiki is not for the faint of heart and finding up-to-date stuff can be quite trivial. The only thing I really would like to see is some sort of scripting interface. There are all sorts of cool one-off stuff you could do with more open access to the ecu's datastream/input/outputs.

As far as enough inputs; sure they are there but I, like you, have a 6 cylinder engine...which on the 8cyl ecu leaves 4 high powered outputs...Wire in the IAC (2), vanos and boost controller and I'm left with one...which doesn't leave much room for additions. I know you can certainly use the lower power stuff to drive relays but that can and does become quite cumbersome. I guess I have an innate fear of burning up the stepper chips and would prefer more high power stuff so maybe that complaint is unfounded. Otherwise, I love it...and for the price point/features ratio it is hard to beat what the system offers.

Sorry if this it too OT; I look forward to your results.

[NOTORIOUS VR]

Most definitely on the knowledgeable people front...but navigating that wiki is not for the faint of heart and finding up-to-date stuff can be quite trivial. The only thing I really would like to see is some sort of scripting interface. There are all sorts of cool one-off stuff you could do with more open access to the ecu's datastream/input/outputs.

As far as enough inputs; sure they are there but I, like you, have a 6 cylinder engine...which on the 8cyl ecu leaves 4 high powered outputs...Wire in the IAC (2), vanos and boost controller and I'm left with one...which doesn't leave much room for additions. I know you can certainly use the lower power stuff to drive relays but that can and does become quite cumbersome. I guess I have an innate fear of burning up the stepper chips and would prefer more high power stuff so maybe that complaint is unfounded. Otherwise, I love it...and for the price point/features ratio it is hard to beat what the system offers.

Sorry if this it too OT; I look forward to your results.

Absolutely! Seems like you're well versed with the VEMS as well... good too see it. I'm trying to make more people aware of this ECU because as you say price/feature it's very hard to beat.

I actually just got http://vems.ca up and running (although still work in progress and website creation is not my strong suit lol) so hopefully that will help as well.

I agree on the outputs, they get used up quickly if you take advantage of most/all the features... I will bring up this point to some of the developers and see what they have to say

Not OT all by the way... I always enjoy a good discussion with level headed people

[wazzu70]

Thanks for the answer regarding the external MAP. I dont think there is really any noticable benefit/delay having one or the other. If you use the proper tube its pretty beefy and does not expand/contract under pressure or vacuum.

That said, because the tube is so beefy the bend radius is pretty large and its hard to route. I was thinking sometime I might try the external sensor due to the fact I can keep the engine bay cleaner. Much easier to route wires versus the tubing!


VEMS is an awesome unit per the price for sure. Documentation is the weak link for sure. I am wanting to help on that end though since I cant really contribute with code. I also think the effort/benefit would be very nice. Like most people though, I just need the time to sit down and hammer some stuff out.

This is also my first standalone, so Im pretty aware what things are not covered well for a n00bs level


I think VEMS needs to disable the wiki since most info is so outdated and direct people to the forum for answers.

[Eloff19]

Absolutely! Seems like you're well versed with the VEMS as well... good too see it. I'm trying to make more people aware of this ECU because as you say price/feature it's very hard to beat.

I actually just got http://vems.ca up and running (although still work in progress and website creation is not my strong suit lol) so hopefully that will help as well.

I agree on the outputs, they get used up quickly if you take advantage of most/all the features... I will bring up this point to some of the developers and see what they have to say

Not OT all by the way... I always enjoy a good discussion with level headed people

Site has a nice flavor; I may be able to lend some spare time if you wanted a hand with anything?

Are you in contact with dev's much? I'm eagerly awaiting their new implementation of knock code....just because.

Also do you use stepper outputs for tach? I used the tach output as per the spec sheet and get some nice jumpy/fluttery activities above 4k (no signal issues and the wiring is good) Just curious.

Much easier to route wires versus the tubing!

+1 and my reasoning for going with it.

Also since you no longer need the MAF, you can just steal the no longer used wiring...so no need to even run any extra wires!

VEMS is an awesome unit per the price for sure. Documentation is the weak link for sure. I am wanting to help on that end though since I cant really contribute with code. I also think the effort/benefit would be very nice. Like most people though, I just need the time to sit down and hammer some stuff out.

I hear you there...I'd be happy to send you my excel file with the pinouts/wirecolors etc hookups that I used for my setup (OBDI stock harness). A nice little writeup on how to wire into obd I e36, (and if notorious had something lying around) obd II e36 and your e30 may have a very positive effect in removing the "scariness" from hacking up a harness. At least these files would lessen the time-investment on your end

Not that it's hard to follow the thorough bmw wiring diagrams but maybe that's too much for some people...Just a thought anyway.

I mean you can have all this customizability, stability (good connectors AND software), datalogging, integrated wideband (just to name a few) without having to build the unit yourself...for less than or equal to the price of remote tunes...It is shocking that more people don't go this route. Especially when it seems quite apparent that people are certainly not afraid of DIY'ing.

Couple this with the fact that the unit runs full sequential and aside from one (possibly less than ideal) hack to maintain the pressure signal from the A/C...You would never know the car is running a standalone.

I think VEMS needs to disable the wiki since most info is so outdated and direct people to the forum for answers.

This definitely should happen.

[NOTORIOUS VR]

Site has a nice flavor; I may be able to lend some spare time if you wanted a hand with anything?

Are you in contact with dev's much? I'm eagerly awaiting their new implementation of knock code....just because.

Also do you use stepper outputs for tach? I used the tach output as per the spec sheet and get some nice jumpy/fluttery activities above 4k (no signal issues and the wiring is good) Just curious.

Sure man, anyone that has more experience with web development would be an asset lol... As I have literally zero. I've thrown together a site for myself before and now this, but that's it. Basic of basic :P

I chat with them every so soften, I see that we think a like since I've been hounding them for the new knock code as well. I've heard it is being worked on though

AS for the tach, no I usually drive it with a 259 output and the newer ecu's even come with an internal pull up now on a dedicated 259. But the tach jumping should be fixed in the later 1.2.x FW's, as the process for the tach output was given a higher priority. The jumping around doesn't mean anything is wrong though if that helps

I hear you there...I'd be happy to send you my excel file with the pinouts/wirecolors etc hookups that I used for my setup (OBDI stock harness). A nice little writeup on how to wire into obd I e36, (and if notorious had something lying around) obd II e36 and your e30 may have a very positive effect in removing the "scariness" from hacking up a harness. At least these files would lessen the time-investment on your end

Not that it's hard to follow the thorough bmw wiring diagrams but maybe that's too much for some people...Just a thought anyway.

Here is an .xls that someone sent me (another VEMS guy put it together... was nice to work from.

E36 OBD2 to VEMS: https://docs.google.com/spreadsheet/...E12WC16d2lIY0E

I mean you can have all this customizability, stability (good connectors AND software), datalogging, integrated wideband (just to name a few) without having to build the unit yourself...for less than or equal to the price of remote tunes...It is shocking that more people don't go this route. Especially when it seems quite apparent that people are certainly not afraid of DIY'ing.

Couple this with the fact that the unit runs full sequential and aside from one (possibly less than ideal) hack to maintain the pressure signal from the A/C...You would never know the car is running a standalone.

Agreed

Oh and chikinhed is good peoples... he got that pin out to me already

[gstuning]

Knock setup is alot more then meets the eye. I have setup a proper ecu for knock and it´s certainly not just tossing in some numbers. Loads of datalogging and work goes into that. More so then the fuel and ignition tables for sure. Even 90% of all ecu´s that have knock code don´t ever get used because the work involved just doubles the calibration cost and proper calibration is costly enough. There is no guarentee that calibrated knock code will safe a engine. Let alone half assed assumption calibration.

I am hoping the new boost control code incorporates some of my suggestions which would make the boost control just about the best there is for targetting and hitting the target without failure. I planned it based on my experience with Drive-By-Wire throttle calibration. I know they are implenting a two solenoid opporation which can be very effective in boost control.

Right now literally I´m wrapping up the circuit board design for M50B25 Non vanos VEMS PnP. Auxiliary connector is a 16pin connector and has the serial wires, 5v out, wideband wires, 3 relay controls, 2 high current outputs and 3 analog inputs. As soon as I have quadrupple checked things I can get cranking on making one. It´ll be 100% PnP in Alpha-N mode but will require plumbing the map sensor for speed density.

Getting autocoding to work is mind boggling when a ecu has matured as much as the VEMS has.

A relay output is enough for vanos activation through a relay as there is no PWM-ing going on and plenty enough time for the relay to ON/OFF. Same with other like functions like manifold flaps and similar devices.

[NOTORIOUS VR]

Knock setup is alot more then meets the eye. I have setup a proper ecu for knock and it´s certainly not just tossing in some numbers. Loads of datalogging and work goes into that. More so then the fuel and ignition tables for sure. Even 90% of all ecu´s that have knock code don´t ever get used because the work involved just doubles the calibration cost and proper calibration is costly enough. There is no guarentee that calibrated knock code will safe a engine. Let alone half assed assumption calibration.

This is very true, but it is something I feel is useful if someone puts in the time to tune it. Even with assumptions and trial and error, some knock control might still be better then none unless you're willing to take out 10 deg of advance to be on the super safe side of things.