aem infinity standalone on e36 obd 1

[bry195]

wow, another issue. so I use the analog output as the lambda feedback. it requires calibrating the aem to the voltage and its in lambda not afr. so the gauge reads like 16.5 when i change the lambda target to .55. if I leave it at .98 it runs at an 18afr according to the gauge. I doubt anyone will have this kind of issue because who runs lambda off an external controller? anyways, AEM is researching the problem. could be in a bad set of injector offsets but I don't know. I created the tables from the raw data injector connection provided me with but I had to mentally interpolate some data points like 30 psi for one which really shouldn't be a derived value.

a few questions have come up over the last month or so about the way Im building this. the OEM engine harness is bone stock. the only changes are that I removed the pressure switch and replaced with a pressure sensor. but the connector to the switch is just sitting open circuit under the manifold. IAT, coolant, all of the other OEM sensors are calibrated and working.

so in theory I can unplug the aem harness adaptor. remove the aem infinity and plug in the OEM DME and have a running car after plugging in the OEM DME. I have not done it nor do I need to. I have no emission testing requirements in Florida.

all sensors that are not OEM are ran through an auxiliary connector to the aem. so if I went to back to the DME the aux connector can be disconnected since these sensors are not required for the DME.

If the project works out and people are interested i might build harnesses or have a custom harness manufacturer build them. what I like about this is that the harness can be a bunch of individual circuits with dedicated channel shielding and a one piece over moulding done. they probably cannot source bmw original wiring and color codes like i did but something can be managed.

[bry195]

Im so glad I built the break out board into the project. this saved so many headaches on working out what goes where. however I think the concept of the break out board being inline rather than T'd into the system is requiring too much space even when the temp board is removed. So I think I will just T it if I can.

[Butters Stoch]

3 months now, this thread makes me no want stand alone.

[pbonsalb]

For some, the project is the most fun part. I think bry195 is having lots of fun.

[Def]

3 months now, this thread makes me no want stand alone.

I went back recently and looked at how long it took me to get mine up and running. ~2 weeks of meticulous planning and documentation (I'm an engineer, used to it by now), ~2 weeks to build the harness working on it a few hrs here and there (probably a solid 10-12 hours total).

Install is like any other harness, so I won't include that, but I did make a new accessory harness in the car for 12V and gauges etc. So maybe an extra 2-3 hrs to give a clean, integrated install here.

30 mins tweaking a basemap slightly, setting ignition timing etc.

Then it cranked on the first go and I was driving it around the block shortly after. No O2 feedback either, like a boss.



So it doesn't have to take a long time.



And not to knock bry too hard, but some of his missteps are specifically spelled out in the manual. We all run into hiccups, but I found that rereading the manual from time to time really helped clear things up as I progressed in the project.



As with anything, there's a learning curve, but it's not that bad. Plus now, I can look at ANY sensor and figure out what's going on with the engine/car. Good luck with a ROM tune. When you've got a problem, most people throw TONS of cash and time and frustration at it. It's just the norm, so people accept it.

[bry195]

Shite guys. I built the harness from scratch. I didn't like the direction the connector came off the base unit so I found where AEM buys all the pieces and parts and did it all over again. Im an engineer as well and definitely enjoy the building part. I even used all oem wire on the oem harness stuff so I could look at the manual and understand what I was looking at. then I labeled every wire even though it was color coded.

I'd say the only thing I would do different is the startup without confirming with a light but I didnt have to adjust it at all.

I spent a year tuning the miller package. but it ran flawlessly in the end and I sold it.

I was in no hurry. Def's timetable is a very good estimate.

I can tell you this. when everything is calibrated (2 minutes a sensor) and going VE tuning is very simple.

turns out the table I received for the injectors was not quite right. I selected ID1000's from the list and it fired right up. no bug, just a bad table.

wait until I spend 2 weeks on throttle response and wall wetting if you want to call someone anal retentive..;^/

runs a flat .887 at idle after 5 minutes of ignition timing table adjustment and a few points in the VE table. its gonna piss me off if everything is that simple.

[bry195]

in the end every inch of the project is well documented and will be very easy to service on the back end (because I will rip it apart again). and I won't question my work. If I did this any other way I wouldn't like the results and building is my hobby. not racing. I have the money and time to make 1000hp car but I chose to integrate waste gate position control and turbine rpm control or tuned exhaust or ....20 other weird things that I have moved on from. the stuff I do doesn't influence the direction of the way others are building things in a good way or bad way so I just do it my way.

[Def]

The tuning is really stupid easy, which is a good thing. Because you can be sure the engine is running well, then you can get into more spicy stuff like traction control nuances, more elaborate boost control schemes, safety schemes etc.

With just 1 person doing the tuning, if you've got to put in 500 hrs just to get the engine running, there's no way you can really delve into the more interesting and complicated things the EMS can do. That's the whole point of making the actual tuning itself stupid easy, and why AEM concentrated on making it that way. That then gives the average tuner/owner the time to go in and do some more interesting things without spending days upon days to do stuff.

It's also what makes it so refreshing compared to a ROM tune that's sometimes a struggle to get it running 100% perfect in ALL aspects, especially if it's a remote tune or just a dyno WOT tune (which is easy). See something wrong? Pop in and make a minor change. DONE!

[bry195]

The tuning is really stupid easy, which is a good thing. Because you can be sure the engine is running well, then you can get into more spicy stuff like traction control nuances, more elaborate boost control schemes, safety schemes etc. With just 1 person doing the tuning, if you've got to put in 500 hrs just to get the engine running, there's no way you can really delve into the more interesting and complicated things the EMS can do. That's the whole point of making the actual tuning itself stupid easy, and why AEM concentrated on making it that way. That then gives the average tuner/owner the time to go in and do some more interesting things without spending days upon days to do stuff. It's also what makes it so refreshing compared to a ROM tune that's sometimes a struggle to get it running 100% perfect in ALL aspects, especially if it's a remote tune or just a dyno WOT tune (which is easy). See something wrong? Pop in and make a minor change. DONE!

YES! the changes I made in the VE table for low rpm stuff was rock solid. not a result with dither or mild hunting. Rock solid 2 decimals deep.

[Def]

Yea, it's a good feeling to be able to look through EVERYTHING on the ECU and see exactly what cells it is referencing and then go from there isn't it?

Glad you're enjoying it, I know I like mine.

[bry195]

anyone have a good drawing for a circuit to run PWM through a 3 wire idle control valve?

single wire pulse width modulated ground out from the AEM. IACV is 3 pin. 1 wire is the common hot. 2nd wire is the modulated ground to open the valve 3 wire is the modulated to close the valve. 2 and 3 should not be on off. one should grow in amplitude while the other is shrinking in amplitude. this controls bounce in the valve. a little opening torque is held on while closing torque is applied.

40 ohm 50 watt resistor on one of the phases would hold the valve to one side but I dont like the idea of mounting a big resistor. the invert circuit and tuning from aem is documented poorly.

[bry195]

like usual just as im convinced something doesnt work I gave it one last shot. the aem voltage invertor works just fine. here is what needs to be done. 1 wire monitors the pwm out to close the valve. one wire gets 12volt dc and one gets chassis ground. the last wire is the inverted pwm that that goes to the open side of the valve. if you did the board right it will work. assign the pwm close out to ls3 or whichever low side (ground) output you want. go to the outputs tab in the tuner and look at ls3 in the m3 base tune. set all the dutys to 0. no need to start the car. blow through the iacv. I measure flow from open to closed. went from 0 to 100 duty and 100 to 0 duty and everything in between. this gave me flow values and hysterises when going open to close or close to open. start the car with all duties 0. start adding duty until the idle comes down. tune PID until changes in duty get very responsive. fill in the rest of the duty tables for different rpms.

[bry195]

I have to say the simple little voltage invertor is finicky but simple and very clever of an idea. I looked around and none of the other ems have this style of control.

[bry195]

PID for air and timing are part of the idle control strategy. I had no idea that timing is used to control idle speed as well. 800 rpm idle with 100lb injectors feels like its going to be a joke to setup.

[Bigmac834]

What ICV are you using? Why not use a stock one?

[bry195]

What ICV are you using? Why not use a stock one?

oh, yea. Sorry, its the stock 3 wire as opposed to the stock 2 wire. Im trying to keep the electronics on the motor harness stock. so far so good.

[bry195]

I have begun the process of moving all wiring off of the break out board and into there permanant location. little tuning delay going on. the fuel pressure is stuck at 20 psi so I ordered a new fuel filter. the old corn juice has plugged up the OEM filter. went to the stainless golan from jayracing. hopefully I'll get it this week.

[bry195]

here is a super good standard on wiring connections. NASA style.
http://www.hq.nasa.gov/office/codeq/doctree/87394.pdf

[bry195]

replaced the fuel pressure sensor and added the golan 10 micron filter. 45 psi with the pump on and the car not running.

[bry195]

[bry195]

Im thinking about putting a small ground ditribution block or splice connector at the 4 corners of this adaptor. 1 chassis ground in say 8awg connecting all of the distribution blocks. then I can drop all grounds to chassis at central point. thoughts?

[wazzu70]

Glad to see this is moving along and you are having fun.

Installing a standalone is not very hard. As mentioned Bry is doing a bunch of "extras" to have some fun and experiment along the way.

Just something to keep in mind so people dont think every install is as complicated.

[bry195]

great point. plus I have the attention span of a squirrel.

[bry195]

It looks like everything will fit in the dme box. Car is running again although it has a slight miss but I'm on base maps.

All of those bullet connectors and wires will disappear eventually. They will be replaced by deutsch connectors once I'm sure avert thing is just right.

I do have a potential voltage and difference in resistance between engine ground and dme ground that I'll need to solve but for now we are good. Or as George bush announced in Iraq mission accomplished.

[bry195]

nice box to hide the components in. this is the idle invert circuit for the 3 wire valve. I added a relay to isolate the power from the ecu. probably mount this in the cabin next to all of the other control boxes.