[Nezil]

Originally Posted by msmiljanic

I don't understand this. If the jet screws into the bung hole by, as an example, 4mm, then the spray hole only has 4.1mm left until the inner wall is reached. Through their diagram, the normal ones would be fine right?
I asked AA about mine and they said 2 mm thickness and a thread depth of 8mm. So I assume the jet will screw 4-6mm into the bung hole leaving 4-6 mm to the inner wall which means the normal ones are good since they say a 6mm max. I would assume this is fairly standard for charge pipes and the extenders are really for exceptions to the rule.

Well there are a few things at play here...

First is that the 1/8 NPT threads are tapered, so how far you can screw in really depends on how deep the bung was tapped in the first place. Unlike a straight thread, there is no real way to know how far the jet is going to end up getting screwed in, and how tight you do it in and if you use any sealing tape will all affect this.

Second is that every brand of jet is made slightly different. The cone shape, and atomisation will all be different.

Third, and probably most important for my application since I'm using Aquamist jets, which are machined in house by Aquamist staff, are M8 x 0.75 thread, which is a metric straight thread with a finer than normal pitch, and smaller diameter than 1/8 NPT.

Aquamist provide 1/8 NPT to M8 x 0.75 adapters and blanking plugs with every kit, and these can be used either directly into tapped holes in the charge pipe body, or into 1/8 NPT bungs already fitted to the charge pipe.

I'd rather not tap new holes when I have two perfectly good bungs already provided, so I'm going to use the bungs. The trouble is that based on the information I got from FTP, which is probably pretty standard in the charge pipe world, is that the bungs themselves are 6mm thick, and these are welded on top of the charge pipe, which itself is 2.1mm thick. A total thickness of 8.1mm.

If you can screw the 1/8 NPT jets you have into the bungs nearly 8mm, you'll be fine. My case is different...

The 1/8 NPT to M8 x 0.75 adapters appear to have about 4mm of thread, as do the jets themselves. My situation is going to be like the 4th or 5th diagram in the chart you quoted because I'm using a third party NPT bung with a 2.1mm pipe (4th diagram), or to look at it another way, the overall depth is 8.1mm (5th diagram).

[boostm3]

This is an awesome project Nezil. Back around 2000 when I had supercharged my e36 M3, pictured in my signature, and increased the boost, I used two forms of cooling: I used a water to air aftercooler, and I used water/methanol injection courtesy of Aquamist, purchased from Active Autowerkes. The water injection was amazing. I just used a single nipple into the intake a few inches in front of the throttle body. But the effects were staggering.. IATs dropped from 50 to 100 degrees when the system triggered. On 80+ degree days, Id routinely see IATs of ambient plus only about 30 degrees! Without the WI, IATs would commonly go as high as ambient + 100. Im sure youll get tremendous results with yours.

[Nezil]

Originally Posted by boostm3

This is an awesome project Nezil. Back around 2000 when I had supercharged my e36 M3, pictured in my signature, and increased the boost, I used two forms of cooling: I used a water to air aftercooler, and I used water/methanol injection courtesy of Aquamist, purchased from Active Autowerkes. The water injection was amazing. I just used a single nipple into the intake a few inches in front of the throttle body. But the effects were staggering.. IATs dropped from 50 to 100 degrees when the system triggered. On 80+ degree days, Id routinely see IATs of ambient plus only about 30 degrees! Without the WI, IATs would commonly go as high as ambient + 100. Im sure youll get tremendous results with yours.

Thanks for the kind words. I'm looking forward to doing the install.

I think most other users of WMI are looking for something only to cool IATs, and only then on long WOT pulls. My needs are somewhat different because I'm looking for an octane boost and perhaps even some supplementary fuelling. These are the main reasons I've gone with Aquamist because it is progressive and aligned with regular fuelling.

I'm happy to see that others are getting great results from more simple systems based purely on boost or controlling flow based on modulating the pump etc., but I'm looking for every day drivability as well as improved and consistent performance.

We'll have to see how well my system achieves those goals once installed!

[Nezil]

I do also have some updates on my progress...

I'm currently on vacation in Maui, returning on Monday 1st, but before I left, and while I've been away, several parts have arrived:

• 1 Gallon empty bottles for mixing water & meth - 2 Received; 1 mixed up with 50/50 Water / Methanol
• 1/8" split ribbed tubing for under hood cable tidying - 10' Received
• EWG Male and Female connectors - 5 sets of each ordered from AliExpress
• TMAP Male and Female connectors - 2 sets of each ordered from AliExpress
  
• Injector Male connectors - 1 pair from a Mini heated windshield washer harness received and tested; work great!
• Injector Female connectors - 5 sets ordered from AliExpress
  
• 6mm carbon fibre braid - 4m received
• 4mm carbon fibre braid - 3m received
• 6mm PTFE tubing - 10' received
• Fuel Rail Pressure Sensor Male & Female connectors received
• Glue filled heat shrink tubing received
• M10 Stiffening bolts (33326768354) - 8 ordered from Brian Keller @ Sun Motors; I know I'm going to have to remove this plate during the install, so better to have the single use bolts ready!
• All my Aquamist stuff, still waiting to be shipped

In addition, I've ordered enough parts to prototype a more elegant fail-safe. I've also designed a small PCB layout for this, assuming everything works, but I won't be ordering that until I actually test it on solder-less bread-board, and then on solder prototyping board. I've ordered 4 sets of components, so if no changes are necessary, I have enough to build a few sets.

I'll post separately about the details of the fail safe.

[Nezil]

The Aqaumist system offers several methods of implementing a fail-safe in the event of flow being either too high or too low based on thresholds that you set during installation:

• A 'map switch' single wire that can be set to go to several different voltages on a fail condition
• A relay which can make or break on a fail condition
• A flow output, which can be analysed by a third party system to trigger a fail safe in that system

The fail-safe setup suggested by Aquamist for N55 EWG motors is to have the 'map switch' wire connected to the relay, and have this bridge the EWG position sensor output as it goes to the DME. Under a fail condition, the DME will be sent a 5v signal, suggesting that the EWG is locked in a single position. This will cause the DME to put the car into a drive-train 'limp' mode, cutting power aggressively and protecting the motor. This works as a fail-safe, and a re-start should fix the limp mode, but it's a bit aggressive for my liking.

I know that simple piggyback tuning devices work by biasing the sensor voltage of the TMAP sensor so that the DME thinks less boost is present than it really is, upping boost to compensate and relying on the remaining engine sensors to keep everything safe.

What I was looking for then, is a sort of piggyback in reverse, where the TMAP sensor is biased higher than actual, resulting in the DME thinking that more boost is present than actual, reducing boost to compensate, and using this as a fail-safe. I posted about this separately in this post: https://f87.bimmerpost.com/forums/sh....php?t=1624996.

So here are the ISTA+ details of the TMAP Sensor output. I'm sure it's not absolutely perfectly linear like is shown here, but that's the design intent:

The output is absolute pressure rather than purely boost, with Atmospheric pressure at sea level being 1013.25 hPa, or ~2.0v on our 2.5bar sensor. Anything less than 2.0v is vacuum, and shouldn't be altered. Anything above is boost, and can be.

Here is the pinout of the TMAP, also from ISTA+:

There are a number of different ways that a circuit could be designed to achieve this behaviour, but I decided to use Op-amps and resistors. I designed the circuit with the following characteristics:

• To act on a fail-safe signal of 0v, a condition OK signal of 5v
• When in 'OK' condition, to not modify the signal at all
• When in 'Fail-safe' condition, to not modify the signal below a set point, and to add to the signal above that point
• To provide adjustment (with a trimmer) to the set point above which the signal is modified, and below which it is not
• To provide adjustment (with a trimmer) to the amount of gain applied to the signal above the set point

I'll post the schematic explaining the operation in another post.

[Nezil]

So here's the schematic I'm trialling first:

I'm using LMV324 Op-amps because they're available as an automotive grade part, and they're able to be driven near rail to rail.

I'm going to be using the 5v power to the TMAP sensor to drive the circuit, and a brief description of the operation is this:

U1A, the first op-amp, effectively takes the sensor input, and subtracts the voltage set by the trimmer RV1. This trimmer is the one that sets the point at which the sensor voltage is biased higher, and anything below is un-changed. It's therefore important to set this voltage at ambient or higher, equal to something above ~ 1.9v.

U1B, the second op-amp, takes the fail-safe output from the Aqaumist system (set to 5v) and subtracts this voltage from the output of U1A. If the Aquamist system is operating correctly 5v will be subtracted, and the output of U1B will simply be 0. If the fail-safe signal is active (0v), the voltage from U1A will be passed through U1B as-is.

U1C, the third op-amp, provides a gain for the signal so far, adjustable with trimmer RV2. The minimum gain is 1.0, so effectively, as we'll see later, the boost amount will be doubled as a minimum for the fail-safe. I'll have to experiment with gain to see how aggressive a fail safe I'd like.

U1D, the forth op-amp, adds the signal from U1C to the sensor signal. In normal use, where the Aquamist system is operating correctly, there will be nothing added because the second op-amp would have reduced the added signal to 0. In a fail safe condition, the signal from U1C will be zero up to the point set in RV1, then increasing with a gain of +1.0 upwards, clipping at the op-amp clip point of just under 5v.

Modelled in Excel, the orange trace should look something like this during fail-safe condition; blue trace is the sensor output or non-fail safe condition:

[Nezil]

Big updates today...

Firstly, my Aquamist parts all arrived from the UK. Fittings and jets look spectacular. The jets in particular are machined in multiple pieces with springs and sections to help improve atomisation. I can see why Aquamist's jets are well regarded. More on this lot later though, because the other updates I have relate to the charts above etc.

My first attempt at a fail-safe circuit didn't work for two reasons:

1. The circuit diagram above has some of the Op-amps connected with the inverting and non-inverting inputs the wrong way round (school boy error!)
2. The rail-to-rail op-amps I selected result in the modified MAP sensor voltage clipping at the supply voltage, giving a DME error that the MAP sensor has shorted to positive; the real MAP sensor voltage never exceeds about 4.5v, so this does make sense

I went back to the drawing board slightly and re-designed the circuit so that it has the same boost gain characteristics, but now clips at a voltage set with another trimmer. I tested this just now for about a 1 mile drive, with the clip point set to 4.6v, and didn't get any errors at all.

I'm going to drive to work and back tomorrow with my test circuit connected. If I don't get any errors on my commute, I think I can safely say that the circuit design is working well.

One of the other things I changed in the design, is I'm going to use a relay to switch between the un-altered sensor signal (when an 'OK' voltage is applied to the relay), and the modified fail-safe signal (when no voltage, or Ground, is applied to the relay.

I haven't decided yet if I'm going to include a relay on the PCB, which would allow it to be used by anyone for any fail-safe purpose as long as they have a 12v switched signal, or to simply use the fail-safe relay included in the Aquamist controller. I'm sure both approaches will work fine, and I'm working out the pros and cons before I go and get PCBs made up.

Here's some eye-candy of the test circuit with factory connectors on my desk, and installed in the car for testing:

[Nezil]

I’ve made quite a bit of progress over the last few days.

Starting with the wiring of the Aquamist controller.

My original plan was to install the controller in under the dashboard on the drivers side, where there is quite a bit more space, but with my car being a 6MT, there isn’t any easy grommets available to run the wires through the bulkhead.

Thankfully, the 2 series body is made for both RHD and LHD markets, and all of the holes in the bulkhead used on my LHD model are plugged with blank grommets on the passenger side.

The front fuse box is also located on the passenger side under the hood on both LHD and RHD vehicles, and uses the large hole for the steering column.

I therefore decided to locate the Aquamist controller just above the Front Electric Module (FEM) on the passenger side underneath the plastic cover in the footwell. The wires then run through a grommet in the bulkhead out to the engine bay, and I can also tap into power from the front fuse box and the lighting signal from the FEM all in the same location.

In case it wasn't obvious, the controller is the black box with a silver sticker that says HFS4 v3.1 (upside down in the photo above).

[Nezil]

Sticking with my goal of trying not to cut or tap into any of the factory wiring harness, I started looking for a 12v switched ignition power source. The front fuse box is the obvious place to start, but there is no space there for an ‘Add-a-Fuse’ and no way of getting the wiring out of the top of the fuse box and back into the passenger footwell without risking the integrity of the waterproofing.

The front fuse box is clipped into the top of a sealed box, with all of the wires being fed from the inside of the car through the passenger footwell using the hole where the steering column would be, so I started investigating if it would be possible to take power from the bottom of the fuse box rather than the top…

You can see the hole through the bulkhead and grommet (not yet fitted back in place) to the left of the fuse box in the image above as well.

I started by consulting the fuse card located in the rear fuse box to see if I could work out which fuses were going to be ignition power only, but found the icons provided to be confusing at best, and therefore turned to the wiring diagrams in ISTA+.

I ended up doing a series of wild card text searches for all of the fuses in the front fuse box, and reading the resulting wiring diagrams to create a better index of what each fuse does in a spreadsheet.

While I was doing this, it occurred to me that this might be useful to others in the community, so I switched to Google Sheets so that the data is publicly accessing here: https://docs.google.com/spreadsheets...it?usp=sharing

To my surprise, not every fuse in the fuse box was searchable in ISTA+; some of the fuse numbers just didn’t show up in the wiring diagrams at all. In any case, I filled in the spreadsheet as best as I could, making notes where no information was available.

One of the other things available when looking at the wiring diagrams, is information on where the fuse is connected to ‘up-stream’, and it became obvious that the fuses were often grouped connected to a common relay, which was activated usually by the FEM, for either global power on when you open the door, or Ignition on when you press the start button. I noted this data when it was clear, in the spreadsheet as well.

After doing all of this work, there were a group of fuses in the front fuse box which were all connected to the ‘15N’ virtual terminal, which is BMW speak for the Ignition On output of a relay. Some of these fuses had no wiring diagrams associated with them in ISTA+, and I wondered if this might mean that they were available to be used for the Aquamist controller.

[Nezil]

As I said earlier, the front fuse box is clipped into the top of a water-tight enclosure, and can be lifted up slightly to look at the back side and the connectors there. To my surprise, many of the fuses present in the fuse box, don’t have any wires going to the back side of them!

I updated the spreadsheet with fuses that were not connected on my vehicle (US Spec LCI M2 with Executive Package), and then selected fuse number 47 because it was not found on any ISTA+ wiring diagrams, it had no wire going to it, and it's part of a group of fuses connected to virtual terminal 15N.

In order to connect to fuse 47, I was going to need a crimp terminal that I could fit into the connector underneath. Fortunately RealOEM turned up a list of 4 crimp parts associated with the fuse boxes for our cars:

• 61130007438 - crimp with wire attached
• 61131378906 - 0.5 ~ 1.0mm crimp
• 61138377732 - 1.5 ~ 2.5mm crimp
• 61138377734 - 4.0 ~ 6.0mm crimp

The first dealer I tried had the third part in stock, which turned out to be the crimp terminal for the larger fuses (fuse 47 is the small type), so that was the wrong part for my application:

The second dealer I tried had the first part in stock, and though I’d probably have preferred just the crimp, and I think the second part number would have been correct, at $4.00 for the part, I wasn’t complaining and this fit perfectly, sliding into the connector and providing me with an additional 12v Ignition switched positive in the passenger footwell.

[Nezil]

I bought TESA brand felt tape and fabric tape to wrap the wires for an OEM look where appropriate. I therefore wrapped the bundle of wires going through to the engine bay in felt tape before the grommet inside the car, and fabric tape under the hood.

I unclipped the pump relay connector so that I was left with a smaller set of terminals to thread through the various areas of the engine bay, and separated this harness from the others because I’m planning to mount the pump behind the front bumper on the passenger side, so this harness will be going straight forward whereas the other harness (Fast Active PWM Valve, Flow Sensor, TMAP, Fuel Rail and Injector connections will all be heading over to the drivers side of the engine bay.

There is a nice plastic tube (centre of the image below) that carries a wiring harness from the area around the front fuse box to the engine bay area, and I was able to feed the pump power harness through there (once I’d removed the relay connector of course).

The four other cables were wrapped in TESA fabric tape, and tucked between the strut brace bar and foam strip to go over to the other side of the engine bay.

[Nezil]

That’s where I’m up to so far. Hopefully I’ll have some time this weekend to remove the front bumper, mount the pump, drill the washer reservoir and run the piping in that area.

I’m still trying to think of an elegant way to connect 12v power to the grey ECU harness, and where to connect ground for the pump and controller. I’m planning to take a permanent 12v for the pump from the feed wire to the cooling fan relay which is located on the passenger inner wheel arch and is bolted in with a beefy ring connector, and I believe there are several grounding points in the same area. Inside the passenger footwell I believe there are grounding points as well, but I’ve not looked for them yet.

[Nezil]

So loads of progress over the weekend. The system is installed, tested and functioning.

It's too early to tell yet if the WMI provides any benefit to the OTS maps. I'm still using the ACN Stage 2 91 map and it's pretty conservative so there's likely no benefit at all. The next stage is to start capturing logs both with and without WMI and see the differences.

There are a few additional things I'd like to finish / improve on with my install, and I'll discuss these in the following posts, but so far at least, things look great.

[Nezil]

Starting where I left off in my previous post. The first thing I needed to do to tidy up the internal wiring was to find a ground inside the passenger footwell, and to connect the 12v ignition switched fused wire I added to the fuse box to both the controller, and the appropriate wire on the grey harness.

The end of the grey harness is under the hood in the area at the back under the plastic cover in front of the driver's side. I therefore needed to run an additional 12v wire with the bundle I'd already made, which rather annoyingly meant un-wrapping the Tesa fabric tape from the bundle and re-doing it.

The Tesa fabric tape is great because it looks perfectly OEM, and is wear and heat resistant. What's not so good about it is the glue is seriously gooey, leaving loads of residue if you try and peel it off. If you think insulating tape is bad, this is horrific, but once it's on, it looks great.

The carpet in the passenger footwell is thick, and under it there are pieces of sculpted foam. I was looking for a ground plate that has several wires going to it, because I was hoping that there would be some spare terminals available to use. I found the plate, but it was fully populated. Fortunately there was no reason I couldn't use a ring crimp and fit this under the nut:

Once this was done, I made sure that all of the cables inside the vehicle had the correct Tesa felt tape applied, resulting in a harness that looks like this:

Last thing on the wiring, I worked out the routing for the sensor wires required for the Aquamist (Fuel Rail Pressure, Injector Duty Cycle & MAP) and built up a harness to completely avoid tapping into the factory harness or connectors. As this was under the hood and around the engine, the factory look called for corrugated split tubing; I used 1/8" for these wires:

The MAP sensor is optional for Aquamist, and not used by default. I'm still working on the fail-safe which will tap into this so for this reason I haven't yet connected the MAP sensor and that's why it's not in the harness above.

The 12v, and sensor wires harness was connected to the Aquamist bundle coming from the controller. I've currently soldered and glue filled heat shrunk these connections, but will be replacing them with a waterproof 5-way connector at some point in the future.

[dtmurf]

I didn't understand a lot of the technical WMI details but I read every word and hope it works exactly as you want it to! I appreciate the attention to detail you put into your projects and the sharing of your knowledge... that spreadsheet is a keeper for sure!

[Nezil]

Originally Posted by dtmurf

I didn't understand a lot of the technical WMI details but I read every word and hope it works exactly as you want it to! I appreciate the attention to detail you put into your projects and the sharing of your knowledge... that spreadsheet is a keeper for sure!

Thanks dtmurf, I feel like I get a lot from the BMW community, so passing it forward is always my aim.

Having said that, finding time is tough! My install was finished over the weekend, and I've not found the time to post up the photos and details.

One thing I can say for certain is that I'm seeing two massive benefits now that I have the system installed:

• IATs drop really really fast - I can be sat with the car idling in the CA heat (about 30C / 85F ambient during the day), and the IATs will climb to over 40C / 105F pretty quickly. This isn't exactly heat soak territory yet, but it's less dense air for sure. If the WMI kicks in, within less than a second, the IATs are below ambient (24C / 75F). I have a CSF FMIC, but with WMI, I'm not even sure it's necessary any more!
• Power! - WMI shouldn't add any significant power without being tuned for it, but I've switched back to running the BM3 Stage 2 91 map, and the difference between without WMI and with WMI is subjectively huge! I've only taken the shortest of logs, and I do need to really focus on that over the next few weeks, but I'm convinced that on CA gas, we're getting well below the potential of most of the maps for our cars. The difference with the ACN map wasn't as pronounced, and that makes sense because it is less aggressive, and less octane dependant. My goal was to see similar performance to that seen by European M2 owners, and I think I am now... and the difference is massive!

[eeyang92]

Originally Posted by Nezil

Thanks dtmurf, I feel like I get a lot from the BMW community, so passing it forward is always my aim.

Having said that, finding time is tough! My install was finished over the weekend, and I've not found the time to post up the photos and details.

One thing I can say for certain is that I'm seeing two massive benefits now that I have the system installed:

• IATs drop really really fast - I can be sat with the car idling in the CA heat (about 30C / 85F ambient during the day), and the IATs will climb to over 40C / 105F pretty quickly. This isn't exactly heat soak territory yet, but it's less dense air for sure. If the WMI kicks in, within less than a second, the IATs are below ambient (24C / 75F). I have a CSF FMIC, but with WMI, I'm not even sure it's necessary any more!
• Power! - WMI shouldn't add any significant power without being tuned for it, but I've switched back to running the BM3 Stage 2 91 map, and the difference between without WMI and with WMI is subjectively huge! I've only taken the shortest of logs, and I do need to really focus on that over the next few weeks, but I'm convinced that on CA gas, we're getting well below the potential of most of the maps for our cars. The difference with the ACN map wasn't as pronounced, and that makes sense because it is less aggressive, and less octane dependant. My goal was to see similar performance to that seen by European M2 owners, and I think I am now... and the difference is massive!

Awesome! Glad it's all working!

Being in CA as well and hearing all the horror stories about our gas I've been hesitant to even think about tuning, but WMI definitely sounds very appealing.

I'm interested to find out what kind of gains you made just with WMI making up for our lower octane.

[Nezil]

Originally Posted by eeyang92

I'm interested to find out what kind of gains you made just with WMI making up for our lower octane.

If you're offering to pay for some Dyno time... I'll pay for the fuel

[Nezil]

The installation has been finished for at least a week now, and I apologise for not posting the next parts sooner... Here's a start:

Now that much of the wiring was in place, it was time to start on the tubing and water / methanol fittings etc.

I'd already scoped out the possibility of fitting the pump inside the front bumper, and using the washer fluid reservoir, and decided to go that route at least initially.

First off, I located what I felt was the best location for a 22mm hole for the tank adaptor. Aquamist, like everyone else, supplies an adaptor that doesn't require you to use any sealant, and works by compressing a rubber seal against both sides of the hole as you tighten the external nut.

I'm not 100% happy with the result. The hole is very close to the edge of the reservoir on the left in the photo below. It leaked until I tightened the seal a little more, and though it hasn't leaked since, I feel that a better seal would have been possible if the mounting face had been totally flat.

In any case, the tank adaptor fitted, sealed and I used some 90 degree fittings to route the 6mm pipe up in the same direction as the headlight washer jet pipe over the top of the wheel arch and around to the front bumper. As you can see, to make it look OEM, the 6mm pipe is wrapped in 1/4" corrugated black sleeving.

I'd initially planned to mount the pump in place of the horn, but in the end decided to mount it behind and to the main bumper support beam. You can see cable ties holding the pump in place in the photo below, but these were replaced before the bumper was replaced.

Aquamist doesn't warranty placing the pump anywhere outside of the passenger compartment, and recommends the trunk. I'm taking a chance by mounting it here, but I live in California, where it rarely rains, and the location I've chosen doesn't get wet, unless I were to drive through a flood. For now, it's a risk I'm willing to take; I'll buy a new pump and re-think the strategy if it ever fails.

With the pump location decided, I cut the Aquamist harness length down, and taped it up to line up with the 12v and ground points I'd chosen.

Bolted to the side of the front passenger fender is a huge relay for the cooling fan. This of course has a big fat 12v supply to it, secured with a nut, and perfect for a ring crimp connector.

Under the fender is a ground point similar to the one I used inside the passenger footwell. Again I used a ring connector under the nut holding the plate in place.

For the 6mm tubing over to the PWM valve, I chose to cover the tube in carbon fiber braid whilst in the engine bay, and 1/4" corrugated split sleeving under the wheel arch. After measuring the needed length, I cut the tube and applied the braid:

The tubing runs up the wheel arch in the same direction in came towards the reservoir, but this time goes up into the engine bay, into the rear bulkhead section and around to the other side of the engine bay where I mounted the PWM valve.

The PWM valve is currently mounted to the cover for the power distribution panel in the engine bay. This is the only place where the Aquamist installation is obvious, and I may change this at some point to make it more discrete.

The install and setup process involves calibrating the flow meter to show 5 or 6 bars when your jets are flowing 100%. To do this you set the jets on the windshield and fill the tank with 100% water. There is a switch on the controller to get it to just run flat out while you adjust the gauge trimmer.

After this, you start the engine, turn down the injector duty cycle % start point to the lowest position (normal operating position is ~40%) and check the progressive behaviour of the system under light load (rev while stationary) and if necessary go for a drive.

Once all that is done, you can fit the jets to the charge pipe. I have an FTP charge pipe, which has two bungs. One is accessible from the top of the engine bay, but the one closest to the FMIC is only accessible if you remove the front stiffening plate (always replace the torque to yield bolts!).

I carefully measured the length of 4mm tubing I'd need, then made up the set with carbon fiber braiding:

It was then a case of fitting the jets to the charge pipe. I used 90 degree fittings to make it neater. Here's the top jet installed:

And the lower one, accessible between the steering rack and sub-frame once the stiffening plate is out of the way:

And that brings us up to date I think. One thing I didn't mention is that I have now connected the MAP sensor output to the Aquamist controller, and fitted a 5 way waterproof connector (a spare connector set I ordered for the EWG) between the engine loom and the Aquamist wiring.

I'll end this post with one more photo... Obviously the PWM valve is visible here, but can you spot the wiring for the Aquamist system in this photo:

[Nezil]

Here are the non-oem / Aquamist PNP harness I made highlighted:

I can't believe there was plastic clips and guides for the 1/8" corrugated sleeving already there on top of the engine!

[eeyang92]

Originally Posted by Nezil

If you're offering to pay for some Dyno time... I'll pay for the fuel

I have no idea how much a dyno costs haha

If it's reasonable, I'd probably be down - after all, you've given all of us an awesome WMI guide.

[Poochie]

I just skimmed through your project, very detail, intricate reverse-engineered here; it's going to take me a minute to sort through all of it..

I have a blank Dinantronics ECU (No harness) for the M2; you can have it for your further research, if you're willing to cover the shipping cost..