NefMoto

I thought I would get this topic started since so many people with B5 S4s want to install meth injection kits.

So some basic info on meth injection:
1) The fluid you inject is going to be a mixture of methanol and water. The water cools the combustion mixture which reduces knock. The methanol increases the octane rating of the combustion mixture which also reduces knock.
2) Methanol is fuel and by injecting it you will run a richer AFR than the ecu is trying to achieve.
3) Oxygen sensors are lambda sensors, not AFR sensors. They read the amount of oxygen remaining in the exhaust gas. When you inject methanol your oxygen sensors will read rich, because the air fuel mixture is rich.

More to come...

Here is an excellent article that was recently posted on another forum.

This is an article written by Labonte Motorsports regarding Meth Injection 101 in Gasoline Engines

http://www.labontemotorsports.com/ontrack/WM101.pdf

I have also attached the file to this post in case the link breaks.

I understand when adding meth you can pull primary and add timing (and boost) but what modifications do tuners typically make for a custom meth tune? Thanks!

I just wrote my first k03 meth tune.  I sent it to a forum member for testing. 
I made a conservative first attempt.  I added about 6-10* timing or so from about where full boost hits and up... i wanted to make sure his meth kit would be in full spray.  Then I pulled fuel from kflbts in the same region.  I'm sure it needs to be tweaked but it should be a good start.
b5 s4 k03 turbos 17-18 psi spike tapering to 14
91 octane very conservative
(http://i714.photobucket.com/albums/ww142/judeisnotobscure/kfzw291oct.jpg)
91 with meth
(http://i714.photobucket.com/albums/ww142/judeisnotobscure/kfzw2meth.jpg)

91 octane
(http://i714.photobucket.com/albums/ww142/judeisnotobscure/kflbts91oct.jpg)
meth/ conservative since we are not working with wideband yet
(http://i714.photobucket.com/albums/ww142/judeisnotobscure/kflbtsmeth.jpg)

You want to add timing until you get significant timing retard WITHOUTH METH, about 6 deg. This will be your fail safe maximum timing value in the situation when you run out of meth or pump fails etc. Then you turn on the meth and enjoy almost knock free engine.

Having said that I run about 24deg @ 7k rpms of timing... just to put your values into persepctive.

Cheers.

Thanks julex
my 93 timing table is a little more dialed than the 91 table (running about 6* up top), but not for long.
I've been working with another member to fine tune the 91 table... I make very conservative changes when i'm tuning, because i'd rather have to make another revision then hurt someones elses car.

I am with you as far as using wmi for making big power, I will get up to about 25* up top after a couple of test runs.

In another thread nyet suggested using intake air temperature to determine if the car had run out of meth. Then you can use the intake air based timing adjustments to reduce timing in that situation.

link to that thread please, or the map names, and or adresses.
that sounds like the smartest way of doing it.

Post #13 by nyet in this thread.
http://www.nefariousmotorsports.com/forum/index.php?topic=474

Is that a kfzw map... I didn't see the iat correction map for timing in my xdf. 
I guess i could use kftarx and just cut load.  that will work just as well.

As far as I know there is no map that allows you to change timing based on intake air temperature. The best you can do is reduce desired load or maximum boost pressure based on intake air temperature.

I agree, and i will incorporate this into my meth tunes.  I can't wait to get meth on my car... should be within a month.

Question is, can you really tell if your WI is empty (or dead) by just IAT?

If you make it "safe" enough, will you lose power even if WI is properly working? Or do you just dial back enough boost to keep your CFs from maxing out when WI isn't working?

My post was hypothetical; I dont' know if anybody has done it.

Personally I would rather tune the maps that reduce load and timing based on knock. If you run out of meth, your knock will go up, which would be tuned to dial back your load and timing.

Good points... I think I will just have to do some testing to figure out the best combo of knock intervention and iat correction.  I'll post back with results after my meth setup is installed.

You already have a map for max load during continous knock - this is your failsafe.  Also, increase the maximum retard from 15 degrees to 15+x, where x is the extra amount of timing added due to your octane booster.

Rick.

Exactly what maps need to be adjusted to take advantage of Meth?

It seems the general consensus is:
KFZW
KFZW2
KFLBTS

Is that it?

TABGBTS is stock at 499.99, and FBSTABGM is stock at all 1. I don't believe these should need to be adjusted, correct?

I have not touched KFFDLBTS either. Does this need adjustment? I do not want to disable any safety features.

I am putting on my ARD bipipes today and will take some new logs. Then, I will be making adjustments accordingly.

Any insight is greatly appreciated.

Josh.... If I have time, I will try to get some WiFi somewhere and take some logs of your file as well.

Timing maps should really be the only thing you should mock around. The base maps (KFZW, KFZW2) are the starting point ECU is taking timing values from. In the absence of knock it will adjust timing towards KFZWOP and KFZWOP2 maps. But it is lengthy process as it will take several hudreds of RPMs for this to make visible change to higher timing...

If you adjust KFZW and KFZW2 to much higher values to begin with (for high load condition), it will give ECU higher timing starting point. ECU will still retard timing even below (up to 12 deg but you can adjust that too) values in KFZW/2 if needed so you're only giving it a "head start" here.

If you run out of meth, no problem as well as it will just retard more.

In my experience, it is a good idea to either tune without running meth until you hit -6 deg retard which will be you safety in case you run out of meth, or tune with meth but make sure you don't hit more than -3deg retard across the board.

This is for every day DD car... now if you're tracking or at least absolutely sure you will never run out of meth... you can go all out with timing...

KRMXN is the map for max timing retard... as mentioned you can increase this as a safety measure.  I will be adding this to my meth tune.

perhaps this is not the right place to ask but as I recall, the air temp sensor is post intercooler so really the only way to use that sensor as a safety measure for WI would be to spray pre intercooler or preturbo which seemed risky but I know tons of people do it, at least on occasional drag type runs but not so much for sustained periods at WOT at the road track.

has anyone bench tested these popular WI pumps? What psi can they produce w/ nozzle to feed KO4s say 450hp? I also wonder where people are spraying? At the high boost levels you need a lot of pressure to properly atomize post turbo. I was always worried that at 50,000 rpm spraying pre turbo could wear out the turbo, spraying post turbo pre IC would cause the meth/water to condense on the intercooler, loose suspension and just turn into water droplets.  Sure this cools the intercooler but it is not the same as injecting atomized meth/water in the motor pre TB.  afew years ago I was spraying at very high pressure with a pump I seacrched high and low for. the pumps most non AM people seem to be designed for high volume low pressure so i wonder if they are the best choice?

 

Its in the TB, post TP.


Not sure where you're going with this?

I believe the popular pumps are 200psi these days.... seems high enough for me.

I currently run 0.7mm nozzles in my injection which are recommended for k04s ~22psi and I tested them while out of bi-pipes. They produce very fine mist, visually looked good enough considering it will be snatched by 100mph (or whatever it is a wot) WARM/HOT air current and evaporated/atomized further.

Going to post some thoughts.

If you really want a big benefit from meth, you need to run a lot of it.
Running a lot of it means that you need take out some fuel, or it will run too rich.
Plus the timing.

Looking at the Aquamist HFS-3 kit, it has a level sensor and a flow sensor, to make meth injection safe. Usually with the Aquamist HFS-3 kit fitted onto these cars, it is wired into the N75 circuitry, disabling N75 if meth runs out.

While this approach is fairly safe, and you won't blow up your car even if running lots of meth, it means the car can't run properly without meth.

---

Here is how it should be solved, and how I solved it on the Audi 2.2T 5 cylinders.
The 2.2T ECU has free logic inputs on the chip calculating fueling and timing.
So I put a relay inside the ECU and made the meth kit (or for example LPG kit) toggle the relay. And the relay switched that port on the processor between ground and Vref (+5V). To have 100% safety against broken wires, I made it so, that +5V needed to be present on the processor pin and the relay needed to be activated to switch to the "meth" maps...

The master table in the ECU was at 0x8800. I added another table at 0xE800 (an exact duplicate of the first one), and then went into the map access routine in the ASM code, and added simple logic, that just adds 0x60 to the high byte of the table address before accessing it based on whether the port connected to the relay is high or low.

After this, I made copies of the maps I wanted to be different between the two tunes and changed the pointers in the second table to point to them.

---

So, I want to do the same on the ME7, because I don't want my RS4 turn into a dog every time I run out of meth.

I haven't looked much at the hardware or the software of this ECU yet to be honest (meaning, I have not loaded it up in IDA Pro yet and I have not looked at the PCB), but perhaps one of you that has wants to help me, and well really everyone around here who is running meth with a kit that has a flow sensor.

So let's start with the questions - does the main processor have any logic ports that are not connected to anything?
If not, we can do this for example by removing one EGT sensor and writing extra code that reads that input, while averaging the second sensor over both banks...

The plus of this approach is, that with a flow sensor on the meth line, you actually switch maps BEFORE knock and possible damage happens. So it is 99.9% safe and it is proactive instead of reactive.

You could use the post cat lambda input.

Okay, that's a good idea, but how is this input presented to the processor itself?
Is it directly available on a pin? Or does it go through some sort of additional I/O processor?
It is probably processed by an ADC, isn't it?

I guess disassembling the instructions responsible for CAT efficiency codes would shed some light on it.

Looking here:
http://www.nefariousmotorsports.com/wiki/index.php/Reverse_Engineering_ME7.5 (http://www.nefariousmotorsports.com/wiki/index.php/Reverse_Engineering_ME7.5)

It seems Narrowband input is on P5.7/AN7.
Considering narrowband is used in ME7.5 for CAT checks, it would make sense that the rear O2's are also available on P5 in ME7.1...

Okay, so we have a free input. Now I need to dig into the code a bit.
If taking fuel out, either an additional map needs to be added for fuel taper or just the main fuel map duplicated.
Same goes for ign map as well I guess.

So:
1. I need to find the map access routine, and see how it works.
2. Patch some code into it.
3. Profit.

Is anyone sharing ida pro databases on here? Or is that still considered hush-hush info?
I don't really care if I have to start from scratch, just takes more time.

I can double check which pin the post cat oxygen sensors connect to the processor on the ME7. I have the info written down on my other computer.

There are multiple map access routines based on if the map is 1D or 2D and what the data types of the axis and map data is.

So far no one has shared any IDA files. I have thought of posting a basic IDA file with the correct config and memory mapping, but I haven't had the time or the motivation to do it.

In the B5 S4 ME7.1 ECU:
C167 P5.7/AN7             connected to ECU pin 69 connected to Pin 4 G130 Oxygen sensor 1 behind three way catalytic convertor - (O2S) 1 - Postcat
C167 P5.10/AN10/T6EUD connected to ECU pin 11 connected to Pin 4 G131 Oxygen sensor 2 behind three way catalytic convertor - (O2S) 2 - Postcat

I traced these connections a long time ago, because I was also planning on using them as additional inputs.

Cool, thanks for the info Tony.

I guess only the main fuel and timing maps really need to be patched.
If running more boost, the boost maps as well.

In M2.3.2 there were only two map access routines - one with interpolation enabled and the other with interpolation disabled, so it was easy mode.
I'll start with IDA once I port my RS4's factory map to M-box. Seems that this is what everyone is using for some reason, and makes no sense to develop elsewhere I guess?

Getting back to this...
Tony, I remember you posting some map access routine names and parameters, but I can't seem to find the post anymore.

With little hopes to anyone replying after so many years, but let's try:

I'd like to run "mild" water/meth injection, just for safety. I am happy with power output on gasoline, I am not happy with EGT (~ 1000C) and I just want to lower it all together under high boost and high rpm.

My car has wideband and lambda control at WOT. If it detects rich, wouldn't it simply cut down fuel trying to meet target lambda automatically? Why would it be needed to change maps for less fuel to compensate for water/meth?

Beacuse at lambda 0.9 you need 6:1 AFR for methanol
So the ECU will try to reach 0.9 lambda but for that you need twice more injected methanol

Look at table, don't forget ecu measure lambda

Makes sense, thank you!

If you run pure methanol...
Adding a bit of methanol on top of gas should be a completly different story.

Essentially this is what I am intending, just to keep EGT below or around 900C tops. My EGT gauge is actually a PID controller with a built in relay and I intend to wire the pump for meth to that and only spray as long as I am over 880C for example. By default the meth is boost controlled, which I don't need. I don't think I will make any changes to the fuel maps with the intended setup.