Fuel controller

Hi there...

I have a home made manometer which uses hydraulic fluid in transparent piping...its very accurate at getting the balance spot on as trailled on a number of different bikes...the only problem such a device has is that its accurate at BALANCING rather than identifying the pressures on both sides (if both vacuums were too low you would not know, if they were equally low)

R.

Sounds like you're on top of that, i got this...... http://thedigisync.com/digi-sync-top-page.html ..... last year, after having inconsistent results with carbtune carb sticks, and find it really easy to use, with good results.

I have AF-Xieds on my bike (2011 Twin cam GSA) and the bike has never run better. I don't know about extra horse power, torque or top speed but it's much smoother through all the gears, pulls like a train, and much easier to ride in traffic.

I've got the original exhaust can and headers on it and i re-fitted the exhaust flap valve after running a couple of years without it. :thumb2
 

Ok had a look at the link. looks like he is using the potential divider to reduce the voltage so that the ecu thinks it is running lean, so consequently the ecu enriches the mixture. But without any control. Also there are no measured AFR figures.
This approach is ok for the likes of us with expensive diagnostic kit who like to experiment, but most riders want a reliable plug and play solution.
Here is a link to the AF-XIED devolpment .
https://advrider.com/f/threads/2004-r1150rt-wideband-o2-sensor-project-and-af-xied-for-bmw.749080/
 
with the paid version of the motoscan app you can reset the idle speed stepper motors.... First do the valve clearances, new air filter and spark plugs, reset stepper motors give it a blast come back and set up the throttle bodies on a hot engine at 2000 rpm.... with a small fan blowing on the headers / and oil cooler.... redo steeper motors, give a spin redo throttle bodies...

Hey @botus ,

Is there anywhere that explains in more detail how to reset the stepper motor with motoscan? I'm replacing my idle air hoses, and apparently that could require that reset, but I don't know the procedure at all ...

Any help welcome,

Yannick

Sent from my AC2003 using Tapatalk
 
An old thread that has come to life again.

Allow me to comment on the O2 sensor manipulation in general and the AF-XIED .

The O2 sensor is a narrow band sensor type. If AFR (Air to Fuel Ratio) is richer than 14,7 the voltage goes high, and if AFR leaner than 14,7, voltage goes low.

The ECU use information from several sensor to send an estimated amount of fuel for any condition, and the. AFTER the combustion is done, the O2 sensor reports back. If the result was lean, the ECU ads a bit of fuel for the next stroke, if too rich, the next stroke goes leaner. So the ECU uses the calculations primarily, and then fine tune the fueling based on feedback from the O2 sensor. I.e, the O2 sensor sees two values from the O2 sensor, high or low. Hence, any attempt to mess with the voltage by means of resistors or diodes or what ever analogue way will not work.

Now, if you look closer on the signal from the O2 sensor
24852912707_711974e867_z.jpg

you will notice that even if the ECU relates to high low, the curve for the voltage vs AFR is not 100% rectangular, even if the most accurate reading with regard to voltage vs sensor temperature is in the center of the curve (where AFR is 14,7).

BMWs main target is to comply with emission control, hence, they stick to the most accurate data and disregard the rest.

But if you take a closer look at the curve, for voltage up to 0,8 is fairly stable for various temperatures, while increasing voltage (and richer AFR) from there on is less accurate.
The O2 sensor is temperature controlled by the ECU, so actually, this means that readings up to 0,9 V may vary between AFR 12,5 to 14,0, but the high low values are defined at a very wide range of temperature, so for all practical purpose, the green curve suggesting AFR 13,2 at voltage 0,9 is accurate enough for all practical purpse.

And here the AF-XIED comes in.
The AF-XIED as far more than an overpriced resistor. It'a a micro-computer. And what it does is that it intercepts the O2 signal, and evaluates the voltage from the O2 sensor. The original O2 signal stops here.
The AF-XIED aims for a stable voltage readout from the O2 sensor, and then it send High or Low signals to the O2 input for the ECU in order to force the ECU to adjust the fueling to get the proper voltage (and AFR) from the O2 sensor.

Adjusting the different levels is merely selecting different voltage levels for the AF-XIED to aim for. The adjuster is a variable resistor, but it does not set the voltage directly, it merely selects different preprogrammed levels for the AF-XIED in such a manner that the lowest level offers no correction, and the highest level will provide an AFR of 13,5, give or take.

There is much more to be said, but Ill stop here. The point being that the AF-XIED does it's adjustments in an intelligent and controlled manner. it doesn't just 'Add a bit more'. And all ECU functions remains intact, i.e no fault codes or limp mode from the ECU.
 
Where are you able to purchase these from as the last time I looked bemer boneyard no longer posts to the UK.
 
Where are you able to purchase these from as the last time I looked bemer boneyard no longer posts to the UK.
Hey,

You're right they don't. Luckily a family member of mine is visiting from mainland Europe soon so I sent it to them. There might be services that do that too?

Yannick

Sent from my AC2003 using Tapatalk
 
Hey @botus ,

Is there anywhere that explains in more detail how to reset the stepper motor with motoscan? I'm replacing my idle air hoses, and apparently that could require that reset, but I don't know the procedure at all ...

Any help welcome,

Yannick

Sent from my AC2003 using Tapatalk

dealer tools, motoscan paid version or reset vehicle adaptions on the last version for GS911 wifi and it resets stepper motors as it completes

on motoscan its under engine or service, you just tick the box and the tool makes them clunk about, running to max min movement and then settle where it thinks it should - takes 30 seconds engine off
 
re AF-XIED, as I understand it they can only richen the mixture (they cant lean anything out) - which on my hexhead would not be a sensible solution.

after an adaption reset I found my bike would work correctly for around 200 miles then always adapt to running like a dog. Eventually I realised this was a 100% repeatable exercise, and so disconnected the CAT sensors at 150 miles for almost useable map that never deteriorates. But I still have naff 1200 to 3000k rpm behaviour at less then 15% throttle application (you can live with it but I don't wish too), and a significant dead spot at 4400 to 5200 rpm on full throttle.

After proving the stop adapting method works long term, it occurred to me I could then reintroduce the PC3 to try and work around those two smallish foibles still present.. and indeed this works longer term too...

But the fix is not just enrichment... the hole at 4400 to 4900 is where is running desperately rich on the stock map. And its there with the bike std, or std and CAT sensors off at 150 miles.... the fault is just there all the time. Its just other bugs that get worse and worse if you leave it the way BMW wants

Over and above the pictures I shared of my PC3 setup in the VX800 post, I'm now running much leaner. I run up to -15 widgets of whateverness on the PC3 all the way from 3500 to 5500... the amount varies but 4800 is the full -15 widgets... so I don't get on my bike how AF-XIED, will do anything other than make it even flatter. If I add the sort of fuel they bring around the 4400 to 4900 mark its easily making 20bph less
 
My two 2008 GS conundrum

I read this thread with interest. Apologies for hijack and please delete, ignore, move etc as you see fit.

I have had my 2008 R1200GS since 2017, let's call this Bike A. Only performance mod I am aware of is Akra end can. Have done 3 European tours two up and weekend stuff.

Bike A has always pulled well, I have never noticed any flat spots in the rev range, no vibes I have ever really noticed.

I have recently acquired Bike B, another 2008 R1200GS, pretty much identical bikes when new. Bike B has full Akra system (decat headers :nenau) and is Hilltopped, and generally a cleaner bike with a little less mileage (30K miles), fresher suspension and some extras like LED indicators, nice splashguards, winglets. When I first rode this bike it felt different, engine a bit smoother.... but I don't know about BETTER.

The plan was to swap some of my farkles over to Bike B, keep it and sell my trusty, well liked, been on tour and has the stickers Bike A.

BUT :rob Now here's the thing, when out for a spin on Bike B at the weekend I noticed some vibes between say 70 - 80 MPH in top gear, right about where we would be cruising when on tour. And then I read this thread and flat spots and bikes not running well and I'm starting to get cold feet about it. Then I thought about moving full Akra from Bike B to Bike A but then I wonder could I introduce vibes to Bike A. That old adage "if it ain't broke" comes to mind. First world problems eh? What would you do?
 
I read this thread with interest. Apologies for hijack and please delete, ignore, move etc as you see fit.

I have had my 2008 R1200GS since 2017, let's call this Bike A. Only performance mod I am aware of is Akra end can. Have done 3 European tours two up and weekend stuff.

Bike A has always pulled well, I have never noticed any flat spots in the rev range, no vibes I have ever really noticed.

I have recently acquired Bike B, another 2008 R1200GS, pretty much identical bikes when new. Bike B has full Akra system (decat headers :nenau) and is Hilltopped, and generally a cleaner bike with a little less mileage (30K miles), fresher suspension and some extras like LED indicators, nice splashguards, winglets. When I first rode this bike it felt different, engine a bit smoother.... but I don't know about BETTER.

The plan was to swap some of my farkles over to Bike B, keep it and sell my trusty, well liked, been on tour and has the stickers Bike A.

BUT :rob Now here's the thing, when out for a spin on Bike B at the weekend I noticed some vibes between say 70 - 80 MPH in top gear, right about where we would be cruising when on tour. And then I read this thread and flat spots and bikes not running well and I'm starting to get cold feet about it. Then I thought about moving full Akra from Bike B to Bike A but then I wonder could I introduce vibes to Bike A. That old adage "if it ain't broke" comes to mind. First world problems eh? What would you do?
I really like your post. To be honest I haven't had experience with any of those farkles yet, so my response might not be very useful.

But, since I'm interested, my first thought would be, as you said to put the akra on your old bike and see if that's the problem. If it is then I guess your problem is solved.

Otherwise, maybe some hilltoppers could confirm if they had the same issue. Maybe you can ask hilltop to remove the remap? Maybe that's free? Not sure.

Even though it might feel like you bought another bike for nothing, if you prefer your old one you could at least get more money for your new one if you sell it... I'd be curious what others think :)

Sent from my AC2003 using Tapatalk
 
Actually, I just remember hearing that adding a full exhaust system would need a remap or booster to make the bike run correctly, as otherwise if would run too lean. Do you know wether the previous owner has done that?

Sent from my AC2003 using Tapatalk
 
BUT :rob Now here's the thing, when out for a spin on Bike B at the weekend I noticed some vibes between say 70 - 80 MPH in top gear, right about where we would be cruising when on tour. And then I read this thread and flat spots and bikes not running well and I'm starting to get cold feet about it. Then I thought about moving full Akra from Bike B to Bike A but then I wonder could I introduce vibes to Bike A. That old adage "if it ain't broke" comes to mind. First world problems eh? What would you do?

my bike a hex head (not the later 2009 to 2012 twin cam - which is a different later engine running different software again) has always had a rough spot at 4500 to 4800 rpm. Its super nasty... altering the fuelling and running super unleaded 99 makes it less bad. I suspect its an inherent NVH issue with the way they made them. Its no worse 26k miles on, just nasty. If I go with a group of others I say go under 75 or over 85mph. Also worth doing up the engine bolts... mine were loose once until I did them up properly (clearly BMW don't know how).

what was different was the old original engine map, a basket case but tweaked with a PC3 was good nearly everywhere. Then 10 years in to its life I had a software update on the whole bike (bringing a new engine map) and it was shit. But wiping the old adaptions and rebuilding new ones helps, so long as you stop them from developing all wrong... and thus the PC3 is back and its OK.

But as per that long vx800 thread its NOT as good as the original map and as its just gone cold its running like a dog again.

I asked but no one said... On my Merc car I reset its adaptions and it has two sets. Cold weather and normal adaptions. I think the bike is like this too... but I suspect its undocumented and clearing your "vehicles adaptions" clears both you don't get a choice.... So of course would need to allow adaptions to rebuild in both hot and cold weather
 
re AF-XIED, as I understand it they can only richen the mixture (they cant lean anything out) - which on my hexhead would not be a sensible solution.

Actually, not quite so.

The AF_XIED (AF from now on) relates to one thing only, the voltage from the O2 sensor.

This means that any time the fuel is richer than selected setting (Richer means higher voltage emitted by the O2 sensor), the AF sends HIGH signal to the ECU (i.e the ECU thinks it gets a feedback from the O2 sensor that the fueling is too rich) , meaning that the ECU goes leaner. And the AF will keep sending the HIGH signal until the ECU has leaned enough to where the O2 sensor generates a voltage lower than the threshold.
This is the wonder of the AF, it base it's commands on a measured signal (from the O2 sensor), it doesn't just add a bit more and disregard the result.

Take a look at this curve, it's a scope readout from a 1200LC with a AF installed, blue curve reads the voltage from the O2 sensor and red reads the voltage generated by the AF and is transmitted to the ECU O2 signal input.
41625718154_f33b00b8e8_b.jpg



Look at every time the red curve peaks, and see what is then happening to the blue curve (voltage from the O2 sensor).

The fueling (and hence the AFR) keeps swinging between richer than and leaner then the reference voltage (or reference AFR if you like). As mentioned before, the ECU only cares about two levels from the O2 sensor, that is leaner or richer than the average (the value is 0,45 V, give or take a few mV, where 0,45V represent AFR 14,7, which is the target for emission control)).

Back to the graph: When the red curve is at it's lowest (the ECU reacts to the signal passing 0,45V either direction), the blue curve also changes direction. When red curve (signal to ECU) drops below 0,45, the ECU decides that the current fueling is too lean, and it starts adding a bit more, and you can see that the blue curve is again rising, meaning that the fueling gets richer.
After the fuel gets even richer, the blue curve keeps rising (voltage from O2 sensor), and it keeps rising until the voltage passes the defined 'high voltage' as defined by the AF, based on the setting selected by the user. Once the high blue line (high voltage) has been reached, the AF sends a too-rich feedback, i.e the (red curve)voltage rise to above 0,45V. And look at what happens when at the top peak of the red curve. By now the ECU has received a feedback that the fuel is too rich (as it reads the voltage from the AF (red curve), and once again, it starts to lean (blue line dropping).

The border for rich/lean (meaning blue curve top/bottom voltage) level is defined by the selected setting in the AF.

Here is an example where the AF is set to position 8. The reading was done a couple of days later, so the probes where switched.
So this time, the red curve reads the O2 signal and the blue line reads the AF output.
48197739222_042b42ed27_b.jpg


At this setting, the AF has defined a closer value between high/low level, and we can see the curve is more level.
A close to steady 0,9V (from O2 sensor, red curve in this chart) means an AFR 13,5 +/-,( ref. the chart published in my post a few posts back.)

So, to sum it up: The AF takes a positive control of the fueling, based on feedback from the O2 sensor. And it does so by maniplating the O2 signals to the ECU. Where the OEM mapping would make the engine run too lean, the fueling will become richer, and where the OEM fueling is too rich, it will be leaner. (The reference for too rich/lean is here defined by the user setting of the AF)
 
nice - thanks for the explanation knutk

so all that fine tuning would end up in the vehicle adaptions area, meaning in the moment the corrections the AF_XIED was doing at first, lessens over time?

any thoughts on cold condition adaptions?
 
The Adaptions do not lessen over time as the long term fuel trims learn from the short term trims that the AF -XIED`s have manipulated.
That is why it takes a few tanks of fuel to adapt the long term fuel trims.
The long term fuel trims also influence open loop.
Cold condition adaptations. When the bike is cold / first started it is in open loop so the lambda sensors are not reporting but it does not take many minutes until it goes into closed loop.
It appears that the later bikes go into closed loop quicker than earlier bikes.
 
...............
so all that fine tuning would end up in the vehicle adaptions area, meaning in the moment the corrections the AF_XIED was doing at first, lessens over time?

any thoughts on cold condition adaptions?

No, the corrections will be stored in the adaptive table and be added to the equation even when in open loop conditions.
And the adaption table is dynamic. That is, as @Mistcat suggest, it keeps changing for the life of the ECU. And the motivation for doing so is easy to understand.

Let's say we are riding on a dusty road, where the air filter is slowly clogging up. If the fueling was set once for all, the fueling would then slowly become richer. But with the dynamic adaption, the ECU will continuously adapt to the changes and the emission standard will be upheld.
So, the old adaptions will be overwritten little by little.
 
the bit that concerns me is how quick can it get it right...

bog standard bike - if I snap open the throttle it gives up at 4700 rpm for around half an hour...

if I get the PC3 right its correct instantly and revs through like a grown up... allowing me to right a bike that's doing as its told...

if its dynamic does that mean it talks a while (aka a split second) to find its feet every time the AF_XIED need to fix the hideous mess BWM wrote in the map?
 
I have never seen a specification for how long it takes for the changes to take place.

But updating the adaptive map is a continuous never ending process. The aim for the ecu is to set the values in the adaptive map to such values that there will be no need for corrections from the o2 sensor. This means that at any point where the adaptive map contains one value, and the O2 sensor still calls for a correction, towards lean or rich, the stored value in the adaptive map will be adjusted in the direction dictated by the O2 sensor. In order to prevent the table from 'chasing values' it seems natural that the ECU is programmed to perform the changes in tiny steps.

The only time I have been able to notice a change in values taking place was when my Euro 3 -15 1200 GSA got the ECU updated to Euro 4.

That time, when i rode from the shop, the engine would run less than perfect. Within an hour and 90km of riding, the less than perfect running of the engine was no longer an issue.
 
the bit that concerns me is how quick can it get it right...

bog standard bike - if I snap open the throttle it gives up at 4700 rpm for around half an hour...

if I get the PC3 right its correct instantly and revs through like a grown up... allowing me to right a bike that's doing as its told...

if its dynamic does that mean it talks a while (aka a split second) to find its feet every time the AF_XIED need to fix the hideous mess BWM wrote in the map?

What gear are you in and at what speed are you doing when you snap the throttle open,
The reason I ask this is when I snapped the throttle open to full throttle 6th gear at 50 mph the bike would stumble, after logging this and looking at the AFR graph it initially went rich as expected but then went lean causing the hole before slowly going rich again.
First thought was the knock sensors, but the data said not.
I took the graphs along to a so called "expert" at a rolling road tuning establishment. He did not know and said it was probably due to the sensors being a bit slow, REALY!!!.
The only conclusion that I arrived at was that the Load calculated by the ecu RPM / throttle position/selected gear etc was causing the hole to prevent excess load on the transmission.
When an appropriate gear is selected before giving it full throttle I have not had a problem the bike revs to the red line cleanly in each gear.
With the PC 3 you are probably dumping enough fuel in to cover the hole then the ecu is learning around it if you have the lambda sensors in circuit.
 
As to being Dynamic I understand that every time the bike fires the result is reported back by the lambda sensor. The information is looked at regarding throttle position, selected gear, rpm, engine load etc and the injector pulse width adjusted next time that it fires.
The data is logged to the short term fuel trims that are applied to the relevant map when these readings become more constant they are logged by the long term fuel trims which are logged to the relevant map.
This is a very simplistic view the way that I understand it, so yes very dynamic.
 


Back
Top Bottom