Hotrodding the GS

<BR>

<table border="0" cellpadding="5" width="650"><tr><td>I hate when my thinking catches a snag.&#8194;

It sounds like a compelling argument. Really, it does. I get caught there myself. But the issue is more a case of 'when a person has a hammer then all problems look like nails', than anything else.


Change the code vs. change the inputs looks like this:

Any changes to the code that affect 'leading throttle' operation must involve changing the fuel/air ratio. That ratio is based on mass. A change to the code operation could be described like this, "Inject more mass of fuel for a given mass of air."

Take a look at it from the other way around though. You can see that if the mass of the air is changed, it accomplishes the same thing. And temperature modification changes the mass of the air.

The same method of 'look at it from the other way around' applies to 'transient enrichment' operation.

It's the classic, raise the bridge vs. lower the water question.
</td></tr></table>
<BR>

I've a question, some of what I've read here causes me to think that the prom doesn't have all the software. That there are instructions in the ecu that can't be directly changed by changes to the prom.

That MAY be true. IF that is true, it is still not a problem. See below...

You are VERY close. You have misstated my position slightly...

I am not in favor of recompiling or rewriting or modifying the ECU software at all. It's more or less fine.

Look back at this post:

Why reinvent the wheel when all you need to do are make some tweaks?

Also, I am not selling a competing product. Maybe someday, but not now.

You are right, of course. But I can make my changes a number of layers of abstraction lower than you. And I can be certain that my changes will not effect anything I don't want them to.

What if spark angle gets advanced with a low oil temp? (It almost certainly does.) Now that you're making the ECU think it's nice and chilly when it's actually up to temp, maybe you get too much spark advance along with your rich mixture...

From my understanding by and large you got it right.

But another aspect that I see as important is the ease of installation/implementation. For a large number of folks, as soon as you mention having to open up the ECU and replace chips, their eyes glaze over and you realize they just aren't going to go there.

But, adding an external 'black box', is no where as daunting a challenge to those very same folks.

And many forget that the Motronic ECU is 25+ years old now. Some of the technology used is no longer in production. That doesn't mean there isn't a work around, but it might significantly increase the complexity of the design and implementation.

Just some food for thought.

JJ

<BR>


<BR>

will your devices work for the R1200S model?

<BR>

The model for the air temperature works on all EFI models. The next device is for motor temp on the 1100/1150. The third device, for motor temp on the 1200 hexhead and camhead, is coming later.


<BR>

I'll add my fifty cents here coming from my experience in the world of Volkswagen TDI engine tuning. Way back when, when the TDIs first came out people did add-ons - mainly because the amount of people who could do chip tuning were very few and these add-ons did had positive results. These add-ons basically fooled the ECU by going between the ECU and said device (Air Mass Sensor for example - similar to the intake air temp sensor mod for BMWs) to modify the signal to change the fueling rate / injection timing. These devices do work, but each and every time chip tuning (when done correctly!) does a better job. Mixing chip tuning and ad-ons usually spells disaster as it usually maxes out a particular parameter and thing go for a coffee and bacon induced poo. :eek1

With TDI chip tuning there is both modifying the ECUs "operating system" to remove things like immobilizers, EGR etc and modifying tables of data: inputs from sensors and outputs to devices. By saying that, I am aware of no software mods that execute the data from the tables any better.

So in the end, both approaches of fooling the ECU and modifying the software tables will result in improved performance.

Needless to say, I am involved in chip tuning and making improvements to things like injectors and breathing not sensor fooling. That being said, I don't do Motronic tuning so... I want my bike to run better so I'll take what I can get!

<table border="0" cellpadding="5" width="650"><tr><td>Do you guys have the ECU 'source code' instructions for the Volkswagen? And are changes being made to the source instructions, or only to the tables?

On the TDI, since there's a turbocharger, there's a MAP sensor available to manipulate. Do you know if that's what the sensor add-on components were doing? That brings a new level to sensor tuning "done correctly".

If either is done correctly, it's like comparing apples and apples.
</td></tr></table>
<BR>

Where is the motronic getting air density? Is it a builtin sensor in the ECU assembly?

<BR>

<table border="0" cellpadding="5" width="650"><tr><td>The Motronic uses a built in barometric sensor for air pressure, and the intake air sensor for air temperature. Those two variables provide the air density.
</td></tr></table>
<BR>

The source code is really a Bosch thing - even for Volkswagen and it is based on an Intel processor designed for cars. So if one spoke assembly language and knew what each of the registers does then one could easily follow the program. If you only know how C++ or Visual Basic or Java works then you have no clue how assembly language works! Each command in these higher level languages is dozens of commands at the CPU level - where you do thing literally one step at a time. The high level languages take your "human" code and turn it into machine code. There really is no simple or even difficult way to go from machine code to human code - it's more like über complex!

Really smart people have written programs to reprogram TDIs and Bosch Motronic (1.8T) where you can visually look at the tables and with a click of the mouse change things... this is after 15 years of TDI and there are a lot more TDIs out there than BMWs. I think I am going to crack open my Motronic 2.2 ECU on my 1100 and see what I can see with one of these programs... could be quite simple!

Also, like I was saying there is the ECU's source code that can also be manipulated to turn off things like the immobilizer or EGR system or if you want to use a 3.0 bar MAP instead of a 2.5 bar MAP that is all part of the "source" for the ECU.

All the sensor add-ons would do is give the ECU a more favorable signal that works better with the original software. So just the like air intake sensor mod for the BMW where the ECU believes that it is cooler than it really is so there for it delivers a different (richer) mixer that makes more power. Once you put chip tuning on there you are in effect doubling those efforts and sometimes (most of the time!) that is too much.

I have a feeling that if someone was to crack the immobilizer routines for the R1200 ECUs there would be some happy customers!

CAN bus programming is also something to look at. 35W HIDs on a 55W circuit makes an error... but... if you change what CAN is looking for (like don't care how many watts) then there's no problem.

<BR>

<table border="0" cellpadding="5" width="650"><tr><td>Lots of things are possible if you get 'inside' the controllers and change simple things like flag bits, or a byte that defines the amperage threshold for the lights. About the 35W vs. 55W lighting; Current sensing control is inside the ZFE box, and isn't related to CANbus.

If you wanted to do things like that, if it were me I would put a serial data logger onto the OBD port, and bring the bike in for service. Log the communication between the BMW Diagnostic Console, and the ECU and other controllers on the bike.



p.s.: I am really fond of those TDI Volkswagens!
</td></tr></table>
<BR>

Thanks - thought I had read something about the sensor being built in but couldn't find it again.

I smell a nerdgasm, and it smells good.

A couple years ago, I had a chance to talk to Steve Dinan about how his company had decoded the BMW car software. That's how they captured the data but they have their own GT1s so there was less subterfuge.

Hey guys, my O2 sensor is disconnected. Will your devices still work on my bike (02 GS)?

<BR>

Using the BMW GT1 diagnostic console, with a serial data logger, makes the most sense.

<BR>

<BR>

Yes.

<BR>

How about the discarded cat code plug crowd? Not to mention the Steptoe mod folks...