So I'm tuning the motronic...
03-24-2015, 08:36 PM
#1
Three Wheelin'
Thread Starter
So I'm tuning the motronic...
Jumped down the rabbit hole with this one. It all started with chasing a rich AFR at WOT. Dyno shop and my LC-2 wideband agree so I trust the result. The rich AFR is present in the stock chip or using Wong's 93 octane chip. Though Steve's chip runs even richer than the stock chip. Engine is healthy, fuel system pressures and all of the sensors check out, so I can only conclude that the rich mix is built into the chip.
To start with, I'm running an Innovate LC-2 wideband o2 along with their SSI-4 to log rpm. Additionally, I have a Moates Ostrich 2.0 eprom emulator which takes place of the chip and contains the binary code. I use Tunerpro RT to interface with the Ostrich.
There are 103 maps loaded into the code. Though many of them are duplicates (part-throttle fuel map for US, RoW, Tiptronic, etc. are the same). There are additional values for the rev limiter, decel activation, IAC trim, etc. At the moment I only have the US WOT fuel map address in my "this is absolutely accurate" map address list. This map kicks in when the WOT microswitch is activated or when the AFM voltage maxes out. At that point, the amount of gas injected into the ports is entirely dependent on your fuel pressure, injector flow, and the values in this map.
Here is what the US WOT fuel map looks like. My new and improved fuel curve vs the stock fuel curve. My changes is the trace with circles:
Notice how coarse the available adjustment bins are?
Here is my tune (solid lines) vs the Wong chip (dotted lines). Much improved!
That's a start. The emulator will trace active cells in the maps when the motronic uses them which at least provides some clues. I've got initial addresses for part throttle and idle maps for fuel and ignition. Just need to sort out which the region/tiptronic versions. I'd like to find all of the cold-start maps and corrections next so that I can get a better starting idle.
I'll be posting up a tunerpro definition file containing all of the maps, but without descriptions of what they are. I'll add descriptions when I'm certain of a map. Each map has a start address, a list of static axis values, and the "tuning" values. For the axes, the rpm actual values are fairly easy to get. I've got the conversion to degrees advance for the ignition maps. However, I haven't gotten to figuring out load, temperature, etc. conversions from the binary data. The only way to really know what everything is would be to disassemble the code, but that's not on my radar at this point.
Hope that this can be the start of a helpful thread on tuning and decoding the motronic!
To start with, I'm running an Innovate LC-2 wideband o2 along with their SSI-4 to log rpm. Additionally, I have a Moates Ostrich 2.0 eprom emulator which takes place of the chip and contains the binary code. I use Tunerpro RT to interface with the Ostrich.
There are 103 maps loaded into the code. Though many of them are duplicates (part-throttle fuel map for US, RoW, Tiptronic, etc. are the same). There are additional values for the rev limiter, decel activation, IAC trim, etc. At the moment I only have the US WOT fuel map address in my "this is absolutely accurate" map address list. This map kicks in when the WOT microswitch is activated or when the AFM voltage maxes out. At that point, the amount of gas injected into the ports is entirely dependent on your fuel pressure, injector flow, and the values in this map.
Here is what the US WOT fuel map looks like. My new and improved fuel curve vs the stock fuel curve. My changes is the trace with circles:
Notice how coarse the available adjustment bins are?
Here is my tune (solid lines) vs the Wong chip (dotted lines). Much improved!
That's a start. The emulator will trace active cells in the maps when the motronic uses them which at least provides some clues. I've got initial addresses for part throttle and idle maps for fuel and ignition. Just need to sort out which the region/tiptronic versions. I'd like to find all of the cold-start maps and corrections next so that I can get a better starting idle.
I'll be posting up a tunerpro definition file containing all of the maps, but without descriptions of what they are. I'll add descriptions when I'm certain of a map. Each map has a start address, a list of static axis values, and the "tuning" values. For the axes, the rpm actual values are fairly easy to get. I've got the conversion to degrees advance for the ignition maps. However, I haven't gotten to figuring out load, temperature, etc. conversions from the binary data. The only way to really know what everything is would be to disassemble the code, but that's not on my radar at this point.
Hope that this can be the start of a helpful thread on tuning and decoding the motronic!
03-25-2015, 12:35 AM
#2
Race Car
Wow nick. Nice work. Looking forward to seeing the results...gimmie a shout when you are going to head over to the dyno. Maybe we can get a couple guys together and share the cost and compare notes...
03-25-2015, 11:54 AM
#3
Nice work! I'm looking forward to reading more about this project. It's not surprising to me that you're seeing a very rich mixture at WOT, since it's likely running open loop? They're probably keeping a safety margin in case of poor fuel. What is the end goal of this activity?
03-25-2015, 01:11 PM
#4
Three Wheelin'
Thread Starter
Goughary- I'll definitely be up for a dyno day! Spring is coming...
HalV- Yes, open loop when the WOT switch is closed, or when the AFM flap maxes out. I was hoping that I might feel a little extra oomph, but not much of a butt dyno difference.
End goal? Well, the motronic has the ability to be just as flexible as a stand-alone efi. There is lots of space on the chip to code in instructions, assuming that we can figure out how to code it. Wouldn't it be nice if ITB's just required the throttle bodies and a chip instead of a whole Motec install? Or how about an easy MAF swap? A supercharger kit without a hokey piggy-back injector system? Bigger injectors for high hp motors? Or just modern injectors with better atomization?
I am finding out some interesting things about how the motronic works. For instance, it is *always* closed loop, bouncing between about 14.2-15.2 AFR unless: the WOT switch is closed or the AFM is maxed (WOT mode), the o2 sensor is warming up just after starting (10-60seconds), or the engine is still warming up (cht sensor reporting temp is below ~70C, I wrote down the exact temperature somewhere...), it then uses the fuel maps with no o2 feedback fuel trim correction. A part throttle acceleration up to 6500 rpm is going to be at a constant ~14.7:1 AFR. Probably not ideal. If you check out my fuel curve, I also had to add quite a bit of fuel at lower rpm to get the AFR down. So it could be tuned to run more efficiently pretty easily. The wildly rich WOT is definitely Porsche's way of erring to caution- edit, or mine has a problem somewhere that I haven't found.
HalV- Yes, open loop when the WOT switch is closed, or when the AFM flap maxes out. I was hoping that I might feel a little extra oomph, but not much of a butt dyno difference.
End goal? Well, the motronic has the ability to be just as flexible as a stand-alone efi. There is lots of space on the chip to code in instructions, assuming that we can figure out how to code it. Wouldn't it be nice if ITB's just required the throttle bodies and a chip instead of a whole Motec install? Or how about an easy MAF swap? A supercharger kit without a hokey piggy-back injector system? Bigger injectors for high hp motors? Or just modern injectors with better atomization?
I am finding out some interesting things about how the motronic works. For instance, it is *always* closed loop, bouncing between about 14.2-15.2 AFR unless: the WOT switch is closed or the AFM is maxed (WOT mode), the o2 sensor is warming up just after starting (10-60seconds), or the engine is still warming up (cht sensor reporting temp is below ~70C, I wrote down the exact temperature somewhere...), it then uses the fuel maps with no o2 feedback fuel trim correction. A part throttle acceleration up to 6500 rpm is going to be at a constant ~14.7:1 AFR. Probably not ideal. If you check out my fuel curve, I also had to add quite a bit of fuel at lower rpm to get the AFR down. So it could be tuned to run more efficiently pretty easily. The wildly rich WOT is definitely Porsche's way of erring to caution- edit, or mine has a problem somewhere that I haven't found.
Last edited by -nick; 07-30-2017 at 07:39 AM.
03-25-2015, 01:15 PM
#5
Three Wheelin'
Thread Starter
If anyone wants to join in, here is what you need:
1. Wideband oxygen sensor (wbo2)
Innovate LC-2 kit.
a. Install wbo2 sensor into cat in place of the stock narrow band sensor.
b. Run wiring from sensor to LC-2 module through left side of engine firewall wiring grommet.
c. Connect LC-2 to switched +12v and ground.
d. Route LC-2 brown wire, analog out #2, into engine bay.
e. Connect analog out #1 into stock o2 sensor pin 2 (the pin separated from the other two, see pic).
f. Connect laptop to LC-2 via usb. Calibrate sensor per manual.
g. Open LM Programmer. Default analog out #2 settings are what you want to emulate the stock 0.1-1.1v sensor. Advanced > warmup settings should be 0v.
2. RPM logging
Innovate SSI-4 module.
a. Connect switched +12v and ground to module.
b. Connect ch1+ to pin 6 on the 55-pin motronic connector, or pin 14 on the OBD1 plug (center row, far left pin) in passenger footwell.
c. Connect SSI-4 to LC-2 (Requires additional cable: Innovate 3812).
d. Connect laptop to usb, accept all defaults.
*The Innovate software is “buggy” to say the least. It seems to always if you follow this process:
Start LogWorks, click “ok” when it says there is no connection. Leave the next “connect” popup there.
Plug in USB
Ignition to “on”
Now click “connect” and start the car.
Every other way of connecting either freezes my workstation laptop, or windows 7 pops up the blue screen of death and reboots.
3. Moates Burn2 chip reader
https://www.moates.net/burn2-chip-pr....html?cPath=94
I'll post Porsche's last chip revision along with the other files when I get a chance. So you don't necessarily need a chip reader/burner. The last revision, #7006, works in all C2/C4/RS America/etc models. From what I hear, the Euro RS and Cup cars not only have different maps, but also different code in the chip. So they're on their own.
a. download tunerpro rt ( www.tunerpro.net RT is free with a timeout, or pay up if you continue to use it!)
b. put chip in burn2 ***notch is toward handle and extra rows are at handle***
d. plug usb into computer
d. windows auto-installs usb drivers
e. open tunerpro rt
f. tools > hardware utilities > moates i/o
g. "read chip", 27c256 chip type
h. "prom i/o succeeded"
i. save buffer to file, name the bin file
j. open file > open bin
4. Ostrich 2.0 eprom emulator
https://www.moates.net/ostrich-20-th....html?cPath=95
a. Download USB drivers from moates.com downloads section.
b. Plug ostrich into usb. Don’t connect to motronic yet.
c. Windows xp automatically installed drivers and set com port to COM8. Go into device manager and change this to COM3 or COM4. My laptop said that COM3-COM7 were “in use”. I selected COM3 and ignored warnings.
d. Set latency=1 instead of the default 16 in the driver port settings. This will speed up com dramatically.
e. Open tunerpro rt and open a bin file and xdf file.
f. Tools > Emulation > Upload current bin to emulator.
Now we’re ready to plug into the motronic.
g. Plug ribbon cable from ostrich into motronic chip socket. ***Red stripe on cable faces the chip notch***
h. Make a notch in the metal motronic cover for the ribbon cable to exit. The cover needs to(?) be firmly attached so that it firmly contacts the ground connection to the ecu board.
i. Plug motronic back into car and fire it up!
You’ll need an .xdf definitions file to open up the maps. The xdf is where all of the arduous decoding happens.
1. Wideband oxygen sensor (wbo2)
Innovate LC-2 kit.
a. Install wbo2 sensor into cat in place of the stock narrow band sensor.
b. Run wiring from sensor to LC-2 module through left side of engine firewall wiring grommet.
c. Connect LC-2 to switched +12v and ground.
d. Route LC-2 brown wire, analog out #2, into engine bay.
e. Connect analog out #1 into stock o2 sensor pin 2 (the pin separated from the other two, see pic).
f. Connect laptop to LC-2 via usb. Calibrate sensor per manual.
g. Open LM Programmer. Default analog out #2 settings are what you want to emulate the stock 0.1-1.1v sensor. Advanced > warmup settings should be 0v.
2. RPM logging
Innovate SSI-4 module.
a. Connect switched +12v and ground to module.
b. Connect ch1+ to pin 6 on the 55-pin motronic connector, or pin 14 on the OBD1 plug (center row, far left pin) in passenger footwell.
c. Connect SSI-4 to LC-2 (Requires additional cable: Innovate 3812).
d. Connect laptop to usb, accept all defaults.
*The Innovate software is “buggy” to say the least. It seems to always if you follow this process:
Start LogWorks, click “ok” when it says there is no connection. Leave the next “connect” popup there.
Plug in USB
Ignition to “on”
Now click “connect” and start the car.
Every other way of connecting either freezes my workstation laptop, or windows 7 pops up the blue screen of death and reboots.
3. Moates Burn2 chip reader
https://www.moates.net/burn2-chip-pr....html?cPath=94
I'll post Porsche's last chip revision along with the other files when I get a chance. So you don't necessarily need a chip reader/burner. The last revision, #7006, works in all C2/C4/RS America/etc models. From what I hear, the Euro RS and Cup cars not only have different maps, but also different code in the chip. So they're on their own.
a. download tunerpro rt ( www.tunerpro.net RT is free with a timeout, or pay up if you continue to use it!)
b. put chip in burn2 ***notch is toward handle and extra rows are at handle***
d. plug usb into computer
d. windows auto-installs usb drivers
e. open tunerpro rt
f. tools > hardware utilities > moates i/o
g. "read chip", 27c256 chip type
h. "prom i/o succeeded"
i. save buffer to file, name the bin file
j. open file > open bin
4. Ostrich 2.0 eprom emulator
https://www.moates.net/ostrich-20-th....html?cPath=95
a. Download USB drivers from moates.com downloads section.
b. Plug ostrich into usb. Don’t connect to motronic yet.
c. Windows xp automatically installed drivers and set com port to COM8. Go into device manager and change this to COM3 or COM4. My laptop said that COM3-COM7 were “in use”. I selected COM3 and ignored warnings.
d. Set latency=1 instead of the default 16 in the driver port settings. This will speed up com dramatically.
e. Open tunerpro rt and open a bin file and xdf file.
f. Tools > Emulation > Upload current bin to emulator.
Now we’re ready to plug into the motronic.
g. Plug ribbon cable from ostrich into motronic chip socket. ***Red stripe on cable faces the chip notch***
h. Make a notch in the metal motronic cover for the ribbon cable to exit. The cover needs to(?) be firmly attached so that it firmly contacts the ground connection to the ecu board.
i. Plug motronic back into car and fire it up!
You’ll need an .xdf definitions file to open up the maps. The xdf is where all of the arduous decoding happens.
03-25-2015, 03:10 PM
#6
Race Car
So beyond my expertise it's ridiculous...I'm loving the thread though..
Will definitely be up for a dyno day, never having had my car on a dyno, would be good to see the numbers. I'm sure it'll lead me to look for a solution to "something"...lol
Maybe late April? Early May?
Will definitely be up for a dyno day, never having had my car on a dyno, would be good to see the numbers. I'm sure it'll lead me to look for a solution to "something"...lol
Maybe late April? Early May?
Trending Topics
03-25-2015, 03:33 PM
#8
Rennlist Member
03-25-2015, 03:44 PM
#9
Race Car
When you say its running rich, how rich, what sort of AFR's is your LC-2 seeing?
Having spent a bunch of time understanding the Motronic (not as deep as you) And having installed a ViPec in my 993 i am interested in what you are working on.
One of the biggest challenges is finding the address for the injector dead times, this is where many seem to stumble. If you end up with bigger injectors the problem is that there is not enough downside adjustment in the tables to compensate for the extra fuel that can be delivered....
Enjoy the rabbit hold, and bring spare batteries for your flashlight
Having spent a bunch of time understanding the Motronic (not as deep as you) And having installed a ViPec in my 993 i am interested in what you are working on.
One of the biggest challenges is finding the address for the injector dead times, this is where many seem to stumble. If you end up with bigger injectors the problem is that there is not enough downside adjustment in the tables to compensate for the extra fuel that can be delivered....
Enjoy the rabbit hold, and bring spare batteries for your flashlight
03-27-2015, 08:42 AM
#11
Three Wheelin'
Nick, ITB's do not require Motec, it can be done with pretty much any standalone ECU, including VEMS and it can be tuned to have excellent driveability with very reasonable cost.
Here's stock 993 non-VRAM engine with GSXR1000 ITB's, 220cpi cats and VEMS:
Here's stock 993 non-VRAM engine with GSXR1000 ITB's, 220cpi cats and VEMS:
01-10-2017, 04:09 AM
#12
Hello there,
I'm starting to look at the tuning possibilities on the standard motronic chips, in order to install a MAF with a larger AirFilter.
Nick, thank you very much for this introduction on what's what. It sure helps to understand !
I have a Hammer Tool for the OBD error codes and this guy tells me I have an error on DME, which as I read on the net, suggest I have a custom chip in there.
When I bought my car, there wasn't any O2 sensor and the cat had been replaced by a silencer without an O2 plug, so whatever chip is in there works on a constant open loop, which sounds like the previous guys wanted the raw maps to be used as is.
Recently, I replaced this silencer by a Sport Decat Tube WITH an O2 port on which I'll plug my Innovate LM2 WideBand O2 Sensor. Actually that was the first milestone I needed to achieve.
Not checked in the motronic yet to see what I've got, but before unfitting anything, I wanted to be sure I have all the tools to get started properly.
So far, I've installed TunerPro and loaded some 964 bin found here or there with the XDF file provided, but I'd be interested to check on your's as well.
Nick, would you also have some other bins to compare ?
I'd be interested in an OEM 964 (1990), just to check on differences with customized maps.
I'm starting to look at the tuning possibilities on the standard motronic chips, in order to install a MAF with a larger AirFilter.
Nick, thank you very much for this introduction on what's what. It sure helps to understand !
I have a Hammer Tool for the OBD error codes and this guy tells me I have an error on DME, which as I read on the net, suggest I have a custom chip in there.
When I bought my car, there wasn't any O2 sensor and the cat had been replaced by a silencer without an O2 plug, so whatever chip is in there works on a constant open loop, which sounds like the previous guys wanted the raw maps to be used as is.
Recently, I replaced this silencer by a Sport Decat Tube WITH an O2 port on which I'll plug my Innovate LM2 WideBand O2 Sensor. Actually that was the first milestone I needed to achieve.
Not checked in the motronic yet to see what I've got, but before unfitting anything, I wanted to be sure I have all the tools to get started properly.
So far, I've installed TunerPro and loaded some 964 bin found here or there with the XDF file provided, but I'd be interested to check on your's as well.
Nick, would you also have some other bins to compare ?
I'd be interested in an OEM 964 (1990), just to check on differences with customized maps.
01-10-2017, 04:13 AM
#13
Are you happy with the Moates Ostrich 2.0 eprom emulator ?
I was wondering if the APU1 AutoProm was more appropriate/necessary ?
http://www.moates.net/apu1-autoprom-....html?cPath=95
What's your opinion about that ?
I was wondering if the APU1 AutoProm was more appropriate/necessary ?
http://www.moates.net/apu1-autoprom-....html?cPath=95
What's your opinion about that ?
01-10-2017, 04:17 AM
#14
(adapters, links, etc...)
thanks !
01-10-2017, 11:18 AM
#15
Three Wheelin'
Thread Starter
Yikes! Almost 2-years and I still haven't finished soldering together my break-out box for this experiment.
Hi Manu,
The Moates gadgets are working well for me. I'm not familiar with the one that you linked to.
I would urge running some kind of a cat. I'm using the Magnaflow #59928 cat. It's about as cheap as a bypass, doesn't inhibit the flow, and still cleans up the emissions.
Send me a pm with your email and I'll send you the xdf file that I've been working on. It's still a little rough, but it's a big upgrade from the current one available via the tunerpro website.
Raceboy has a nice system put together. I believe that his is megasquirt based. From my megasquirt-2 days, getting factory drivability was not an easy task. Perhaps it's better with version 3.
Hi Manu,
The Moates gadgets are working well for me. I'm not familiar with the one that you linked to.
I would urge running some kind of a cat. I'm using the Magnaflow #59928 cat. It's about as cheap as a bypass, doesn't inhibit the flow, and still cleans up the emissions.
Send me a pm with your email and I'll send you the xdf file that I've been working on. It's still a little rough, but it's a big upgrade from the current one available via the tunerpro website.
Raceboy has a nice system put together. I believe that his is megasquirt based. From my megasquirt-2 days, getting factory drivability was not an easy task. Perhaps it's better with version 3.