9.5.2004 I managed to get a possiblity to have some track time at Ahvenisto
track (sorry, that site in finnish only). Here is picture of
the equipment all
and preparing for the
first track session of this season. Here is a photo of my favourite
current state of development. And here, Corsa on track and
getting ready to deal
the next corner.
is some track activity seen from the pit
lane (wmv-video, 347 k). I installed my small digital camera
(Minolta Dimage) to the side of the fairing and made some recordings
while driving. Here are some selected clips recorded with the on-board
cam showing Corsa doing some hunting
of japanese products (wmv-video, 2453 k). A solo ride
(wmv-video, 2118 k) along
free track. These
videos should be
playable with the Windows Media Player by just clicking the links. If
your browser/player is configured to some specialized manner, you might
need to download the video-files onto your harddisk before playing by
clicking the right mouse button and selecting save.
is performing very well.
The smooth power delivery and high torque already at low revs make the
bike very easy to drive and on the track it is therefore possible to
surprise many other bikes having much higher maximum horsepower :)
April 2004, Riding season started 14.4.2004
warm spring days that I was hoping for above finally came and I have
been so busy riding for the last weeks that I haven't had time to
update this site. However, here come some impressions after about 1500
km with My16M.
The very first ride straight out from the garage was about 100 kilometers long. The bike starts well cold or hot. Idling seems to be more stable than ever with the original WM unit. I think that I have managed to stall the engine only once in some street crossing in the beginning when my fuel injection map was not yet adjusted at all. As a starting point I used the stock 1100i Sport map provided by Cliff and downloadable from his www-site.
the stock map everything was fine and smooth above 3000 rpm and bike's
response to the throttle is very crisp. Between 2000-3000 rpm the
bike suffered some "roughness" or even misses when cruising with
the throttle hardly open at all. This problem was alleviated to a
certain degree when I synchronized the injector bodies very carefully.
For this job I have used a Carbtune
II vacuum meter, which is a great tool because you can tape it on
your tank and ride around observing at the same time the difference in
vacuum readings. I tune the injector bodies nowadays with the air
screws fully closed and try to get the readings even around 2500-3500
rpm. There will be a small difference at idle around 1000 rpm, but I
prefer to have the bodies in optimum sync at the driving rpm rather
than at idle.
with good sync I still had some misses at low revs. The Optimiser
together with Communicator and the lambda probe revealed that I was
having lean mixture at those rpms. I fiddled with the map and this
seemed to cure that problem. At the moment I am still optimising the
fuel injection map, but it seems that both between 2000-3000 rpm and
above 6500 rpm full throttle the bike runs better when I have increased
the fuel injection 10-20 % depending on the map location. I have also
done some adjustments to the ignition advance map mainly just smoothing
out some parts of the map.
reasons why the stock map was not suitable to my bike maybe mainly
because I have the Termignoni exhaust system instead of the original
one. I also use a K&N filter in the original airbox instead of the
basic paper filter. These mods propably mean that the engine breathes
in a slightly different manner than a stock one. It seems that the
original exhaust "mystery" below the gearbox has some kind of
"resonance" region around
3000 rpm preventing efficient gas flow.
At the moment I am running closed loop above 3000 rpm. It works just fine. The Optimiser shows nicely what kind of correction factors are used. When collecting this data during a run with the Communicator, it is easy to see what parts of the map are not optimal. Barometric correction works also nicely. At this time of the year the temperature may be in the morning only a couple of degrees above zero and in the afternoon reaching even +25 C. The bike seems to work ok in both cases.
have still some optimising to do, but already now the bike works very
nicely and I haven't had any electrical or other problems with My16M
other than those related to the non-optimised maps. The connection
between the ECU and a PC-computer works ok and it takes only a couple
of minutes to connect the computer to the bike in roder to load a new
map into the
ECU. A couple of times I have managed to unintentionally edit the map
when playing with the Opimiser functions during a ride.
February/March 2004 - Connecting a Nokia phone to the ECU
Well, what's next ? It was
here in Finland
and, therefore, I could not take the bike out and start driving and
tuning work. After looking a little bit closer into the communication
protocol between the My16M and Optimiser (or a PC computer), I got the
idea that maybe I could hook a Nokia 9210i Communicator with the
My16M. After quite a bit experimentation I managed to plug the
Communicator phoneto the RS 232-adapter using in between an additional
self-made adapter providing suitable power to the special kind of
RS-cable of the phone. The phone cable needs electrical power, which it
usually steals from the RS-port of the host computer, because the cable
itself has some active components inside the 9-pin plug in order to
change the voltage levels to become suitable for the phone. Here you
can see the phone
talking to the ECU over the RS 232-bus.
am running a terminal emulator software in the phone and with it I can
datalog all the stuff flowing from the ECU into an ascii file. A 64MB
memory card in the phone provides enough capacity even for a longer
test ride (> 30 hours). All sensor readings and ECU parameters are
recorded at 5-8 Hz into the phone memory. Now I have a very convinient
datalogger that I can keep in my pocket when driving around with the
bike. The next thing was to write a program that decodes the
hexadecimal strings coming out from the ECU. Thanks to the instructions
by Cliff, I can break the hexadecimal codes into meaningful parameters,
like rpm, injection pulses in milliseconds, ignition timing in degrees,
lambda reading in volts, etc. This software is still under development,
but in the end it allows me to study the datalogged info with
three-dimensional graphs to find out what parts of the fuel injection
map are out of tune.
very happy with this kind of somewhat unexpected development, I decided
to go on and put a lambda probe into the bike. A real time lambda value
data logged during driving into the Communicator should help to find
optimal settings for the fuel injection map. Here is a
picture of the lambda probe
installed into the right hand exhaust pipe. Also the Optimiser is able
to use lambda values for tuning, as you can find out from Cliff's
next thing is to calibrate the lambda probe against a car service shop
lambda-meter, but I need to be able to drive there first, so the snow
has to go before that. I have already checked the signal from the
lambda probe using a digital
oscilloscope. The probe seems to give rather steady signal, but
when then air-fuel ratio rapidly changes (when flipping the throttle
quickly, for example ) the probe seems to be able to react with about
10 ms response time. I also experimented shortly with the closed loop
control provided with the My16M, and the ECU seems to be able to
correct the fuel injection values in almost real time depending
on the readings of the lambda probe.
is now nothing much more I can do before I can get on the road
and start the actual map tuning work.
February 2004 - First engine start with
Okay, in the end of February
I finally serviced my bike. This included, among other things,
adjusting the valve clearances, setting the TPS (150 mV) and
synchronizing the injector bodies using the original WM16M to run the
bike. I also modified the link between the injector bodies with a
somewhat better joint
original plastic one. Then it was time to replace the original
ECU with My16M. Here you
see My16M ECU installed to the bike. After carefully checking once more
all the wirings I hold up my breath and turned the ignition key on. The
fuel pump operated for a couple of seconds and then silence. This was
exactly how it should do and nothing seemed to be heating up or smoking
in the My16M or anywhere else. So I thought that my wirings must have
been quite ok. Next thing I pushed the start button and after a few
cranks the engine bursted into life !
Wow, it took only a couple
of seconds after the first start and the engine was already idling very
beatifully. I checked the readings from the Optimiser
installed into the
cockpit and all the sensors seemed to be giving correct readings. The
motor responded to the throttle very nicely, without any noticeble
misfiring or any other problems. I was really amazed that it went so
well and smoothly just simply using the maps provided by Cliff and
optimized for his stock 1100i Sport. I quickly hooked the My16M with my
PC computer running the ECUController program and it also worked
without problems. The engine responded also immediately to all changes
made by the ECUController program.
January 2004 - Bench testing
Here you can see my the “test setup", where the ECU unit is powered from the bike’s battery for initial testing and the Optimiser and the ECU are already “talking” together nicely . I have also tested the coupling of the ECU via the RS unit with a PC running the ECUController program. Everything seems to be ok so far.
The next thing is then
to try the unit with the bike, of course. Before this I will, however,
need to do some minor service with the bike including synchronization
of the injector bodies and calibration of the TPS (throttle position
sensor). I will do this by running the
bike with the original WM16M and after that I will switch to My16M. I will be reporting here about my first
impressions when I get the bike running first time with the My16M. So
far everything looks very promising and especially the capabilities of
the Optimiser are very intriguing.
December 2003 - Kits arrive
The kits arrived without any problems in a few weeks time and they looked like this. Here you can see a picture of the original WM16M circuit board taken out of the WM box and to be replaced with the My16M unit. The first thing that I did was to cut the original WM connector loose from the WM box/circuit board and to modify the backside of this connector to more easily adopt the new wirings. This modification included the use of a plastic support plate (shown in the picture) and epoxy glueing to support the pins upwards compared to their original position directed downwards, i.e.towards the original printed circuit board.
Cliff gives detailed instructions on his www pages how to assemble the ECU, the Optimiser and the RS 232 unit. However, in order to build these by yourself starting from the component level one should have at least some basic knowledge about electronics. Also decent soldering equipment makes the job easier. After some work the units started to come together and this picture shows the fitting of the bare My16M board inside the WM box. It is a rather tight fit but it goes there.
Here you can see the assembled My16M ECU together with the RS 232 link. The RS link makes it possible to connect the ECU to a PC computer via a standard 9-pin serial cabel (COM1). From Cliff’s pages you can learn more about how the PC can be used for tuning of the ECU unit.
Here is the Optimiser seen
from the outside,
and a closer look of
powered up. Here is a closer look
inside the ECU and the finished wiring
going into the WM
connector. Wiring this connector is maybe the most annoying tasks when
assembling the ECU. I have used a BNC
connector in the ECU to
provide connection with an optional lambda sensor. The same picture
shows beside the BNC the computer-type connector to connect the ECU to
the Optimiser, or via the RS 232 unit to a PC.
November 2003 - Winter time, some background
Check out My16M web site by Cliff Jefferies to see the huge amount of development work that he has put into this subject in order to provide an alternative for the Weber Marelli 16M ECU unit (herein WM16M). WM16M is used for powering many Moto Guzzi and Ducati models and it is known to have certain “features”, that the owners necessarily do not appreciate. For tuning purposes, WM16M also does not allow any easy way of changing the fuel injection and ignition timing maps. The maps have to be changed by changing an EPROM chip. This limits severely any tuning activities.
Because there was
actually nothing needing fixing, or even major service in my recently
acquired Sport Corsa, therefore,
I decided to fiddle around and learn something about ECUs during winter
2003/2004. What would be a better way to do this than to build one by
yourself ! Luckily, I managed to source a
broken WM16M unit for these exercises (originally from a Centauro,
thanks Moto Italia). This makes it very easy to revert back to use the
original WM unit, if ever necessary. For any future engine
modifications, a "tunable" ECU is also very interesting.
In December 2003 I ordered
full kits for the My16M ECU, the Optimiser and the RS 232
unit from Cliff.