Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Ducati 2 Valve 1100 - Individual Cylinder Fuel maps?

Status
Not open for further replies.

gt6racer2

Automotive
Oct 10, 2005
97
0
0
US
Hi, I'm about to purchase a fuel map device for my Hypermotard.
As background, the stock system has a single O2 sensor located after the two header pipes come together which provides a signal for closed loop feedback at low throttle and rpm. (<20%throttle, <5,500rpm). Above this range the stock system then goes to map based open loop.
My new device (Power Commander V) will not change basic process for low throttle operation, but will move the map base a little more rich. For higher throttle/rpm, the new device will have an autotune capability, whereas the signal from an additional O2 sensor(s) will be used to automatically develop the fuel map to a predefined air/fuel ratio map.
My question is whether it is considered that the fuelling needs of the two cylinders are likely to be different enough (due to temperature difference etc..) that I should have two sensors, one for each cylinder, or if one would be sufficient. My concern would be that if the difference between the fueling needs of the two cylinders is significant, with one sensor I'm running an average - ie one cylinder rich and one lean. Cons of two sensor setup are cost and complexity - I'd need three O2's - the stock one plus one in each header.
Thanks, Andrew
 
Replies continue below

Recommended for you

Due to the uneven firing pattern, the fuel demands for each cylinder won't be identical, but it's pretty likely that the stock ECU takes account of this (if it is meaningful). The Power Commander just adds or subtracts a percentage from whatever the stock injector pulse duration is, so whatever logic that the OEM ECU has in it gets preserved and passed through, just with a plus/minus superimposed on it.

Although the fuel demands won't be identical, they're unlikely to be different enough to be meaningful. The engine still has a separate throttle per cylinder, and that overwhelms all tuning effects at part throttle. It has a separate intake runner per cylinder drawing from a common plenum ... at full throttle, the effects of that plenum will be a fraction of what the individual runner tuning does, and those are the same for both cylinders.
 
It uses narrow band O2 sensors. These will tell you which side of stoich you're at but not by how far reliably, see attached graph. I wouldn't trust autotune using narrow band sensors for other than the low rpm/low load operating area - which is exactly what the factory did. Mr. Locock is spot on IMO - plug reading tells you what's really going on. The factory also has separate maps for each cylinder in some of their ECUs, typically they run the vertical/rear cylinder a little fatter than the other.
 
 http://injectoranalytics.com/images/narrowbando2.jpg
No, no, no. The OEM O2 sensor is a narrow-band sensor. The Power Commander autotune uses a wide-band.

I haven't used a PC autotune on a vehicle that was originally equipped with an OEM O2 sensor. It should come with a means of eliminating the OEM sensor so that you only need one sensor (the wide-band) in the exhaust system.
 
Thanks for the replies. I agree that a pull of the plugs is useful to check for appropriate mix.
The posted replies suggest to me that the stock system likely has different high end maps for the two cylinders, and also that the Power Commander makes its change as a delta from the base map. Putting these two together implies to me that I'll still have the stock fuelling difference with the new set up.
The instructions on the Autotune imply the ability to select AFR from 12 to 14.7, apparently confirming Brian's last comment.
Based upon this, it appears the one sensor Autotune is an effective route. Thanks for the help.

Brian, I can get to one sensor but that requires to have the stock ECU reflashed to change it to open loop so that the Power Commander drives the full range. However, on my bike, the stock set-up does a decent job at cruise, albeit feeling a little lean, and so I don't really want to change it. I haven't seen any options that allow the stock ECU to use the Autotune sensor.
Given the helpful replies so far, I'll share what I'm up to in case there are better ideas. All I want to do is to remove the CAT. By doing this I can save some weight and let the engine breath a little more easily - as an engineer, a win/win. However, as the stock system is open loop at high RPM, I will be adding more air to a system tuned to run pretty lean already, and it will not adjust for it. Other have tried the mod without any fueling changes and the plugs have shown it runs lean. What would you do ?
 
Doesn't the autotune kit come with either a sensor-eliminator, or a harness from the PowerCommander that allows it to emulate the signal from the narrow-band?
 
Brian, Wanting to have a clean solution I had asked DynoJet about this and they replied :
"If using Auto-tune, you need to make sure that Auto-tune is disabled in the stock closed loop range to keep the Auto-tune O2 sensors from fighting with the stock O2 sensors. You do this by entering a Target AFR of zero (0) in the stock closed loop range. You will notice that all of our predeveloped map files for the bike model are already setup like this.
You need to keep the stock O2 sensors in the exhaust and active. If you are adding Auto-tune you would need to add additional O2 sensors, but the stock sensors will still need to be in the exhaust. We have no way to eliminate the stock O2 sensors, short of reprogramming the stock ECU."

 
I don't know anything about Ducati ECU's but a lot of bikes these days (like my Triumph Speed Triple) can be easily and cheaply reprogrammed with just a lap top and a USB to OBD2 cable. It's a much cleaner, cheaper and versatile system than any of the piggyback systems.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.
 
dgallup, no such luck with my Italian machinery. I can pay for a reflash, or buy the Power Commander (which then gives me the ability to map it myself). Cost is similar, but of course with the latter route I can then make future updates as I please. The second step is how to deduce the new map - either get a dyno run done, or get the Autotune. As I'll have bought the Power Commander already, the Autotune is the cheaper option, and of course it can adapt to later changes, as well as automatically compensating for environmental variations.
 
There are other options for reflashing the stock ECM on the hypermotard:


These guys are pretty good to deal with. I've used their stuff on my Triumph, worked great.

If you are concerned about the cost of doing the autotune with multiple sensors, can you just tune 1 cylinder at a time, and swap the sensor between the 2 pipes?
 
Yes, I've seen Tuneboy, and there's also Bazazz. Each seems to have supporters and detracters - my belief is that they all work well, but some users are more skilled at implementation. Cost is a consideration, but it's more the efficiency of the execution that I'm looking for. Your posting does give me an nice idea though - in that I can perhaps remove the Autotune and sensor(s)once I've developed the map.
 
TuneBoy is completely different from Power Commander and Bazazz. Tune Boy lets you interact with your ECU with a PC via a USB to OBD2 cable. You can read all the codes, sensors, balance your throttle bodies and edit the mapping. Virtually everything in the maps can be changed; fueling, ignition timing, rev limiter, fan operation, disable O2 sensor, etc. Pretty much anything the ECU does you can change. I even corrected the speedometer on my Triumph.

The piggy back systems like Power Commander are very crude by comparison. You have to splice into the wiring harness and the things that you can control are limited.

There is a freeware competitor to TuneBoy called TuneECU but I don't think they support Ducs at this time.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.
 
Your previous post seemed to indicate that you thought you could only buy a reflash, or a piggyback. The Tuneboy allows you to custom map the factory ECM, not just a canned reflash. At one time, they also had a feature to allow you to import PC maps into their software, I'd assume it is still doable. If not, it is easy enough to manually import.

The selling feature for a bazzaz(to me) is the shift kill feature. I have a CBR1000RR with their full setup. And a Z1000 with a full PC+ignition module. But for efficiency of execution, you won't be able to beat the leaving the factory ECM in control and remapping it with the Tuneboy.
 
Update : Based upon the comments I purchased the Tuneboy system, which has allowed me to flash new TP/RPM/Fuel maps direct to the ECU. Each cylinder has it's own map. I also purchased a wide band O2 sensor, from which I can collect TP/RPM/AFR data, and I can then use the Tuneboy software to adjust the fuel map based upon this data. Unfortunately I can't ride at the moment (non-related issue) but I have flashed a new map for my modified set-up and the bike runs great in the garage. Through the flash I've also removed the low throttle closed loop, so I now run open loop, and (at least under no load), the throttle response is now much more linear.
Thanks to bradrs and dgallup for the lead.
I do have one question from all my research. This map based fuel injection system is commonly referred to as "Alpha-N". This gives a single map value for fuel dependent only upon TP and RPM. However, I expect the fueling requirement at a specific TP/RPM is different under acceleration, steady state and slowing. I have found in my research devices which claim that Alpha-N strategy does not account for these differences, and that a load based modifier is necessary ( EJK electronicjetkit.com for example). My suspicion is that the claim is technically correct - ie pure Alpha-N strategy does not in itself allow for this, but I also expect that the motorcycle ECU has a strategy above pure Alpha-N that compensates - ie nobody is actually running pure Alpha-N. For example, I've read that the fuel is cut when the throttle is shut - which would not occur under pure Alpha-N. Does anyone know what the motorcycle ECU typically does to adjust the map for load when running Alpha-N ?
(In case it's of interest my ECU is the Marelli 5AM.)
Thanks, Andrew
 
Glad Tuneboy is working for you. I don't know the internal workings of the Marelli ECU, but I know the old SAGEM MC1000 system definitely had multiple simultaneous strategies to adjust fueling beyond the map values based on other sensor readings. It would adjust based on engine temp, throttle position, rate of change on throttle position, etc. In particular, it had quite complex programing to account for wall wetting of the intake manifold. During transients there is fuel mass going into and coming out of the "puddle". So yes, I am quite sure the ECU is doing a lot more than just look up numbers in the static tables.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.
 
The wiring diagram shows an atmospheric pressure sensor which probably measures the pressure in the intake after the throttle blades (ie engine vacuum). You can more or less determine the engine load and required fuel combining that with rpm, throttle position and coolant temperature.
 
Thanks for the posts - it appears that my suspicion was correct and there's math being done on top of the Alpha-N map. Now looking forward to a chance to run the bike and see how she performs with the changes...
 
Status
Not open for further replies.
Back
Top