CO setting

[Lucky Phil]

Was reading one of Brad Blacks reports yesterday and came across something I didnt know.

One very important thing to know about the 1.5M ECU is that it runs rich for the first 3,000 rotations every time you start the engine after turning the key or kill switch off and then on.  I just thought (while typing this) that I don’t know what happens if the engine stalls and you restart it – I’ve never tested that.  So best to assume it does it every time.

And this is not just when it’s cold.  Every time!  Even when it’s hot.  If you start it up and check the idle mixture quickly and it’s 5% CO after idling for 3 or so minutes (3,000 rotations at 1,100 RPM idle is about 3 minutes) the mixture will drop to maybe 1% CO or so and it’ll idle like crap.  I usually allow at least 4 to 5 minutes for the mixture to settle after starting the engine.  It’s very annoying when you’re in a hurry, as you have to turn the engine off then back on to initiate the setting procedure with the diagnostic tools, but that’s how it is.

For the full report and hours of interesting reading...... http://www.bikeboy.org/ducati2vthrottleb.html

Ciao

[stewgnu]

That’s interesting.  Would that be why, if the bike doesn’t fire up properly and stalls once or twice, that when it does so it’s really lumpy for a minute or so?

[Tomchri]

4 hours ago, Lucky Phil said:

Was reading one of Brad Blacks reports yesterday and came across something I didnt know.

One very important thing to know about the 1.5M ECU is that it runs rich for the first 3,000 rotations every time you start the engine after turning the key or kill switch off and then on.  I just thought (while typing this) that I don’t know what happens if the engine stalls and you restart it – I’ve never tested that.  So best to assume it does it every time.

And this is not just when it’s cold.  Every time!  Even when it’s hot.  If you start it up and check the idle mixture quickly and it’s 5% CO after idling for 3 or so minutes (3,000 rotations at 1,100 RPM idle is about 3 minutes) the mixture will drop to maybe 1% CO or so and it’ll idle like crap.  I usually allow at least 4 to 5 minutes for the mixture to settle after starting the engine.  It’s very annoying when you’re in a hurry, as you have to turn the engine off then back on to initiate the setting procedure with the diagnostic tools, but that’s how it is.

For the full report and hours of interesting reading...... http://www.bikeboy.org/ducati2vthrottleb.html

Ciao

Ok, interesting, meaning I can't wait to find out what Meinolf's mapping will do. Actually they both fired easily, not idling perfect, but I know I have to do the sync when the weather permitting to do it outside . More later. Time for another Chang beer.

Cheers tom.

[68C]

Although this thread does refer to the WM15 ecu,  it seems to only discuss the Ducati bikes that use that ecu.  I know the WM15 ecu fitted to both Guzzi and Ducati are the same, I am using a Ducati 750ecu on my V11 for tinkering and playing with the map, the maps of course are different.   Is this 3000 rotation determined by the map or is it hard wired into the ecu?  As this is the first time I have seen any reference to this 3000 rotation I wonder if it is relevant to the Guzzi V11.

[docc]

Yes, it applies to the V11. I’ve seen references to it in the past, but don’t recall which threads. 
 

Meinolf could tell us where the function is in the ECU.

[Meinolf]

2 hours ago, docc said:

Yes, it applies to the V11. I’ve seen references to it in the past, but don’t recall which threads. 
 

Meinolf could tell us where the function is in the ECU.

Hi,

 you called me, the genie comes out of the bottle.

Gentlemen, let me set things straight.

There's in fact a warm-up function which in all Marelli 15xx BINs I've come across runs for 2000 engine revolutions after start, regardless of bike manufacturer.

Neither the warm-up table nor the other trim tables mentioned below are hard-wired. They can be easily changed. I suggest one should know what one is doing before making any changes, though.

The factor applied to the value which eventually will determine the injector opening time is in a 2D table. The x-axis is revolutions since start (every start), the 2nd axis is engine temperature. Search for warm-up table and you will find pictures of the table. Or use Tunerpro and load a BIN.

The typical warm-up table has a factor of ~ 600% at -18°C engine temperature for 2 revolutions after start, the value tapers off with increasing revolutions. At higher engine temperatures the values are lower and taper off more quickly. Either way the warm-up function ends once the revs since start, every start, have passed.

But....

There are two further temperature trim tables, also factors, which are part of the fuel calculation. One is a 1D engine temperature-indexed table, the other one a 1D air temperature-indexed table. Plus another other trim tables for air pressure and acceleration/deceleration.

So the fuel calculation looks roughly like this.

Fuel map value x factor engine temp x factor air temp x factor warm-up x factor air pressure x .... = final fuel value -> converted into a injector pulse width measured against end of pulse in relation to crankshaft as determined in the fuel phase table.

See, no magic involved, very basic stuff.

I've found that, as with many other trim tables, the warm-up and engine temp trim tables in Guzzi (and all others) BINs are not as dictated by the laws of physics and the circumstances of the engines (V11 and California) I've logged. Accordingly the warm-up and trim tables in my BINs are corrected.

I've stopped my work on the V11 BIN 2 years ago and moved on to the Jackal and currently the 5AM ECU used in the Guzzi CARC models, the Aprilia Mana and many Ducatis and Morinis. So the more recent posts and results of the reverse engineering can be found here:

https://wildguzzi.com/forum/index.php?topic=93758.0
or
https://www.ducati.ms/threads/5am-bin-analysis.718031/
or
https://www.apriliaforum.com/forums/showthread.php?336210-ECU-tuning-Looking-for-information

Any further questions? No, then back into the bottle

Cheers
Meinolf

[Meinolf]

Hi,

as chance goes I've been revisiting the engine temperature related trim functions on the Jackal, since the current temperature are below 0°C and the respective engine temp breakpoints could be reached. The Jackal is also using the 15M, so it's seems relevant to this thread. 

The start enrichment map and the fuel trim engine temp map were already changed several years ago on the V11 and Jackal, but at that time the logging equipment was from Innovate. Unreliable product and poor support led me to changing to the Zeitronix ZT2s. Those we hooked up to the Norge, but I got several more ZTs and mounted them on the Jackal.

The basic measuring procedure was already planned last year. The ZTs have two 0-5V inputs. But, both are being used during the normal logging. The first one is attached to the TPS, as the standard input only provides for measuring TPS as a integer %-value. Which is not good enough in the lower third of the TPS range, %-values overlap the degree values. The 2nd input gets the lambda signal measured by the 2nd ZT2. So the 2nd ZT2 was temporarily disconnected and the engine temperature sensor connected instead. 

As mentioned above the engine temp related fuel trims are taken care of by two mechanism. The first is the start enrichment map, indexed by engine temperature and number of revolutions since ignition on. I've changed the standard 2000 revs in my BINs to 1000 a long time ago to minimize the interference during the normal logging. The index is to be understood working in reverse, to the first step is 1000-984 = 16 revolutions. And so forth, until the total number of revolutions, specified in a scalar, has passed.

The fuel trim engine temp map is a 1D, only indexed by engine temperature. Both functions overlap, so a trim by this table is factored together with the current start enrichment factor. One of the pictures shows both tables.

The 2nd picture shows the log of a cold start at ~65°C, which is the voltage coming from the sensor at that temperature. First line is lambda, then rpm, then TPS and then the voltage from the engine temp sensor.

As preparation a look-up table voltage vs temperature needed to be created. This was done by running the ZDL software and GuzziDiag at the same time and noting down the corresponding values. The measuring procedure is starting the engine with and logging the data. The inital data, while the rev countdown is still running, is used to check and correct the  trim values in the start enrichment map, then shutting the engine down and repeating the procedure at every engine temp breakpoint. Which takes 20-30 tries, as the engine temperature keeps rising by 5-8°C after the engine was shut off.

Longer runs are used to get lamba values across the temperature range without the start enrichment influence. These are used to check and correct the fuel trim engine temperature table. And this opportunity is also used to re-evaluate the main fuel value at the lowest rpm/lowest TPS breakpoint, which I don't normally reach while driving. My idle speed is higher than that, but it is nicely reached during cold starts.

It will take another day or so, but the first results are very good.

The original values in the start enrichment map are much to rich, a lambda of <0.7, down to 0.6, is standard. Which is puzzling, as this is below the point at which the mixture will ignite at all. Good engineering by Guzzi, maybe. The corrections already made brought lambda in the range of 0.84-0.92, which I will try to equalize as good as possible. 

The fuel trim values are more or less finished also, at least in the range from 0° to 90°C. The higher values will be done while road logging, Beard kindly added a temperature filter function to the database program he wrote to analyze the massive data volume created by the ZT2, close to a million data points in one hour.

Cheers
Meinolf

[bbolesaz]

cold start at ~65°C

??????   150°F ?

[Meinolf]

Hi,

It might not be that cold, but the breakpoints go up to 125C. So yes, 65C.

Cheers

Meinolf

[Meinolf]

On 2/7/2020 at 5:28 PM, bbolesaz said:

cold start at ~65°C

??????   150°F ?

Hi,

on 2nd thought your comment deserves a more detailed reply.

The name cold start is not well chosen. Start-up fits better. The reason why the start-up function is needed and why it's temperature and to a lesser degree revs since start dependent is the following.

After pressing the starter button the injectors begin injecting fuel, the nozzle and the fuel jet are pointing towards the opposite side of the intake pipe . At that time there's almost no air flow and a very low air speed moving through the intake, hence the first fuel shots are creating a fuel puddle on the opposite surface of the intake tract. A surplus of fuel, which is caused by the trim values, during that period ensures that enough fuel gets into the combustion chamber to achieve a ignitable mixture. This effect is engine and air speed dependent, so a temporary enrichment for some revs is needed at any engine temperature.

Once the engine is turning over at a higher speed, the resulting higher air speed moves more of the injected fuel into the combustion chamber. And the increasing temperature of the engine increases the evaporation of the fuel, again resulting in more fuel getting into the combustion chamber.

My current conclusion, based on the measurements, is that the critical values in the start-up table are those used during the first 10-50 revs. Once the engine is running on its own the start-up table values can be replaced by fuel engine temp trim table.

Cheers
Meinolf

[bbolesaz]

Got it.  Kinda surprised the values are still changing above around 25C.