Interesting thing that Gunter found, check it...


So, as I tried to rip open the throttle to get some WOT data, the ECM decided it had other ideas. The WOT corr., which is normally 100% until the throttle (plotted as load) is wide open [see note 1], where it changes to 110%, is reduced to 71.5%, then ramped back up again. I included the injector pulse widths also to show the fuel being reduced.

When this happened, I just thought "damn big engine wasn't ready for that handful", cos it just bogged down...

This is the Noise Abatement Logic, what the White Wire Mod fixes. It is only the 9s that people do the White Wire mod to, but this data shows you that it is worth doing on the 12s too...

Easy to fix, see the How To section... I have already set the whole of byte x0194 to 0.

Notes:

1. The WOT threshold is defined by a throttle / RPM relationship - on my Uly with an IB310 ECM it is:



The defintion of this is held in bytes x0058 to x005b in the order:

_________________
08 Specialized Langster

Can't really do much with the WB sensor measurements at the moment due to:

1. Noise abatement logic polluting the WOT data
2. Concerns with measurements on the front cylinder

But, we can remap the rear in the closed loop area with the stock narrowband sensor.

So, this is what I did:

1. Put the log into Excel, saved it under another filename and deleted anything with an OL Corr. or WOT Corr. greater than or smaller than 100%. You can probably filter it out in MegaLogViewer, but I have not really mastered that yet.
2. Changed the calculation in MLV from...

NB_recommendedCellValueFormula=(cellValue*[Field.egoCorrection]/100*[Field.AFR]/afrValue)+(cellInterpolated*[Field.egoCorrection]/100*[Field.AFR]/afrValue)/2

...to...

NB_recommendedCellValueFormula=(cellValue*[Field.egoCorrection]/100)+(cellInterpolated*[Field.egoCorrection]/100)/2

...becuase the stock solution includes an AFR calculated from the NB characteristic, that I would argue is inaccurate, and besides, the egoCorrection (Gego) is a correction calculated from the O2 sensor Voltage by the ECM, so you are double accounting with the first calc.

3. Set the MegaSquirt.properties file like...

O2Volts=O2
AFR=AFR(WBO2)
TP=Load
egoCorrection=Gego
pulseWidth=PW
RPM=RPM
Time=Time
MAP=MAP
coolantTemp=CLT
speedMPH=MPH

4. Ran the VE analysis in MLV on the REAR map only

And I went from this...

...to...


The stats that MLV created...


Nothing alarming from this nor a cursory glance at the new rear map.

Next job is to put this into the ECM and see how it rides, should be smoother at part throttle now as it is the rear which is controlled and hence provides the closed loop. Then, following on, the WB sensors will be put back in and some more data logged...

_________________
08 Specialized Langster

As I wait for bits to arrive for Son of Grenade, it is back to Ulytune.

I am combining the exercise with some beta testing for Gunter's new ECMSpy for Mono software.

I have relogged data for the rear cylinder and used the new O2 analysis feature in ECMSpy (anyone can tune now - it is a really useful tool - Gunter can explain how it works better than me - as seen here), but before posting results I have a bit of confirmation to do.

One issue I have noted is that I am sure the Uly runs better WITHOUT an aftermarket right hand scoop. I bought a Free Spirits one to try, and yes, the fan is less likely to come on when on the motorway, but the range in which the engine operates, when warm, is now larger. For instance, at 50 to 70mph at cruise, I am looking at 170 deg C and around town I am looking at 215 deg C.

But when hotter, I swear the Uly runs better, the throttle is more responsive and the engine seems less hesitant at part throttle. These traits could easily be put down to "character", but if you have ever ridden a well fueled Buell, you will never be satisfied until yours runs that well.

Suffice to say, the right hand scoop will be coming off...

Shouldn't change my current maps as they have been set up predominantly at 30 to 40mph.

Front cylinder closed loop will be worked next...

_________________
08 Specialized Langster

LH scoop ?.......surely you mean RH ??

_________________
Mithered ta death.
92 MB
96 S2T
98 S1W
00 M2
01 X1
03 P3
10 CR

The right side scoop topic had been addressed in an early Fuell magazine article. Don't know which issue, but a google search came up with this: http://www.sporttwin.com/index.php?/top ... e__st__120
Perhaps the raise temperatures are another side effect.

_________________
"It is often said that before you die your life passes before your eyes. It is in fact true. It's called living."
Terry Pratchett

Yeah, that's what I'm on about...

_________________
08 Specialized Langster

A water cooled engine temp is controlled by a thermostat around 88 to 92 deg C, nice and tight...

_________________
08 Specialized Langster

Just to remind us, this is for an 06 MY Uly with K&N filter and Remus (off an XB12S) with the quietest baffles in...

Started out with the standard map again, but with the Noise Abatement turned off, then logged some data with the engine nice and hot, ran it through ECMSpy for Mono beta version and ended up with this:


Applied these factors to the rear cylinder only...

Took Ms Pash to work, then went to visit Daywalker, logging data as I went. I actually logged three sets of data, but used the one with the engine running hottest, i.e. around town cos it was at this point that I thought I would junk the RH scoop. I ended up with this:


Again, I applied these factors to the rear cylinder only...

At this point, I didn't really feel like I was getting anywhere cos the bike still ran rough at part throttle. It has been a few years since I did the same with the Firebolt so I had forgotten where I had started to reap the benefits. Gunter said that once the lambda sensor went into the front pipe, I would start to notice a difference. Anyway, a narrow band sensor went into the front header:


And I went for a ride, this time to see bikerdude666's trike, then logged data on the way back. I ended up with this:


I applied this to the front map. Next day, I went to work on the bike, then went to Riders to pick up some stuff and logged data on the way back, ending up with this:


Need to update the map again, however, one thing to note is that the areas where there appears to have been no effect from updating the map is due to the number of samples taken (or hits), for example, the point 2900 / 100 only had 13 hits, compared to 2400 / 20 having 1014 hits.

Current maps for those who want to try...

Front:


Rear:


Still need to do the open loop stuff, but the bike runs much much better at part throttle now...

I wouldn't bother trying these maps if you have:

1. An XB9
2. An 08/09/10 model
3. An S3/X1
4. An XB12 with different headers

_________________
08 Specialized Langster

Interestingly, the Uly is very consistent on fuel consumption, apart from when you travel at 90+.

Since changing the maps, it is returning about 6% more mpg.

_________________
08 Specialized Langster

Was enjoying riding Uly today so I logged some data from the front cylinder (the last time I will be doing this)... I did this using a bluetooth dongle lent to me by my favourite virtual friend, Gunter, and a Palm Tungsten T5, which acts as a hard drive when plugged into the PC, armed with ECMSpyBT (free download from the ECMSpy website).

After pumping up the tyres using ECMSpy for Mono, I logged a good 30 miles of data ranging from around town riding to 70mph on the ring road. I got the data off the Palm, ran it through Bin2MSL (free off the ECMSpy website) then went into Excel, where I stripped off all the data with a a WOT correction of 110 and all the data with an OL correction of 105, then I stripped off all the data where the bike ran cooler than 180 deg C...

This is what MLV thinks the front cylinder map should be:


Next job is to reconnect the rear cylinder and repeat the exercise, then move on to the WB sensors...

_________________
08 Specialized Langster

As an aside, whilst logging data, I thought it would be a good idea to see how well the Special O.P.S heat barrier works...

So, I logged some data (got a bit hot eh?)...



...then fitted the blanket...



...just waiting to finish work at a reasonable hour now...

_________________
08 Specialized Langster

I will be really interested in the post blanket results - I've been considering trying to insulate the airbox on my 1125R ...

... although god only knows why, it's not as if it isn't fast enough already !

Or stick some on the frame to help stop the fuel boiling (on the 1125)...

_________________
08 Specialized Langster

Wired up the rear lambda sensor, logged a good 40 minutes of data using the Palm over Bluetooth with around town and ring road / motorway riding...

Converted the log using Bin2Msl, stripped off all data lower than 180 deg C, stripped off anything with an OL or WOT correction then ran it through MLV.

MLV tells me my rear map should be:



Last things to do are the OL and WOT areas. I will do this with the WB sensors and a marked throttle so I can try and keep the running to the table breakpoints at 100, 125, 175 and 255 TPS 8 bit...

_________________
08 Specialized Langster

Temperatures without the blanket fitted...



Temperatures with the blanket fitted....



I was astounded by this. I went the same route, but did it backwards. The pre-blanket test was done with town first, followed by the ring road, the second the other way around. I was thinking this test would be unfair as today, it is about 7 deg C warmer than the first test.

The first set of results show just under 60 deg C inlet temp around town, the second set, with a hot engine, show less than 40 deg C... The blanket doesn't appear to make much difference at high speed...

_________________
08 Specialized Langster