| Author |
Message |
    
Arctic_firebolt
| | Posted on Saturday, January 03, 2009 - 09:08 am: | 
|
All I know is that so far my '08 XB12R runs like S%$T, has the worse MPG of any bike I ever owned whether I ride it easy or higher up the RPM range. Yet it has the highest HP:weight ratio and so far I am thoroughly disappointed with this bike's mechanical ability. I believe they went completely overboard engineering the FI system.
The bike is at the dealer AGAIN! This time it has a loud clunking sound as I apply the front brake, which is in time with the ground speed. Probably a warped rotor or bad caliper. Asked for them to check on any ECM updates. AND.....the clutch cable upgrade/recall, remember that one? Well even though it was done the cable is STILL melting. What will it take to get this bike to run properly for a street machine?
(Message edited by Arctic_Firebolt on January 03, 2009) |
Macdiver
| | Posted on Saturday, January 03, 2009 - 09:22 am: |
|
The post by pkforbes gives the calibration mode region. The AFV section states
The AFV will be learned when the bike is operated at engine speeds between 2500 and 3500 rpm at road speeds in the 40 to 60 mph range under a steady light load (no down grades or steep upgrades, decelerating or accelerating) for 2 to 3 minutes.
Since it does not learn while going up or down steep grades, this may be the source of the problems in the mountains with rapid elevation changes. |
Arctic_firebolt
| | Posted on Saturday, January 03, 2009 - 09:34 am: |
|
You make a good point Macdiver. I realize your post was in response to someone else but I wanted add to it. I live in Florida, where I believe the highest elevation in the state is about the height of a normal speed bump. LOL. I have ridden my XB12R for extended miles on the straightest, flat road imaginable. So, I am convinced the DDFI-3 is still an issue on the '08 models.
After listening to many of the '07 owners it seems their bikes just perform better.
Time will tell how Buell confronts the situation.
(Message edited by Arctic_Firebolt on January 03, 2009) |
Paint_shaker
| | Posted on Saturday, January 03, 2009 - 09:38 am: |
|
DDFI 3 uses a cam position sensor to set timing, as static timing on the 08's and newer has been eliminated.
I did have a stumble and some surging in the mid RPM range on my 08 Bolt, which has since been corrected by my local ecmspy wizard.
Prior to the ecmspy adjustment I ran my bike in the Ga mountains through various elevation changes (up to 6000 ft) with zero problems.
I also have had no problems with the clutch cable melting since the installation of the upgraded part. |
Id073897
| | Posted on Saturday, January 03, 2009 - 11:36 am: |
|
DDFI 3 uses a cam position sensor to set timing, as static timing on the 08's and newer has been eliminated.
DDFI and DDFI-2 use a cam position sensor. DDFI-3 uses a crank position sensor. |
Blake
| | Posted on Sunday, January 04, 2009 - 01:36 pm: |
|
quote:The AFV will be learned when the bike is operated at engine speeds between 2500 and 3500 rpm at road speeds in the 40 to 60 mph range under a steady light load (no down grades or steep upgrades, decelerating or accelerating) for 2 to 3 minutes.
That is inaccurate. |
Blake
| | Posted on Sunday, January 04, 2009 - 01:37 pm: |
|
At least based upon what I've personally witnessed. |
Dogdaze
| | Posted on Sunday, January 04, 2009 - 08:56 pm: |
|
Blake,
as a noob looking to learn, what has been your experience with the AFV learning parameters? |
Swordsman
| | Posted on Sunday, January 04, 2009 - 11:14 pm: |
|
Id,
"XB don't switch to OL at 3500 rpm. "
Okay, so the "official" data posted earlier says:
"CLOSED LOOP
Cruising between 40 - 60 mph at operating temp.
(1500 to 3500 rpm)"
Since there's no sorta-kinda' loop, I would have assumed the switchover to open loop would be beyond the 3500 rpm range. 'Course, I also notice it states open loop is 4000 rpms+, so the document is leaving a 500 rpm window that it doesn't account for either way.
But see, this is exactly why I started this post. This person says one thing, that says another, the document says AFV takes 2-3 minutes to adapt, Blake claims it doesn't... there's no end-all-be-all definitive answer to EXACTLY how the system is put together. It's annoying.
~SM |
Iamarchangel
| | Posted on Monday, January 05, 2009 - 12:22 am: |
|
Part of the problem is that you're looking for a specific rpm point for the changeover and forgetting that temperature has a role too.
The ECM is measuring data from six sensors. You're trying to establish it based on just one.
Even carburetor jetting would vary by season, day, altitude, etc. The ECM has just digitalized all that info, and more.
Florida data today is not going to match Ontario data, but we would expect both bikes to run equally well right now. (Except I'm not going out in the freezing rain/snow tonight, ha.) |
Paint_shaker
| | Posted on Monday, January 05, 2009 - 12:26 am: |
|
Opps... my bad Gunter. Crank Position Sensor it is! |
Id073897
| | Posted on Monday, January 05, 2009 - 12:30 am: |
|
Okay, so the "official" data posted earlier says:
The ECM doesn't care what's written in some f*cking outdated document. It cares about what's written to it's EEPROM. |
Bombardier
| | Posted on Monday, January 05, 2009 - 06:17 am: |
|
Gunter, Does the ets have the same function in the DDFI-3 as in DDFI-2? |
Id073897
| | Posted on Monday, January 05, 2009 - 08:45 am: |
|
It still sticks in the rear cylinder and monitors engine temparature. |
Bombardier
| | Posted on Monday, January 05, 2009 - 10:05 pm: |
|
I meant does it still provide the basis for changing injector pulse widths and ignition timing?
Has anyone coated the outer body of the ets with a heat resistant coating to reduce heat soak? |
Id073897
| | Posted on Tuesday, January 06, 2009 - 01:49 pm: |
|
I meant does it still provide the basis for changing injector pulse widths and ignition timing?
I think so. |
Xoptimizedrsx
| | Posted on Tuesday, January 06, 2009 - 07:11 pm: |
|
playing with pulse widths is a very touchy area. you get it good you think then you find out its way off. the stock widths are very good. if you go up much at all you get drips. the stock injectors just cant apply more than the nozzle can handle. more is not always better.
look at cams bigger isnt always better.
look at the 551 e cam which is in the buell as our basic profile. and the 536 cams. a lot different yet the 536 will provide more peek hp on a totally different timing than the 551e. look at many engine building sites. stick with the stock pulse widths.
the base injectors are good up to 140 hp. if you need more or want better get a better atomizing replacement injector. there are many drop in upgrades.
the ecm has many movable areas with the correct knowledge. you will need a few guages to read everything in temp , flow, o2, and co2
all matched you can get the timing spot on as well as the fuel. then you pretty much set the power of the mechanics of the engine by controlling the afr. the bikes on NA like 13.2 max fuel above 3000 rpm.
if you fighting the learn closed loop just move the scale where it learns or set the learn time up to read longer before updating. these buell ecm are not rocket science compaired to mant car ecm's i work with. these are very basic. controlling them is only having the correct guages to read the engines functions.
mike |
Blake
| | Posted on Wednesday, January 07, 2009 - 06:29 am: |
|
"Blake,
as a noob looking to learn, what has been your experience with the AFV learning parameters?"
Following an AFV reset to 100, I've seen the AFV change and stablilize almost immediately after just seconds, not minutes. One clarification, the engine was already at operating temperature. |
Id073897
| | Posted on Wednesday, January 07, 2009 - 07:38 am: |
|
I've seen the AFV change and stablilize almost immediately after just seconds, not minutes.
Maybe this is a slight inaccuracy which made it into the document because the AFV is written to EEPROM only once a minute. I don't know, if the AFV has to be set for a minute to save it, or if it's simply written every minute. Changes visible in the software will therefor not necessary reflect in the EEPROM data. The delay is configurable. |
Moosestang
| | Posted on Wednesday, January 07, 2009 - 05:49 pm: |
|
Following an AFV reset to 100, I've seen the AFV change and stablilize almost immediately after just seconds, not minutes. One clarification, the engine was already at operating temperature.
Have you data logged to see how the AFV changes? I have and it will go up to 105 one minute and then down to 95 the next, but still be 100 when I shut off the bike and look at the AFV in ecmspy. Just looking at it in ecmspy is misleading. |
Id073897
| | Posted on Thursday, January 08, 2009 - 02:48 am: |
|
 |
Blake
| | Posted on Friday, January 09, 2009 - 04:45 am: |
|
I wouldn't care what the AFV is after shut-down. I care that it is adjusting appropriately while the bike is running. It does. |
Blake
| | Posted on Friday, January 09, 2009 - 04:49 am: |
|
Thanks for those plots Gunter. Very interesting, eight AFV adjustments over 69.25 seconds? Can you elaborate on the associated conditions that generated that behavior? |
Id073897
| | Posted on Friday, January 09, 2009 - 09:16 am: |
|
All input had been synthetically generated on the bench. The missing peaks of O2 voltage are not "missing", but "missed" as polling and the O2 frequency generator were not synchronized.
AFV is adjusted in steps (~105%) and the EGO correction looks more as takeing place in waves, where regions of lesser variation alternate with regions of larger variation. This is probably a result of the ECM's PI controller.
There are 5 "periods" of EGO corrections, and 8 steps of AFV changes, so assumably AFV and EGO correction use different algorithms. |
|
