One more time on ECM

[osvaldo]

Hi Guys,

I am a stranger (Italian) Bueller, new in the forum. (you will may be excuse me for my English)
I own a 2009 XB 12 Ss long and I am also an old guy, from the carburetor school.
I love my Buell, but I fell she can go much better than she is now, despite the 'downgrade' after 2007.
So I am hardly working to get in 'deep as I'll able' in the EMC word, and I'll be glade for what you can help.
I will post in times some questions that I could not answer looking in the overwhelming internet info.
Not all will be directed straight to tune, but some will just be for a global understanding, and some other will may be so obvious that is the reason I could not already found answer in the big readings I have done.

The first two are:

In alpha/n system like our Buells the job start from a TPS/RPM map (ex. 12*13=156 cells) where you can find numbers that are related with the injector pulse width. I ask, in the areas between cells are used the same value or the ECM is used to calculate trough a math equation interpolated value?

I feel as a very bad think for the engine the narrowband 02 sensor changing every moment the AFR to get a correct (14.7) medium ratio. Can't it be the reason many Buells (mine too) in the closed loop range run with a sort of sweetened 'hit and miss' mode?

 

[njloco]

This will not be an answer that will help you much with what you've asked but, both of my Buells ( 2008 Ully and 2007 SS ) are pretty much stock with the exception of one ( 2007 xb12ss ) having a Buelltooth with some minor changes, they both run perfect !

Welcome to the forum.

 

[osvaldo]

Hi,
thanks for welcome me.
Mine is not running bad, but I feel she can do much better.
I'm waiting too for a ordered Buelltooth so I will start at least to read the logs just to see in which condition the engine is working.

Have nice trips

 

[Outlaw]

"Downgrade after 2007 " ? What chu talkin bout ?

 

[Cooter]

The number in each cell refers to milliseconds (mS).

Of course you want stoic 14.7:1 while cruising in closed loop. If your bike runs funny, I would suggest to fix it first, then go to tuning. Check the usual suspects starting with the AFV. That will tell you if the ECM is correcting rich or lean and by how much. If you jump to tuning right away, you could just mask an existing issue, making everything a frustrating, expensive, time consuming nightmare.

The basics man. K.I.S.S.

Hey outlaw, maybe he never saw this? LOL.

 

[osvaldo]

Hi Outlaw,

according to what I read in 2008 engines there were some changes.
On one side there were true improvement like on lubrication system and DDFI III ECM that has the potential to better 'drive' the engine.
On the other side, due pollution/noise regulation, Buell guys had to put catalytic muffler, smaller header, and probably much leaner maps.
All that killed engine performance and made the unlucky engine extremely hot.
I never driven a 2007 down bike but I strongly suspect that it has also been lost lot of drivability.
Did you hear a different history?

Have nice trips

 

[osvaldo]

Hi Cooter,

thanks for your advice.
Yes, I am not going to modify anything straight away.
I am just in hurry to have at home my Buelltooth to logging, read, and think about what can cause some 'unfunny' moods of my engine.
I am looking for strong and net pull, not worrying so much for max Hp as I mainly run on Alps roads, as well as a 'clock' idle.
After that I’ll may be going to tune up.
But first I want to answer a few question in order to have a firmly understanding of ECM strategies and where they can be not the best for engine drivability.
I need also to have clear in my mind which software to use, and how to install and connect them; learn that on internet ocean is not easy for me.

Aside my first two unanswered questions, another one is that I am wondering to use one (may be two) wideband 02 sensor to get on log the actual on time AFR, that seems to be impossible with a narrowband.
But, I have not realized yet if and how the signal(s) from the controller can be sent to EcmDroid keeping on the narrowband to have true engine working condition. Any idea on that?

It is a nice and very hot day here, I am going now to feel some Alps air on the face!

Thanks and good trips

 

[cossack84]

Osvaldo welcome to the forum. I love Italy, very beautiful place. Excellent food, wine and very warm people.

First question. Computer extrapolates data in between the cells. Just depends on exact rpm and load. ECM is used to calculate through math equation interpolated value.

Not 100% sure exactly how narrow band is used to measure the O2, but it does pretty good job of keeping air fuel ration where needed. I can watch wide band in real time as narrow band makes corrections, it’s pretty accurate. I don’t think it’s the reason for hit and miss mode.

 

[cossack84]

Hi Cooter,

thanks for your advice.
Yes, I am not going to modify anything straight away.
I am just in hurry to have at home my Buelltooth to logging, read, and think about what can cause some 'unfunny' moods of my engine.
I am looking for strong and net pull, not worrying so much for max Hp as I mainly run on Alps roads, as well as a 'clock' idle.
After that I’ll may be going to tune up.
But first I want to answer a few question in order to have a firmly understanding of ECM strategies and where they can be not the best for engine drivability.
I need also to have clear in my mind which software to use, and how to install and connect them; learn that on internet ocean is not easy for me.

Aside my first two unanswered questions, another one is that I am wondering to use one (may be two) wideband 02 sensor to get on log the actual on time AFR, that seems to be impossible with a narrowband.
But, I have not realized yet if and how the signal(s) from the controller can be sent to EcmDroid keeping on the narrowband to have true engine working condition. Any idea on that?

It is a nice and very hot day here, I am going now to feel some Alps air on the face!

Thanks and good trips

Thats an excellent question and I was wondering that my self. Maybe some one with 2009 can chime in and tell us if there is an unused analog input.

 

[Outlaw]

Welcome Osvaldo, no cats or small headers here in the USA just pure Buell fun.

 

[cossack84]

The number in each cell refers to milliseconds (mS).

One slight correction, those numbers are injector pulse widths. Each point equals to 58 microseconds. Say table calls for 56 points then injector will be open for 3248 microseconds or 3.248 milliseconds.

 

[Cooter]

^^^ Yes. Pulse width, the time the injector is open. I'm just trying to keep it as simple as possible with the learning curve and language barrier

Datalogging with a wideband O2 is vastly faster and more accurate. WELL worth the small expense.
You can add a bung in the head pipe to datalog the front separately if you want. Twice the trouble for 2% accuracy (Based on the front fuel table usually being set 2% different).
Yes, DDFI-3 has at least one extra pin for analog input.
You can datalog and tune with a NB. It will take much longer, but IMO it will end up being accurate enough for smooth running and long as you allow the ECM to make EGO and AFV corrections afterwards (not locking it out).

IMO,YMMV,IIRC,LOL :eagerness:

 

[cossack84]

^^^ Yes. Pulse width, the time the injector is open. I'm just trying to keep it as simple as possible with the learning curve and language barrier

Datalogging with a wideband O2 is vastly faster and more accurate. WELL worth the small expense.
You can add a bung in the head pipe to datalog the front separately if you want. Twice the trouble for 2% accuracy (Based on the front fuel table usually being set 2% different).
Yes, DDFI-3 has at least one extra pin for analog input.
You can datalog and tune with a NB. It will take much longer, but IMO it will end up being accurate enough for smooth running and long as you allow the ECM to make EGO and AFV corrections afterwards (not locking it out).

IMO,YMMV,IIRC,LOL :eagerness:

If I understood osvaldo correctly, he would like to data log with all the original settings and sensors in place/functional, plus add wide band too.

 

[Cooter]

Ya he can do that using a single WB O2 in the stock rear location.

The WB O2 can be used for datalogging (CL off) and with it's NB emulator, can be used for normal CL operation as well.

 

[osvaldo]

Osvaldo welcome to the forum. I love Italy, very beautiful place. Excellent food, wine and very warm people.

You welcome!

First question. Computer extrapolates data in between the cells. Just depends on exact rpm and load. ECM is used to calculate through math equation interpolated value.

Ok, it is as I was feeling. I don't know how closed to each other are TP (°) and RMP values used in every calculation but they must be much closer than what we can see on the maps.

not 100% sure exactly how narrow band is used to measure the O2, but it does pretty good job of keeping air fuel ration where needed. I can watch wide band in real time as narrow band makes corrections, it’s pretty accurate. I don’t think it’s the reason for hit and miss mode.

ASAP I will try to log and see myself what's going on my engine. According to what I read narrowbands, are working moving ECM to constantly supply the engine with alternatively leaner and reacher mixture as its signal is too easily running away abruptly from 0.45 V (14.7 AFR).
I am wondering if the up and down AFR is not the cause of the slightly fluctuating idle and the differences between engine pushes, almost cycle by cycle, when in closed loop. But, it is also true that narrowband is used almost everywhere and usually you can't feel any misfire or so...

Thanks a lot

 

[osvaldo]

Welcome Osvaldo, no cats or small headers here in the USA just pure Buell fun.

Lucky guys, i did not know that.
Thanks for welcome me.

 

[osvaldo]

One slight correction, those numbers are injector pulse widths. Each point equals to 58 microseconds. Say table calls for 56 points then injector will be open for 3248 microseconds or 3.248 milliseconds.

^^^ Yes. Pulse width, the time the injector is open. I'm just trying to keep it as simple as possible with the learning curve and language barrier

Yes, I was aware of that (from ECMspy guide). This is for DDFI and DDFI II, but I think it should be true even for DDFI III.

Thanks to both

 

[osvaldo]

If I understood osvaldo correctly, he would like to data log with all the original settings and sensors in place/functional, plus add wide band too.

Hi cossack84, Yes, you perfectly understood my point!
I saw that set up on a really nice job from Twin Motorcycle https://www.twinmotorcycles.nl/artikelen.asp?cid=7&aid=67
Unlikely it seems images have gone lost.
You can see something similar on https://www.twinmotorcycles.nl/artikelen.asp?cid=7&aid=135
But this was a Dyno set up and I have no idea which expensive software has been used.
As you catch I am wondering if it is possible to send trough ECM, BuellTooth, up to ECMdroid additional signal(s) from additional wideband sensor(s)

Ya he can do that using a single WB O2 in the stock rear location.
The WB O2 can be used for datalogging (CL off) and with it's NB emulator, can be used for normal CL operation as well.

It is not fully clear to me. Did you mean, running during CL with NB emulator signal (so you should not be able to data log any true AFR) and log the WB signal only during Open loop?
If so I think you should also keep always this set up because, at least in some guys opinion, it seems that if you fit back OEM sensor this can read in a different way than the emulator, so giving you troubles if you changed your maps. What do you think about?

Question aside: on EFI, in general, can't be a WB sensor used for a fully closed loop operation? For sure I am wrong but to me is looking so nice a system based on target AFRs, according to Load, rpm, acceleration, deceleration, and so on, with a constant feedback (and related EGO correction) from a WB sensor. Where am I wrong?

 

[osvaldo]

Yes, DDFI-3 has at least one extra pin for analog input.

Hi Cooter,

can you give any idea on which one is and how to use it?

 

[d_adams]

DDFI-3 can log the data via pins 1 and 33 on the gray connector. You'll have to add the wires to the harness by taking the connector apart and adding them in by hand. It's a bit of a pain to do, I've done it a couple of times.
That said, the ecm is limited and CANNOT (at this time and I doubt it ever will) use the wideband outputs to auto-adjust the ecm or run off of them. It (the ecm) will only use the narrowband signal.