XFI not going to target AFR while cruising?

[Blazer406]

Trying to tweak the XFI on a friends stage II powered IROC while riding around on the streets of BG....


Senario: Cruising around steady state not accelerating... or decelerating.... say 40 mph....

Car feels lean.... kinda buuuu ba buuuu.ba...buuuuuu


On the base VE table.... Actual AFR showing most of the time 15.94:1 .... and o2 correction -3% .... it is still pulling out the fuel???? Look at the target AFR tables.... and the targets are in the mid to high 12's......

What gives? Why would it not go to the target AFR? It is definately in closed loop... because the o2 correction number is pulling 3% fuel....

It has me stumped....

 

[ijames]

I don't know about the target afr part, but remember if you are getting actual misfires they will fool the o2 sensor into thinking it is even leaner than it is. Never trust an o2 sensor on a motor with a miss.

 

[Blazer406]

I don't know about the target afr part, but remember if you are getting actual misfires they will fool the o2 sensor into thinking it is even leaner than it is. Never trust an o2 sensor on a motor with a miss.

With that in mind..... wouldn't that actually have the ECM over-correcting and making the actual real AFR rich?

This thing is showing lean... AND continuing to pull out fuel.......????

 

[Blazer406]

Anybody else with any theories?

This thing should correct the AFR to achieve the target.... within the specified correction limits..... it appears to be going the wrong way while cruising.

It has me stumped.

 

[Craig Smith]

There can and will be short periods during AFR swings where correction appears to be backwards in a given frame of data. The PID loops that control closed loop correction have to be a little on the slow side to prevent major oscillations in fuel delivery. My guess is that if you look at your log, it is not correcting the wrong way for an extended period of time. Just a guess, though... please tell me what you think of this theory.

 

[Blazer406]

There can and will be short periods during AFR swings where correction appears to be backwards in a given frame of data. The PID loops that control closed loop correction have to be a little on the slow side to prevent major oscillations in fuel delivery. My guess is that if you look at your log, it is not correcting the wrong way for an extended period of time. Just a guess, though... please tell me what you think of this theory.

I'm not referring to a datalog..... just riding at a constant speed observing the base VE table..... where you can see the actual AFR.... and the O2 correction on the dashboard for that screen. This is occurring continuously..... AFR showing 15.94:1 .... o2 correction -3.1%. We finally just kept bumping up the values in the ve table while cruising.... so that it was rich enough to not buck because of the lean condition..... I don't want to do it that way.... but have no choice until we get this figured out....

 

[Craig Smith]

Does it ever NOT show 15.94:1?

What is the UEGOS value in the main dashboard?

 

[Blazer406]

Does it ever NOT show 15.94:1?

What is the UEGOS value in the main dashboard?

Yes... it shows 12.x most of the time... just sometimes when cruising... it jumps up to the 15.94 value... and just stays there... with the O2 correction pulling 3.1% fuel out.

I'll have to check on the UEGOS value. I have never paid particular attention to that. I am just learning the FAST.

I did see Cal playing with the setting where you tell it what type of O2 sensor it has.... it didn't seem to make any difference. What should it be set on for the stock sensor that came with the XFI when it was new...?

 

[Craig Smith]

What are your closed loop high/low RPM settings?

Read a sensor called BPW in the main dash. That's base pulsewidth. What is it (roughly) when the problem occurs?

What is your injector opening time set to?

What size injectors?

 

[Blazer406]

1000/800 High/Low Closed Loop RPM settings

Injector opening time .6 mS (Cal had us tweak this until the lower left cell in the base ve table was around 30)

I'll check the base pulse width when that occurs

160# injectors

 

[Blazer406]

One interesting note. The injector size is set to 150#/hr. I asked Cal about this... he said it was fairly normal to list them in the software like that.

 

[Dusty Bradford]

One interesting note. The injector size is set to 150#/hr. I asked Cal about this... he said it was fairly normal to list them in the software like that.

The 160's are pretty tricky. We found they idle better with higher than normal base pressure. 47-48# Maybe the spray pattern is better at higher pressures. I have the injector size set to 145#

Set the target afr at 12.5 or so at cruising and idle. The VE table looks a little wierd but it's the only way I could get cruising and idle. The cells on the VE table where the car idles most of the time are normal. The line above it however had to be raised more than usual to maintain a steady idle with the 160's. For example

M
A 44 46 46 47
P 41 41 42 43


RPM

Watching the car, it was doing similiar to what you describe. The car would idle, pulse width was around 1.4-1.6%. The afr would shoot quickly to 15.9 and the car would stumble. Moving the VE #'s this way fixed the problem.

 

[Craig Smith]

Also, which injectors are they? The blue tops IIRC were more controllable at the low end than the brown tops.

What I am concerned about is that you are at the very bottom end of the dynamic range of your injectors, hence my line of questioning. Once the pulsewidth gets small enough, it just can't open and close fast enough to fuel the engine consistently and properly.

I don't mean to contradict Dusty's words of wisdom and experience, but I believe there is another way to see if this is impacting you, and that is by lowering fuel pressure and raising the fuel injector constant. This should lengthen the pulsewidth and at least somewhat alleviate the problem. Try lowering it 10 PSI and lower your current fuel injector constant by 10%.

To complicate this a little bit, the phenomenon Dusty mentioned could come in to play when you do this. I think that such a short pulse from a huge injector will inherently have relatively poor atomization qualities, and lowering the operating pressure will only serve to worsen that. There may be a bit of a tradeoff here between increased pulsewidth and decreased atomization, as each one will have opposite effects on the quality of idle and driveability.

If you have an opportunity to give this a shot, please let us know what you find. Hope this helps.

EDIT: After re-reading my post, I realized it got away from the original issue... oops. I am not sure why you would be seeing what you are seeing unless there is some other parameter disabling closed loop operation. Something that I had on my wish list dating back to my early days at FAST was some sort of indicator in C-Com for closed loop. A virtual LED or something to that effect would have been helpful in getting to the bottom of many questions in a much quicker fashion. Wish I could tell you more right now.

 

[Dusty Bradford]

We think alike Craig. At first I tried going to 35# base pressure to try and get more pulse width. The problem got worse and I could not get the surging out of the car. It was surging worse than before. When you would remove the vacuum to the regulator the idle would improve. So we tried more pressure. 48# seemed to really help with the idle and the way the car behaved. My thinking was that the lower pressure was causing the fuel to squirt rather than mist.

Also, the car ran better with 1.0 for opening time.

You say the afr goes to 15.9 and stays there while the correction is 3%. How long does the afr stay at 15.9? The car must be running awful if it stays for a while.

Craig, in FAST, if the afr goes to 15.9 and stays there, will the correction also stay at 3% until it sees the afr move? The reason I ask. My work involves control loops. Our controllers will not respond if the actual reading reaches a limit range, 15.9 for example. The controller will hold at it's current value until it sees the afr reading respond or move. Is it possible FAST does the same.

 

[Craig Smith]

Dusty,

AFAIK the PID loop doesn't care if you are at the end of the range, but this may have changed with the XFI.

The injector opening time is a value that gets tacked on to the pulsewidth calculated by the speed/density algorithm. In theory, a pulsewidth equal to IOT + calculated PW = right on the money for what you are targeting. Increasing the IOT therefore increases the PW, which has the possible effect of moving the injector pulsewidth into the controllable range. In the end, it has the same effect as increasing the VE numbers or lowering the target AFR numbers - longer total PW.

Blazer406, hope this doesn't seem like a hijack. I'm hoping this will all help you understand what's going on inside there. If I get to talk to Lance in the next couple days I'll ask him if there's anything else that impacts closed loop operation other than what I currently know about.

 

[Blazer406]

Blazer406, hope this doesn't seem like a hijack. I'm hoping this will all help you understand what's going on inside there. If I get to talk to Lance in the next couple days I'll ask him if there's anything else that impacts closed loop operation other than what I currently know about.

Absolutely not.... I'm very happy to have guys that know there stuff trying to figure out what is happening here. I am not able to go play with the car daily.... so I probably won't get to go tinker until either this weekend... or maybe next week.

Would timing during cruise effect this lean condition? Our spark map shows around 36 deg at the highest vacuum... and say 3000 RPM's..... right about (on the map) where 65 mph would occur on the highway. Eric Marshall suggested potentially advancing this some would potentially lower EGT's....

I am wondering if the two issues are somehow related. Our EGT is showing 1200 deg traveling at say 65 mph.... steady state.... I'm just wondering if additional timing in this portion of the map will change the way the car needs fuel....??

Does this make any sense?