Experimenting with Open Loop vs Speed Density
#1
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
Experimenting with Open Loop vs Speed Density
So last night I decided to test out how my car would run in open loop rather than regular speed density. I did some little re-tweaks on the VE so I was sure it was all in line before I started, and I then turned the closed loop enable temp to 250. The OLFA table was left stock to keep a little fuel cushion in there in case things didn't go as well as I expected.
Notes:
Commanded AFR was around 13.6 or so during normal driving.
Actual AFR ranged from 13.4 - 15.2, so AFR error percentage was obviously substantial.
The car felt very smooth and drove well, even with AFR being all over the place.
I leaned on the throttle while watching the WB, and it richened up as it should, maybe a little more than expected even.
After this experimental drive, I reset the closed loop enable and checked again to make sure things were still accurate, and it was all 0's and 1's just as it was before.
What would/could cause this variance since nothing was changed except the O2's were shut off, and is it normal for open loop driving?
Notes:
Commanded AFR was around 13.6 or so during normal driving.
Actual AFR ranged from 13.4 - 15.2, so AFR error percentage was obviously substantial.
The car felt very smooth and drove well, even with AFR being all over the place.
I leaned on the throttle while watching the WB, and it richened up as it should, maybe a little more than expected even.
After this experimental drive, I reset the closed loop enable and checked again to make sure things were still accurate, and it was all 0's and 1's just as it was before.
What would/could cause this variance since nothing was changed except the O2's were shut off, and is it normal for open loop driving?
#3
Kleeborp the Moderator™
iTrader: (11)
If you are running headers, it could be that your narrow band O2 sensors are a little on the flaky side. Hard to tell what to believe though...your wideband or those. If your wideband isn't calibrated correctly, that could be causing troubles in and of itself.
My open loop tune doesn't swing that much, so no, I would say it is NOT normal for what you saw to be happening. Then again, I don't even know if mine is correct. It becomes a game of what you trust most...
My open loop tune doesn't swing that much, so no, I would say it is NOT normal for what you saw to be happening. Then again, I don't even know if mine is correct. It becomes a game of what you trust most...
#4
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
Try and see whether there are distinct regions of high
error (e.g. decel, some band of MAP or RPM, etc.). If
you scatter-plot AFR vs the prime indices using Excel
it may show you where to focus, and you can decide
whether this is a "don't-care" or something worth a
tweak. Look for patterns, not outliers, in the data.
The min/max histograms are useful in their own way
but boiling down the data into pure statistics hides
the texture of things and just tells you "red".
error (e.g. decel, some band of MAP or RPM, etc.). If
you scatter-plot AFR vs the prime indices using Excel
it may show you where to focus, and you can decide
whether this is a "don't-care" or something worth a
tweak. Look for patterns, not outliers, in the data.
The min/max histograms are useful in their own way
but boiling down the data into pure statistics hides
the texture of things and just tells you "red".
#5
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
I've got a sneaking suspicion that my O2's ARE a little on the flakey side (got a DTC on both of them a time or two), but they seem to be switching just like they should. I was using the wideband AFR % to double check my VE being correct, I probably should have clarified that.
jimmy, I remember I tried filtering out FTC's to narrow down pertinent data. I think I filtered out FTC>5 and most of the leaner numbers were still present. I'll keep looking for some patterns and see what I can come up with. I might also throw some scans in here shortly as well.
jimmy, I remember I tried filtering out FTC's to narrow down pertinent data. I think I filtered out FTC>5 and most of the leaner numbers were still present. I'll keep looking for some patterns and see what I can come up with. I might also throw some scans in here shortly as well.
#6
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
Log files... Not sure how I managed to save the first log as a .hpt, but whatever. It plays back fine.
O2's turned off:
http://www.unrealhp.com/041506_O2_test1.hpt
O2's on:
http://www.unrealhp.com/041506_O2on_test2.hpl
If you look at the first one, the AFR Error % is all over the place. The variance makes it pretty apparent that something wasn't quite right, but I have no clue why fueling would be so different.
O2's turned off:
http://www.unrealhp.com/041506_O2_test1.hpt
O2's on:
http://www.unrealhp.com/041506_O2on_test2.hpl
If you look at the first one, the AFR Error % is all over the place. The variance makes it pretty apparent that something wasn't quite right, but I have no clue why fueling would be so different.
#7
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
I looked this over and it's pretty much a "cloud" with
respect to the prime suspects - only MAP shows any
kind of correlation at all, to the eye, and that's weak
at best. Nothing to do with RPM, airflow, fuel trim cell.
Oddly only 3 FTCs are hit at all, but I guess in open
loop the question is "why not just one?". Anyway, to
me it looks like maybe there is a big noise problem.
The MAP dependence, I wonder whether it's indicating
fuel pressure, injector table or something. But this whole
excercise has turned up diddly, for clues.
respect to the prime suspects - only MAP shows any
kind of correlation at all, to the eye, and that's weak
at best. Nothing to do with RPM, airflow, fuel trim cell.
Oddly only 3 FTCs are hit at all, but I guess in open
loop the question is "why not just one?". Anyway, to
me it looks like maybe there is a big noise problem.
The MAP dependence, I wonder whether it's indicating
fuel pressure, injector table or something. But this whole
excercise has turned up diddly, for clues.
Trending Topics
#8
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by jimmyblue
The MAP dependence, I wonder whether it's indicating
fuel pressure, injector table or something. But this whole
excercise has turned up diddly, for clues.
fuel pressure, injector table or something. But this whole
excercise has turned up diddly, for clues.
The only thing that could potentially be an issue (which I have pretty much decided is a faulty gauge) is my fuel pressure. After driving for a while (more than 15-20 minutes) the fuel pressure gauge shows that pressure slowly starts to bleed down to as low as 53-54psi unless I stab the gas a little, and it snaps back in to place. I'm 90% sure this is a problem with the FP sending unit though, as AFR's never reflect a leaner condition when the gauge shows a pressure drop.
The wierd thing is, during warm-up in OL the AFR's match up perfectly with the OLFA table commanded AFR.
#10
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
Why in the world would my voltage drop while I'm driving around? My voltage holds a little over 14 volts when I'm driving around, even with the amps running hard. I'm almost positive it has to do with heat soak on the FP sending unit, which is mounted to the FPR on the strut tower.
#11
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
Alternator regulator electronics drift the voltage output
with temperature. Fuel pump output pressure follows this,
is modeled somewhat but who can say, about accuracy?
You would think the sending unit would be temp compensated
but maybe check by dousing the case in cold water after
heated up, and see if this alone produces a response.
Presumably any system voltage concerns would only hit
the pump at the upper end of delivery and the regulator
should buck all of that until the pump runs out of steam.
Is the regulator manifold-referenced and the IFR table still
contains vacuum-based flow profile? Something like that?
That would account for MAP dependence at least.
with temperature. Fuel pump output pressure follows this,
is modeled somewhat but who can say, about accuracy?
You would think the sending unit would be temp compensated
but maybe check by dousing the case in cold water after
heated up, and see if this alone produces a response.
Presumably any system voltage concerns would only hit
the pump at the upper end of delivery and the regulator
should buck all of that until the pump runs out of steam.
Is the regulator manifold-referenced and the IFR table still
contains vacuum-based flow profile? Something like that?
That would account for MAP dependence at least.
#12
TECH Fanatic
Thread Starter
iTrader: (6)
Join Date: Jan 2002
Location: Springfield, IL
Posts: 1,194
Likes: 0
Received 0 Likes
on
0 Posts
Are you talking about my Commanded AFR?
I have the stock OLFA table, which for my operating temps around 170, they range from 1.02 at 20kPa to 1.13 at 100kPa, then there's PE in addition to that.
If not, my FPR has a vacuum port but I'm not using it. I've got it set to keep a steady 62psi.
I have the stock OLFA table, which for my operating temps around 170, they range from 1.02 at 20kPa to 1.13 at 100kPa, then there's PE in addition to that.
If not, my FPR has a vacuum port but I'm not using it. I've got it set to keep a steady 62psi.