big afr swings

[phongshader]

I think I've gotten my VE table pretty dialed in, however I get some pretty big fluctuation in afr when in a steady state, that is the throttle position, rpm, and map remains the same but the afr fluctuates from 13.6 to 16.5:1. I've disable AE, xtau, and set ego controller authority to 0. Any ideas to what would cause this? Here's an image of the mlv log and an msq if you want to take a look.
Thanks

[Matt Cramer]

If you could post the whole data log instead of a JPG, I'll be glad to have a look. The screen capture is missing a lot of traces I would like to check.

[phongshader]

Thanks for looking

[TheMonkey]

your voltage looks nice and steady, but your swings in AFR look like they are driven from MAP kind of erratic. in that same part of log where you first shared pic, your TPS is rock steady at 15 but the MAP range wiggles from 47 to 58, which wiggles your pulse widths from 1.4 to 1.8 which i think might be driving the AFR swings. where is your sensor plumbed in? what kind of intake, or how lumpy o cam do you have?

[phongshader]

this is the throttle body:
http://www.auto-nomics.com/cgi-bin/show ... 0850-002-D
with throttle body inserts to reduce it to 400cfm
on an Offy 360 intake
42lb injectors
The vacuum hose for the map sensor is attached to one of the tubes on the base of the TB and the MSII unit is about 24" from the TB under the dash.
The cam is 270° - 451 lift http://www.crower.com/cgi-bin/detail.cgi?prod_id=50231
all in 215 olds v8
http://usera.imagecave.com/phongshader/Alfa/619e.jpg
Can the fluctuation in the map really cause that big of a fluctuation in the afr? Can a 11% change in map really cause a 30% change in afr? Assuming that map is 0-100 and 47-58 is 11% and afr 10-20 or 10 between the 2 points so 13.6-16.5 is about 3 or 30%... I don't know if this is correct or not but it makes sense to me...but what do I really know...;-)

[TheMonkey]

hey, that's a nice looking motor. tough to fit in your car?

Can the fluctuation in the map really cause that big of a fluctuation in the afr? Can a 11% change in map really cause a 30% change in afr?

looks like map fluctuations are driving pulsewidth fluctuations, and from my experience a change in pulsewidths between 1.4ms & 1.8ms could easily fluctuate AFR that much (depending on what actual net opening time is).

so, i guess what i'm saying is that to me it looks like PW wiggles are causing the AFR swings. so what is causing PW wiggles? most likely it's MAP wiggling between 47 & 58, but there are other inputs, and this could be tail wagging the dog (AFR swings changing MAP). VE table looks smooth, but if it's slanting the wrong way at that point, it could cause oscillations.

the RPMs are not fluctuating with the AFR/MAP/PW fluctuations, so my first guess would be that MAP signal is a bit erratic, but your setup seems pretty conventional.

maybe Matt will have some other ideas. i'm not familiar with your size motor or what the specs of that cam might be like.... would you classify it in the nasty category, or well behaved?

maybe read up on the MAP sampling rates and lag factors to slow down the reactions to fluctuating MAP (if indeed your MAP is bouncy).

[phongshader]

Thanks
I had to use different headers and cut them up a bit to fit the engine in but other than that it wasn't too bad.
I wouldn't say the cam is extreme at all, I can get a fairly smooth idle @ 650 rpm
I'm really unsophisticated about all this but just looking at it, the afr isn't consistently changing with the map or the pw. Look at the time from 539.953>541.656 both the map and the pw are oscillating their respective amounts but the afr is fairly stable @ around 14.5 then from 541.656>543.437 it bumps up to 16.20 but neither the map or pw have changed, that is they are still oscillating in the same range. I guess to me it doesn't look as though the map and pw are driving the afr to fluctuate the way it is because the map/pw fluctuations are consistent, more or less, but the afr has much bigger fluctuations staying high then low then spiking high or low and there is no ego correction, AE, or xtau in the equation...I think, unless I'm missing something.
Thanks again for the help.

[phongshader]

Just to add to the to the confusion if you look at time 207.609>231.281, with the exception of one spike the afr is fairly consistent but the pw/map is oscillating in the same range as it is here 701.875>779.250 but the afr is much less stable.

[TheMonkey]

i agree entirely that the frequencies are not the same. BUT, in a steady state like that, injector pulsewidths should stay pretty constant. on my car, injector pulsewidth swings from 1.4 to 1.8 would have similar impact on AFR.

...Look at the time from 539.953>541.656 both the map and the pw are oscillating their respective amounts but the afr is fairly stable @ around 14.5 then from 541.656>543.437 it bumps up to 16.20 but neither the map or pw have changed, that is they are still oscillating in the same range...

just looking at that range, the ~average~ of the wiggling PW looks like it might be slightly higher than the following couple seconds of 16.2 AFR. but not by much.

<< 207.609>231.281 >> looks like the PW range is much less at about 1.4 - 1.5.

realizing that every car is different..... as a matter of reference, during a similar steady state with throttle steady at 15% my pulsewidths will remain within a range of a tenth ms for awhile (4.25 - 4.32).

[TheMonkey]

look how nice and steady your map and pulsewidths are at idle. should be similar at a steady cruise.

another quick unrelated comment... looks like your alternator maintains 13.7v at 1000 RPM and above. at 800 RPM idle, it drops to 12.4. could maybe swap a pulley to speed it up a bit at idle.

[41Willys]

I checked your MAPdot signal in your msl file. It's really all over the place. That little sucker will really amplify whatever your MAP does. It's the rate of change of MAP. I don't know if it's causing your problem, but I just put the fuel filter and orifice on mine, and it really smoothed out the MAP a lot. In my case, the MAP wasn't really showing it, but my MAPdot was starting to oscillate wildly at certain RPMs and was screwing up my accel enrichments. Just a thought.

Dan

[phongshader]

I don't think the mapdot is affecting AE or afr, but I could be wrong, because I have the threshold set too high and AE is essentially turned off

[gboezio]

Lean out is usually amplified by lean misfires, it will make the O2 content spike like this and add to the already lean situation, may try Hybrid Alpha-N at lower RPM, worked wonder on my bike, it idles like a clock now.

[phongshader]

so whats's happening is similar to what's described in this thread:
viewtopic.php?f=31&t=32609&p=200774&hil ... ut#p200606
It's only under lite throttle that the afr oscillates so much, under acceleration or more throttle/map the afr is fairly stable. I've been trying to find out about how to controll decel fuel cut (maybe wrong term) but I'm not understanding it. When there is no throttle the afr drops to around 18:1 and I get all kinds of back firing, under light throttle it oscillates as much as 12:1 to 18:1. In my afr table the highest afr is 15:1@ 20kpa, so why do I get afrs @ 18:1 all the time on decel and how do I smooth the transition between no throttle to lite throttle? Point me to somewhere in the manual that explains this.
Thanks

[phongshader]

After some more datalogging I can see the pattern. It appears that around 35-50kpa the afr get unstable and oscillates by as much as 13.7-18.1. Anything bellow 35-40 kPa the afr stays lean but stable and above about 50kpa the afr is fairly stable. What can I do to smooth out the afr in this area? I guess if I was to average the two extremes the afr would be right about what the target is in the afr table but I have the ego controller authority set to 0% so what would cause the oscillation only in this kPa range? Also why does the afr go so lean...18, even below 35-50 kPa, when I don't have anything that lean in my afr target table. Is this a function of decel fuel cut? I'm using MLV for tuning and datalogging with TunerStudio. I've read through the tuing docs but I must have missed something somewhere...if so point me in the right direction.
Thanks.

[phongshader]

Really ...no insight? Looking at my pw I'm seeing pw down to 1. Could this be part of the problem? My injectors are too big?

[phongshader]

Thanks Lance,

Then the EGO feedback isn't doing anything, and the AFR table isn't doing anything, so the fuel is determined entirely be the VE table and opening time, etc. As a result, you shouldn't expect the measured AFR to follow the table AFR.

Just a general comment - you shouldn't be trying to smooth the AFR feedback. That is putting too much faith in the EGO sensor, especially at low engine speeds. Instead, try to get the engine running as well as possible (smooth, minimum MAP kPa for a given throttle opening and speed, etc.)

Lance.

I realize that with the ego feedback turned off that the afr table won't affect the ms but
1. because ego control is turned off what would cause the oscillation if MS is only running off the ve table with no ego correction and only between 37-50kpa?
2. what I was referring to was that running the datalog through MLV, which I believe does reference the afr table, doesn't seem to affect the very lean afr in the low kPa. That is why does the afr lean out to 18:1 when there is nothing instructing it to. Is there another function, command, setting that affects this directly that I've missed or is it that MLV can't catch it or that it's not consistent so that if MLV fixes it in one instance that it's off in another therefore having to average the 2, 3, or 4 instances.

I'm not really trying to smooth out the afr but the afr is the feedback that I have. The afr oscillation is the symptom and I'm trying to find the cause. All of the tuning I've been doing has been done with no ego control, and running data logs through MLV. I wouldn't mind the oscillations IF it didn't affect the drivability of the car but in that 37-50kpa range the car bucks and surges as the afr swings between rich and lean.

Just one more question, maybe a stupid or obvious one but here it is, how would I tune out something like this that is a transient. what would be some of the techniques to do so...richen up the cells in that kPa range to eliminate the lean areas (would probably be way too rich)?
Add resolution in that area...
Thanks again

[phongshader]

Thanks Lance,
I guess I'll have rope a friend in to do a bunch of driving while I tune.
BTW I just got back from a 800 mile vintage car rally and the engine/MS ran flawlessly with the exception of the light throttle fluctuations, unfortunately the laptop didn't so I wasn't able to take advantage of the time on the road to tune.
Thanks again

[phongshader]

I replaced the 42# injectors with 36# injectors and changed the throttle linkage to progressive linkage on the 4 barrel throttlebody , now I don't get the wild oscilations in afr on lite throttle just a lean spike as the secondaries open. I haven't used map based AE but that seems as if that would be the best way to tune out that lean spike. What do you think?

[fordfanboi]

I have this exact issue. It's always when cruising at 50-40 MAP and it will oscillate my AFR 2 pts. MapDOT is oscillating wildly. My AE is TPS only. What the heck is causing my MapDOT to go crazy. The MAP in the log looks fine - looks like anywhere else in the log - just the MapDOT is wigging out.

Any idea how was this solved in this thread?

Thanks!