Analyzing a log file...

C

carbuff

New Member
Joined
Sep 30, 2001
Hi guys,

Me again. Today I made 4 dyno pulls. After each, I was playing with a few things trying to make the engine better. I adjusted timing a little, and fuel too... Now I'm going back to look at the log files, and I'm wondering if anyone can shed some insight on how to use them. I imagine others might like to hear this also.

If you want to look at my log files, see the files in this directory:

Log and GCT File

Now, I have 14 parameters in this log captured, and I think I have most all I need. I'm trying to look things like:

actual AF vs. target AF

O2 correction

injector DC %

Now, my real question is this. How far behind, in terms of frames, is the actual AF vs. the target AF? Obviously it takes some time for the adjustments that the ECU makes to propogate down to the WBO2 sensor. I would like to use this to adjust my VE tables once I get my fuel delivery problems sorted out.

In addition, when I do a dyno pull, would I be better to do it in open loop and just look at the actual AF to adjust the VE tables? For example, when you step on the throttle, you get an extra bump of gas from the MAP AE and TPS delta tables. This causes the O2 to read very rich, and it pulls out fuel. Then you have a period of time for that removed fuel to get back. It seems like the computer is chasing its tail in cases like this...

I guess what I'm asking for is advice from others who have been down this tuning road on what I should actually look at, what I should do for WOT tuning, or for all tuning for that matter... One thing I have tried to do is sitting in the garage, just open the throttle and look at the O2 at various places in the VE tables (the ones you can get to without load on the motor anyway) and try to get the VE entries to provide an actual AF that I like. Is this reasonable? And it only works for a few places in the table anyway...

I would love some feedback on all this rambling! Any and all thoughts appreciated as I start the hard learning curve...

Bryan
 
Engine specs

Small Block Chevy, 377ci, big bore motor. Solid roller cam (252/268, .672/.672, 114ls). Canfield aluminum heads (227cc intake ports, 2.100" intake valves, 1.600" exhaust valves, 306cfm intake port, 215cfm exhaust port at .700" lift). I'm running an Edelbrock Victor Jr. intake modified to accept the fuel injectors and rails. I'm using SVO 36# injectors and a 1000cfm Holley throttle body. I have a complete MSD ignition system (Billet distributor and 6AL ignition box). And finally, I have a FAST bank to bank setup with the wideband O2 sensor.

If I missed anything, please take a look at:

http://carbuff.homeip.net/engine.html

Thanx!
Bryan
 
Bryan, I looked at one of your log files, now I'm no expert so don't take me too seriously, but the log I looked at at about 5500 rpm, you are at 80 percent duty cycle on your injectors, and I've allways been told that's max duty cycle you would like to see, but you are seeing up to 130 percent duty cycle, which means you are going really lean, and you can see this in your log file. Looks like you need some bigger injectors first before you burn that thing down, and after lookin at the pics if your motor, it's a nice piece ya got there, so get some bigger injectors first and then get that thing tuned in. By the way Bryan, what ya gonna put that motor in? Chris.
 
Chris,

The motor is in my 1970 Camaro. Here's the web page:

The Hunk o' Metal

The duty cycle went way up because of losing fuel pressure (I have another post talking about that, sorry I didn't mention it again). I'm down to about 20psi above 5500 or so. :( So I'm fuel pump shopping at the moment...

I didn't know that 80% was the target max duty cycle on the injectors. Maybe I need to jump up from the current 36# injectors that I have... Hmm...

I'd still love to hear some more generic ideas about how to look at and use a log file for tuning. Surely some of you guys out there have done this once or twice... :)

Thanx!
Bryan
 
80% is fine, or at least it will be when you get a pump that can deliver the fuel you need. As long as you aren't going static and have some breathing room, high duty cycle is a good thing.
 
Carbuff-

The procedure is pretty much the same for all the FAST units. When you know that you have adequate fuel delivery, then the procedure goes like this-

Play back the run overlapped on the VE table (ask if you don't know how to overlap). Step through the run frame by frame, and see which VE values need to be corrected (as indicated by O2 correction).

At WOT there is very little lag time between actual and indicated O2 readings (maybe 0.2 seconds, if even that much). Just to be safe, your best bet is to pay particular attention to the portion of the run in top gear, where engine RPM's are changing relatively slowly. If you don't make it all the way through 3rd gear at WOT (and not many street cars do), then use the portion of the run from 2nd gear to correct that part of the VE table.

If you run in open loop, just make a judgement of actual A/F vs. target A/F instead of using O2 correction as your guide.

If you have a knock sensor, in theory you can do the same thing overlapping the timing table and seeing when there is knock. In practice you might want to take an additional 1-2 degrees out "just in case", you won't lose that much power.

Of course this assumes your mechanics are working properly (fuel pump etc). Get that fixed before you even bother trying to mess with any of your WOT VE values. You will know when your fuel system is sufficient by having your injector DC% under 80 under all conditions, and you are able tune to 0% (or negative) O2 correction.

Under conditions less than WOT, O2 response is slower, depending on how much airflow you have.

-Bob Cunningham
bobc@gnttype.org
 
Bob,

Thanx for the response. I understand everything you said, but maybe I didn't ask the correct question.

When the ECU is doing O2 correction, it takes a LOT of time it seems for the correction to go away. As an example, if you look at my log, you see a huge spike when the throttle is initially pressed. This is due to the AE from MAP and/or TPS. But the O2 sensors sees it as really rich, and starts pulling out fuel. So I have a big lean bias from the O2 correction.

As the RPMs start going up, the O2 correction slowly comes back down, but it is still there from the initial throttle press. The O2 correction oscillates around the desired point, but it's always catching up.

I understand what you are saying about WOT runs, but I wouldn't think the VE vs. O2 is that fast until the RPMs are up pretty high. ie: at 2000rpm and WOT, the air still isn't moving to the O2 sensor that quickly, like it is at 6000rpm. Does that make sense?

The O2 correction issue was one reason I asked if it would make more sense to do the dyno runs in open loop... That way the only thing affecting the O2 reading should be the VE entry after I get past the AE point. (btw, my AE may be off, but it's set to the same things as in the demo.gct file I believe...).

And I agree completely on the mechanicals. I will be installing a new fuel pump this weekend, and start tuning again next week. I'm just trying to understand what and how to do it before I get on the dyno again! :)

Thanx for the feedback!
Bryan
 
Jim,

Thanks, I did not pick up on that when I originally looked through your manual. Please don't forget to send me a zip file whenever you get your laptop out of the car. That is a very nice manual, and I would benefit from it greatly while in the car! :)

psikillz@hotmail.com

Clayton
 
There is a setting that controls how quickly the O2 sensor reacts to an error. I played with that a long time ago and forgotten about it... There are a P and D term somewhere in there that control how quickly the sensor reacts. If you don't think those terms are set well, then you probably want to do your tuning in open loop like you describe.

If you have that much of a spike you might want to cut back on your TPS AE and give it more MAP AE to compensate... just guessing. But that way you might get less of a spike in your indicated A/F, and therefore you won't run as lean in closed loop. Maybe a turbo guy can help a bit here (I run NA).

When you say that your O2 corr oscillates around the desired point, I assume it's osscilating around zero?

Once you're at WOT, even if still at a fairly low RPM, there is still so much airflow through the exhaust that the O2 will be very close to what's happening inside the motor.

I'm curious what fuel pump you end up with- I've had nothing but problems in that area.

-Bob Cunningham
bobc@gnttype.org
 
Play with the GAIN setting in the closed loop parameters. That will help your 02 sensor to respond faster if it's never catching up. Raise the gain a little. It's probably at 15, try it at 20. If it starts to constantly "overshoot" the target AF ratio, turn the gain down until it doesn't. Hope this helps.
 
You must log in or register to reply here.

Similar threads

D
runs with stock ecm, not with xfi 2.0
Replies
19
Views
499
DGC
D
8
Sd2 should I spend my money on a gnecu
Replies
12
Views
719
Mr.Spool
M
S
Volt Blaster vs Boost a pump
Replies
15
Views
1K
GNVYUS 1
GNVYUS 1
TireFryer
What s10 parts!????
Replies
3
Views
309
TurboBob
T
SlowGrandNational
Small exhaust leaks causing this much headache?
Replies
0
Views
64
SlowGrandNational
SlowGrandNational
Top