Since I got my car up and running, I've been using TunerproRT and a Moates APU1 to emulate and datalog. The basic setup is an 8.5:1 CR 340CI engine, 24# injectors, Victor-type intake, Crane 216/228/112 0.454/0.480 cam, 2.02/1.60 iron heads pocket ported, manual trans. In carbed format, she made 260hp /292lbft at the rear wheels. I have come across two situations that I need some help sorting out.

First, I was seeing a saw-tooth pattern on INT as I would try to hold the engine at a steady RPM... 128 dropping to 90, then back up again, repeat. I take this to mean it's tending rich, and the ECM is trying to lean the mixture. Watching the WBO2, it's rich (13-14:1) when at 128, then the mix gets very lean (16-18:1), engine stumbles, back to 128, engine speeds up, repeat.

Second, and this has happened on the last couple of logs, the INT will basically sit on 128 for a long time, regardless of RPM or MAP, or just bounce along in the 124-128 range while the WBO2 swings from 13-18:1.

I'm using an LM-1 to provide an analog NBO2 output to the ECM, and it seems like it's doing what it should. I'm using the output from the datalogs to find times when the INT is fairly stable in the range that MAP and RPM are stable and adjust the VE tables with the INT numbers, but I'm losing confidence in the INT reporting. If the BLM's sit on 90 long enough, do the INT numbers just quit meaning as much? I tried clearing out the ECM which brought INT and BLM back to 128, but it wasn't long until it was doing the same things again.

Clair

 

Your LM1 is NOT doing what it should, for the NBO2 simulation.

When it "sawtooths" it's going to the limit, then setting a fault, resetting INT, repeat.

With the LM1 connected correctly, your AFR should go from 14.3 to 15.

You probably connected the 0-5V linear output, or the LM1 is programmed very wrong. You can reprogram your LM1 to have both outputs be 0-1 V 14.2-15.2 AFR, or whatever the other 0-1V is. Then it won't matter which wire is connected.

Edit: When you see 18:1 and INT is 90, you KNOW that INT is very wrong. Don't adjust your VE to a 'wrong' INT. Only adjust VE to INT/BLM if your O2 is swinging up down up down about 1-2 cycles per second, and the AFR reads an average of 14.6 and stays within about 14.2-15.2. Otherwise something's not right.

 

THanks for the reply, RG. I'll pull the LM1 tomorrow and see what I've got it set on, looks like I can't just check my previous settings through the software without having the unit physically connected to the computer.

FWIW, the AFR is only at 18:1 or so when it bottoms out, and it takes ~2-3 seconds or so for it to do so. The bulk of my tuning for the last couple weekends (maybe 3 hours total) has been trying to track the AFR rather than INT, so I don't think I've made any wild changes at this point. I'm really trying to get my idle sorted out well since I think that will give me some decent hints on how the lower end of the VE table will tie in to the upper end. My WOT limit is based on DD2K, more or less verified by the dyno pulls I made with the carb.

Clair

 

On checking the LM1's settings, analog 1 was set for 0.103V @ 1.022 lambda, and 1.100V @ 0.958 lambda. Analog 2 was different, 1.002V @ 0.680 lambda, and 2.004V @ 1.360 lambda. Those are the factory defaults AFAIK, with the second one being set up for DVM readout to see 10:1 to 20:1 AFR. I've got them both set for 0-1V now, so I guess I'll see if this makes a change. Not sure how long it will be before I get a chance to start the car again... weather has been miserable, and monkey boy goes to bed early.

Clair

 

Update...
Was able to spend an hour or so with the car on Sunday, and the sawtooth situation was still there, so I pulled in to a parking lot and tinkered for a bit. What I ended up doing was ADDING fuel to the VE table, mostly because I just wasn't comfortable taking any more out, and that seemed to really calm things down. I found this by trying to hold the engine at any given RPM, and it would pull the same crap by dying off coming back, etc. By adding about 12-15% to the cells where I was holding the RPM, it smoothed right out, and was able to sit at 14.7 by the ECM alone. I did this for a few different locations on the lower and upper tables, and then adjusted the tables to fit these new "norms" when I got home. It'll probably be the weekend before I get to go for another drive to test the new VE's.

I'm guessing that maybe it was lean enough to cause an occasional misfire which would spike the O2 towards rich, but I can't swear to that. The idle quality seemed pretty good until it got lean enough to stumble, and it doesn't seem right that a relatively brief moment of rich would push the INT to make it that lean. Still doesn't explain why the INT wouldn't go above 129 or so in any event. Hmmm...

Clair

 

If that Analog 2 was connected to the ECM, then the ECM looking for a swing past .5 volts or so, and never seeing anything below 1.002 volts would lean as far as it could, then set an error. Are you getting error codes ?