Did some more work on my PROM today. I took the car over a couple mountain passes and datalogged the runs as well as did VE learns. My VE tables are really starting to come into shape now that I replaced my 02 sensor. Its a bosch narrow band $20 sensor and I've heard there are reliability problems with these but it's been good so far. I'm saving for a WB so I can do my AE.

Anyway, the datalog is attached. There are two key areas that I need help with.

First issue:

At 3:50 I am steadily climbing the grade. Car is in overdrive, 40% TPS, MAP is about 88 KPa, spinning about 1600 RPMs, and moving at about 53-55 mph. Then at 3:52 exactly, for some unknown reason, the engine enters PE mode. As soon as it enters PE mode the knock sensor starts picking up knocks. It knocks steadily until 3:56, at which point the car downshifted into third and the knocking stops. It stays in PE mode for a good while after the downshift with no problems/knock.

My questions about this issue are, why did the car go into PE mode? Why did it take 5 second (with the engine knocking) in PE mode before the transmission downshifted? How can I take out a few degrees of PE spark advance for the appropriate RPM/MAP range? Once the car downshifted and went up to like 2600 rpms the MAP dropped to around 79 KPa and the knock stopped so I only want to take timing out where I have to.

Second Issue

At 4:14 I reach the top of the mountain and start descending. This is obvious because the TPS drops to zero and the MAP drops like a rock. TCC unlocks and my RPMs drop to around 1200 as I "coast" down the hill. The problem is very obvious at 4:19. At this point the TPS is still zero (this problem ONLY occurs at 0% TPS) and the RPMs start bouncing between ~900 and ~1300. TPS is zero throughout the bouncing, and the IAC steps are steady at 27 throughout the bouncing, so it must be fuel related. Any idea what can cause this and how I can fix it? I get this problem not just coming down hills, but all day long any time I take my foot off the throttle in the city or the highway.

Any help with either of these two issues would be really appreciated!

Thanks
Andy

 

For the first issue, PE mode, the TPS% increased a tad (40% to 41%) at the point the ECM went into PE mode. Check this table for the PE mode TPS% threshold: 'PE - TPS% Enable threshold.' It is most likely close to that value.

Since it was only knocking in PE mode I would change the PE SA table. It is this one: 'PE - SA vs Gear/RPM.' With an auto only the gear 0 row is used. Reduce the SA at the RPM where the knock was taking place (around 1600 and then lower).

At a minimum the TCC unlocked. The TPS% increased a little (44%) and the MPH dropped a little. These two together can cause the TCC to unlock and the transmission to down shift.

The 2nd issue is as you mention from fueling. In this case it is a toss up between proportional gains and re-enabling async injector mode. The reason is that the PW's are small, the BPC is moderate (101), the INT is moving up & down. This area is tricky, and makes me want to create an open loop decel function. If the RPM wasn't so low I'd recommend DFCO.

The PW is swinging wide, so I'd try the proportional gains and INT delay first. In the 36-43 gms/sec (and lower) airflow range of the INT delay table (INT - Delay vs Airflow) increase the delay about 20%. Can use the TunerPro toolbox and multiply those values by 1.20

Then reduce the prop gain vs airflow in this table: 'PRP - Gain Multiplier vs Airflow.' Again is the same area as the INT table above.

The affect this has is to reduce the PW swing, and thereby the AFR swing. Slowing down the INTegrator allows the prop gains to work without additional corrections.

RBob.

 

Attached is Darkshot's #22 log file in a text readable format. Un zip it and it becomes #22.ebl, which is text and can be viewed in Notepad or Wordpad. This is for anyone that wants to follow along and see what is being talked about.

RBob.

 

You've somewhat lost me when you start talking about the INT, prop gains, and async, because I haven't read up on what any of those things are yet. However I know enough to follow your instructions and make the appropriate changes in the tables so I'll give it a shot. In the meantime can you give me, or point me in the direction of a good explanation on what async is, prop gains, INT, and that stuff? What do you mean by the PW "swinging wide"? I need to know a little bit more in order to fully understand what is happening to my engine under these conditions.


I've made the changes you suggested and took the car for a spin. I increased the range you specified by 20% like you said but I was not sure how much to decrease the prop gain vs airflow, so I reduced it by 20% (multiply by .8). Should this be decreased more?

After driving it the "bouncing" effect seemed somewhat diminished, however the RPMs still occasionally bounced on decel. The engine almost stalled upon coming to a stop once, I think the stall saver was the only thing that kept it running.

 

In the Tuning Guide Book there is a link named 'Fueling Logic (O2 swing points)
' under the '8063/'7747/'8746 section. That covers a good bit of the proportional, INT, and O2 stuff.

>> the PW swinging wide

This is more the AFR is swinging wide, with a good swing on the PW. Look at the area of decel where the RPMs are bouncing. The O2 volts is hits the upper and lower extremes. Then look at the PW. There is a large varience in PW. This is from the combined efforts of the prop gains and the INT. Need to get the INT to stop interfering (by slowing down the updates), and then get the prop gains under control.

RBob.

 

Well after driving it too and from work I've concluded that those corrections had no effect on the problem. Should I reduce the prop gain vs airflow some more? Or increase the INT delay vs airflow some more?

 

Can you post a data log?

RBob.

 

Okay, finally got around to datalogging. Attached is a zip containing two logs.

In #31.dat, you can see the AFR swing occurring. I was cruising through the city, and went closed throttle @ 5:04. The AFR swing did not occur until 5:10. It continues until 5:22 at which point I gave the car some throttle and it went away. This delay is typical, I've noticed on more than one occasion. I take my foot off the gas and all is normal for several seconds, then the AFR swing starts.

The second datalog, #30funkyidle.dat, is a closed loop idle on the side of the road for a minute or two. I threw in a couple revs also. The problem in this case is the idle is very rough. It often idles like this when warmed up. If you watch the PW it randomly ranges from ~.9 - ~1.1. I've seen it jump up to 1.4 while idling. These changes are consistent with rough/choppy idling. You can see the RPMs change, coinciding with the PW change. Any ideas on this? The IAC steps are high I know. But sometimes the IAC steps are good at idle (~10). When they are good, the car does not idle crappy. What could cause this?

What is async, and could it help with these situations?

 

Oops. Forgot to attach it.

 

doh, ttt
tomorrow (sat) is my day off and i'll be fussing around with this before I start my worrk week

 

Good morning, I sent a BIN update to your email addy. It implements an open loop deceleration mode. This will correct the surging issue when slowing down. During open loop the proportional gains and INT are not used. Convert your current BIN then set the Option Word 3 bit 5 flag as outlined in the email.

Your idle issue is a little more difficult. You mentioned that it doesn't always have an idle issue, just sometimes it is rough. After a throottle blip the AFR hangs a tad rich with the result being the idle smooths out. A rich AFR is easier to light off, suggesting an ignition problem.

From the data log the battery voltage (from ignition switch) at the ECM is fluctuating. While the fuel pump voltage is steady as a rock.

That voltage fluctuation from the ignition will cause all sorts of weird behavior. It may just be that the ignition switch is wearing out. Or that there is a wiring problem. Also double check the spark plug wires, cap, and rotor.

RBob.

 

I've noticed some weird things involving that weird idle. Sometimes while idling if I activate different accessories the engine will respond with roughness. My cooling fan and AC are the worst. Often times when my cooling fan kicks on my engine damn near dies but recovers. Same with the AC system, once I actually did kill the engine when I turned on the AC at idle. Possible signs of a bad alternator?

 

I went for a nice spin today. Open loop deceleration works flawlessly, I love it. Not a hint of the AFR swing problem. Car comes to a stop so much better now.

I did catch slight, and I mean SUPER slight, like I would not have noticed it if I wasn't looking for it slight, hints of the problem on some part throttle situations, during which I could look over and watch the NB O2 voltage swing back and forth. Would further tweaking of the INT and prop gains round out the rest of the problem?

 

Two possible things that can cause the motor to want to die are spark and fuel. If the engine goes lean or doesnt have enough or too much spark under low speeds at high loads, itll want to die.

 

I'm familiar with those basics. It seemed to me that RBob was implying that alternator/battery voltage that varies too much too quickly can cause the ECM to behave strangely, leading to my rough idle. And nothing causes my voltage to drop more than activating certain high-amperage accessories like fan and AC system. So I am trying to determine if there is a problem in the regulation of my car's electrical system, or perhaps if my alternator is starting to fail, because I know its not normal for my car to act like this.

 

For the slight pronlem during park throttle, data log it and tap the space bar when it does it. The space bar taps will show up as green highlights in the Analysis screen. Go to that area of the data log, check that it isn't going into async mode.

Then check to see what the INT is doing. If it is bouncing up & down a little, I would adjust the prop gains (less), and possibly the INT delay (longer). This is at the gms/sec airflow as shown in the data log.

Have you disabled async mode at all?

{edit: forgot abobut the electrical} I don't think it is the alternator, although can never rule anything out. Play back the FunkyIdle log and watch the voltage gauges (far left). Note the BAT (ignition voltage) is jumpy, while the PMP (fuel pump voltage) is steady. With this I'd first go for a bad connection/switch in the igntion switch circuits.

RBob.

 

The A/C doesnt place alot of demand on the electric system (which you probably will have initially as the coolant fan comes on with the A/C), but teh compressor really drags on the motor. The coolant fan also has a large stalled rotor current, which causes alot of alternator drag when it starts up. I notice this on my car as well. Doesnt want to die, but it does cause teh idle to dip momentarily while the fan is starting. Definatly not a bad idea to check teh electrical system to be sure, but these two conditions can put the engine into an area it never sees with an automatic. Low speed, high load operation. If the calibration isnt right there, the motor can get stuck in that area and have a hard time recovering. Might also want to check and add momentary IAC counts for the A/C/cool fan coming on, assuming its an option.

Another check may be to rev the engine to arund 1500 RPM and then try the A/C/cool fan. If it still wants to take a dump, then it very well may be a fault with the system. If you have a DVM, measure the voltage at the alternator. It should be fairly steady at 13.5-14.5V. If it is, and teh same is true at the battery, then the fault probably is in the wiring or the ign. switch. If it fluctuates, and there are no loose conections, then another alternator may be in order.