I suspect my Dyno tuner was using an O2 sensor that was reading rich.

Here is the situation:

1. Car starts and drives fine. I still have some tinkering to do with the idle and cold start, but am making progress.

2. Car has less power than with MS, but keep in mind the MS was yielding terrible gas mileage. This is clear from the way it drives, and the dyno chart which showed only 160 rear wheel horsepower. That feels about right as you drive it. It feels like a fairly strong naturally aspirated slant with a four barrel, and slightly bigger than stock cam.

3. I have not checked the gas mileage. I have not trusted it yet to drive it very far for very long. I am still building a relationship with the car again.

4. There are three O2 sensors in question, my wide band dash gauge, the ECM feed narrow band, or the dyno tuner's wide band. The wide band gauge I have shows lean. The narrow band ECM feed shows lean. The O2 targets in the fuel map are all set at around 14.8:1, and the closed loop operation is trying to correct all the time. I am not sure what it means, but on the correction scale, it shows lean, and I assume this means it is correcting a lean situation. There is a function in callmap, where you can click "auto cal" and it will automatically change the VE cell where it is currently operating to eliminate the need for correction. When I did this on one cell, it bumped the number in the cell from 4.5 to 6 or so, or about 30%. The thing is, it seemed to run rougher when I did this. So I don't know for sure where to head with this right now. Maybe getting the fuel corrected also calls for a timing change of some kind. But maybe the mixture is right, and my sensors are wrong. It is a logic problem of sorts here.

When I asked Fred which O2 sensor he used, he said he used his own when they dyno tuned it. It seems to me as if his O2 sensor was measuring rich. Do you have any ideas about how to figure this out?

It would be nice to have two different calibration files, one for economy, and one for power. Right now, I don't even know if this tuning will yield good economy, but I suspect it will be better than I was getting out of the MS. It couldn't be much worse than that was. Some years back I had a 2 barrel Stromberg carb off of a 65 318 which gave the car terrific power, but gas mileage dropped from 22 to 15. I wasn't willing to make that kind of sacrifice, so I took it off. But it put it in my mind that these slants really run pretty strong with richer mixtures, but no longer yield very good economy when tuned that way. My 360 gave as good economy as the slant with that Stromberg carb.

The bottom line is, I am having fun with the car. It starts better and idles much better than with the MS, and I have hopes of getting this figured out. I think I will read the MS Megamanual on line for some ideas about tuning in general. I threw my old Megamanual away, not realizing it might come in handy as a general strategy guide for later. Maybe I will print it out again. Also, I had the good fortune of talking to Eli at Accel, and he seems easy to talk to, very bright, and eager to help. So that is good news also.

Any ideas here about how to get this sorted out? How can I figure out the real AF ratio at this point? I could take it back to Fred, but his tech, Jason, is pretty sure everything he knows is right (NOT), and I was not encouraged by my conversations with him after tuning. He said some things that I knew for a fact were wrong, and kind of stupid, but he was not about to open his mind to any kind of challenge of his ideas. Any question I asked, he seemed to take as a direct challenge to his authority, knowledge, and manhood in general. So I don't want to go there. I am grateful they got the thing running, but I think their usefulness might be played out at this point. Anybody got a textbook they can recommend? I will give more details about timing and fuel maps as questions arise or comments are made here. Thanks for any and all help.
Sam

_________________

I do not know your EFI system, but I have a question or two. Why is your target A/F 14.8:1? You say the system uses a narrow band O2. A narrow band toggles at 14.64:1. You cannot use a NB in a feedback system to tune for anything but 14.64:1. So I do not understand what you are doing here.

Where are your oxygen sensors located? How far from the cylinder head? Are you running a catalytic converter? Where in the exhaust system did the dyno operator put his wide band?

Your EFI is a centralized FI or a port FI? If it is a wet flow design, then you have the additional task of having to deal with wetting of the runners. This is where fuel condenses onto the manifold and port wall surfaces.

With regard to the delimma about actual A/F, if you calibrate for 14.64:1 and your system actually achieves this, then you should see your narrow band continually toggling about 0.450 volts. In fact, if you use a DVOM that can average readings for you, then you should see the average voltage output from your NB O2 at or around 0.450 volts. This should agree with your wide band readings.

Mitch

Lean is good for cruising & gas milage but 14:81 or anywhere around here is to lean to make good power in most engines(depends on combustion chambers ability) . I would guess that around 11:50 to 12:50 is where you will find the power. But when going richer I think you might also need more timing to give time for the mix to burn because it will burn slower than the lean mix. So having the lean settings should be good in certain parts of the map. But as you go heavier on the throttle & more load taking it to richer mix with more timing should give you the power. Having 14:81 all the way across the board as a target doeasnt make much sense to me.

As far as the true calibration of your sensors compaired to the dyno sensors it depends alot on where the operator was putting his sensor & where your sensors are located. I guess having anouther wideband and putting it in the same place as yours is could get you a idea how yours are working.

When you used the auto cal to set the one cell & it changed the VE. You stated it ran rougher. It should after a bit of time run the same because before the change it is running the same A/F but having to correct it & after the change it should just get to where it is not having to correct it, but it may take a bit of running for it to make these ajustment and get leveled out back to the target A/F. So it may run off on the A/F because it is running on the corrected amount that it had been using and as you run it the ECM will make ajustments until it is back to target A/F. Then if it doesnt have to correct to reach this target you will know the VE table in that cell is correct then. How much time this would take I am not sure but you can tell by how much correction factor it shows for that cell, when it changes from what it was reading before and quits changing what it says it needs for correction then that is how long it takes. But no matter what the Target A/F is keep ajusting the VE table until it takes no corrections to maintain A/F, when you get it where the VE table is done the VE should not be ajusted agian. Only ajust other things and dont changes VE anymore when you achive this all the way across the table.


If I am remembering right you have a turbo is this correct. If so you do not want this 14.80 target when it is under power. That is way to lean. 11:50 to 12:00 is safe and 12:50 is kinda pushing it.

Jess

To give more information:
The A/F ratio is different across the operating range, I just thought I said too much already in the first post. The narrow band sensor is about 8 inches after the turbo. The wide band is about two feet below that, but before the split into duals. The set up is multipoint injection with injectors right at the end of the manifold runners.

The target AF ratios look something like this:

14.00 at idle, and high vacuum.
14.51 at high vacuum and higher RPM all the way to 5500.
13.76 at 77 KPA through all rpms
12.86 at 100 KPA (0 vacuum)through all rpms
up to 12.2 at full boost of 7 lbs through all rpms.

It changes gradually from 13.76 to 12.2.
In fact all targets change slowly and gradually from cell to cell.
Jess, thanks for your guidance here. I will procede in the way you described, Your logic seems solid. Since adjacent cells are averaged by the processor, maybe having one cell very different from the ones nest to it could account for the roughness as well. I think maybe one key here is to always save the present calibration as a new file, and keep the old one to go back to if things start to fall apart in some way.

_________________

This is not a reply about how to tune your system, but about your sensors. I have had problems with one wire sensors getting cold at idle and light cruise (20 mph) with a free flowing exhaust even with the sensor before the turbo. I would look into a three or four wire heated sensor before or even after you get this worked out, they really work much better and can help with cold running and low speed drivability. With the four wire you have a good direct ground connection to the ecm instead of through the exhaust. They go closed loop right away and stay that way. I would think about moving your ecm sensor to the inlet side of the turbo as close the header/manifold as possible. I don't know about the wide band, but maybe moving it up might help as well. Old or polluted sensors read leaner as their condition worsens, so the engine runs richer.

I run single wire O2 sensors in several modified vehicles with success. There are some caveats however. Bosch O2 sensors have a much shorter life in marginal situations. I prefer NTK or Delco(best).

I run the sensors in the collector of long tube headers in one application and have no trouble. The warmup for the sensors is far faster than the engine....you need a good cold start/run program and dont really need to go closed loop until at least partly warmed up.

Problems can occur on going closed loop to soon. My fave is when you get timing retard caused by startup blowby.

The learning process the ECM goes through during closed loop is far slower than you might think. It takes my 747 ECM half a tank to adjust from 89 octane blended(alcohol) fuel to straight 87 octane Western Canadian regular. The idle speed isnt usually set in the ECM until the vehicle has achieved 30 mph(you do have a VSS?)

I would manually work through your tables as Another 6 suggested before trying closed loop.

Whenever I start with a new tune( I use chips) I always run it with the ECM cleared and the O2 sensor disconnected to see just how close the chip tune is to optimal....then whem Im happy I hooker up and let the ECM fine tune.

Yer gettin there!

_________________
Yeah....Im the one who destroyed this rare, vintage automobile.....

I took it out tuning this afternoon, and the closed loop function did not seem to be responding, in spite of the fact that it is turned on. The sensor is giving a voltage that changes with the operating conditions, so that is not the fault. It seems to still be running really lean if I can believe the narrow band O2 sensor. It was reading .06 at idle, and would go up to .6 and .7 under acceleration. I did not notice if boost was in the picture at the time. Again, I do not know if I can trust the narrow band sensor. It is a four wire sensor, or hego as they call it. I think I might see if my local shop has a tail pipe sniffer to see if the 02 voltages correspond to what the tail pipe shiffer reads. Maybe it is not lean at all. I will also call Eccel's support line tomorrow and see if they can tell me why closed loop stopped working. It was fine yesterday.

_________________

I may be thinking wrong but if its to lean it may keep it from going closed loop, just not enough heat I think.


You Wideband is in a good location for widebands, you dont want them to get to hot it tends to kill them and they seem to heat up with ease on there own. You just about cant get a wideband to far back. I bet your wideband is telling you the true story. Maybe the narrow band O2 sensor is dieing or has a problem, it would acount for bad running also.

I think thats a good idea to go get a snifer hooked up to it to see what is going on. You need to know because your just chasing your tail if that narrow band sensor is not working right. Once you know if the wideband is working right then you can use it to go by and no more worrying with it.


I think Sandy may have a good piont about just running it open loop untill you get it tuned, with the wideband working & going off its results & open loop this would be alot easier to get VE tables set and all the rest. If the Accel was running off wideband that was working correctly it ajusting corrections may not be that bad but with the way things are going I think its hurting things. Get that snifer report so know if the wideband is doing correctly so you have something solid to go by. After all thats what that puppy is for. And if its telling the truth you know the narrow band is goofed so open loop is what you need until you get the narrow band fixed.


Aint it fun, just keep at it youll get it. Just takes awail.


Jess

For the price of a narrow band sensor you could just replace it. That would have to be cheaper and easier than going to a shop. If the old one is still good you have a spare.

How much blowby do you have?(Im thinking valve seals here) That can drive you nuts when working with O2 sensors.

_________________
Yeah....Im the one who destroyed this rare, vintage automobile.....

Hi Sandy, I think this motor is pretty tight. It only has maybe 10K on it, and it does not accumulate the blowby muck around the oil filler or PCV valve.

_________________

Remember that "narrow band" is a good term for that device. The midpoint voltage occurs at (or pretty darn close to) stoichiometric, but if you deviate toward lean by even the smallest amount, the voltage will plummet. Look at the typical voltage graphs and you'll see an incredible non-linearity... it was designed as a switching device, and for that it does an excellent job. For tuning a carburetor, knowing whether you're rich or lean is often a lot better than "seat of the pants" driveability feel. But for designing a fuel map on an EFI system, you're shooting in the dark.

In other words, seeing a low voltage in a narrow-band sensor will tell you that you are running lean, but won't tell you how far lean. Admittedly, it sure is easy to get scared while reading a dashboard gauge, when you forget that little fact...

- Erik

_________________
Lots of early Valiants and Barracudas have crossed my path.
Also a handful of other toys for variety now and then.

Thanks Eric. I am going to start a new thread on this.
Sam

_________________