had a ok tune now my AFRs are crazy without changing anythin

[Supernova_6969]

hi!
I bought a 280zx with a megasquirt II v3 (i think). the previous owner told me it used to run like a stock car before he took the engine apart and rebuilt it. after that, he never readjusted everything to get the tune right according to the new set up. he told me all I needed to do was to tune it and i'd have 250-300hp.

that was last year. when i bought it, the afr was ALL OVER; sometimes extremely rich, sometimes crazy lean. I worked on the tune by playing with the VE table for all of last summer. I managed to get something nice enough when i'd leave in the morning to work, but if I were to use it again during the day, I found that the AFRs were not quite right anymore. It felt like the fact that I was starting it hot was doing something wrong. using it only at 5pm, after 8 hours of sitting there, was fine.

The car was in the garage this summer, and I took her out again 2 months ago. I kept working on the tune, and I felt I was getting somewhere; I don't remember the weird AFRs on hot starts being that much of an issue (the weather was also colder though). I had stable-ish AFRs pretty much where I wanted them. I then started setting up the EGO, and that was ok. I'd get lean spikes after giving it gas, but nothing serious and I kept looking for a solution to this common issue. 2 weeks ago, I took out the IAT/MAT sensor, to check it out (it worked, I just wanted to Identify what type it was), put it back, and all was fine.

Then last week, i started the car one morning, all was fine. an hour later, I got back for another drive, and the AFRs are now on average too rich. we're talking 2-3 points below what I wanted (14.7 became 13s and 12s, and under boost, my low 11s became low 9s). and the AFRs were all of a sudden crazy unstable; cruising at 40 mph on flat ground with a steady gas pedal yields afrs that changed from 12.5 to 15.5, with spikes up to 17-18, with the surges that go with that.

since then, I tried changing some of the values to make up for this, but the AFRs don't seem very affected by changes on the VE table.

I'm stomped. I checked all the inputs (temp sensors, really), checked the ego (which works the way it's supposed too. turning it off does not help). I checked the x-tau and it's working fine (I actually set it up for the first time and it got rid of most of my lean spikes after acceleration, and most of the richness after I let go of the pedal, which is a bonus). I sprayed carb cleaner all over, and I don't seem to have a vacuum leak.

at this point, I'm thinking either something is wrong with the tune and I just don't know, of something mechanical broke. either way, I'm not certain of what to do. I've been suspecting that at least one of my injectors wasn't working well for a while now, but I don't get how one injector would affect the system so much (and create both lean AND rich issues).

anyone have any suggestions?

[BigLou240sx]

First thing - you are using B&G firmware. This forum is focused on support for MSEXTRA firmware. After you figure out this issue (or as part of your forthcoming troubleshooting), you may make the choice to use MSEXTRA, and these forums would be able to help out. Otherwise, you may getting better assistance on the B&G forums

General note - it would help aid in troubleshooting if you also included a datalog of normal operation, and a datalog of the issue you are describing.

Having a quick look through your msq and being unfamiliar with B&G firmware, I don't see anything too far off. But take that with a grain of salt. In any case, sudden drastic changes to running condition would force me to look at mechanical (engine mechanical, wiring integrity, etc). I know you said all of the temp sensors checked out ok - did you verify this with the ecu connected and viewing tunerstudio?

[Supernova_6969]

Hi lou.
Yeah, i know that im running b&g, but i didnt know it was that different. Id love to make the switch but im afraid i wouldnt know how to reprogram the new software properly. Im also afraid of breaking somwthing while doimg it..
Thanks for the input though.
As for the sensor data, thats what i meant when i said all the data was good... I ran the car with tunerstudio open and all the values were fine. Ill try to log a few miles on it and post it...
Thanks

[DaveEFI]

How are you measuring your AFR? According to your MSQ, you have a narrow band O2 sensor. You need to check this in both EGO control and Tools, calibrate AFR table. And turn off EGO control until you get a decent tune.
But as has been said, you need to use the B&G forum (MSefi) and the Mega manual there, as B&G code and Extra are very different.

[Supernova_6969]

oh crap.

I have a wide band that I hooked up myself. I had to calibrate it manually because the default voltages an innovate LC1 just didn't give me the right values in tuner studio (based on what the innovate gauge gave me; I suspect that it's doing this because I didn't ground the senor at the same place as the MS2 ecu. something to do eventually).

Thanks for pointing this out, Dave, I'll change it as soon as I get home. that would explain why the EGO was overcorrecting everything...

that being said, the car seems to respond better to changes in the table after 2 weeks of rinning erratically. I still don't know what is causing the issue and it's still running rich, but if I can fix it by re-tunning until I figure it out.

[lagos]

In order to have consistent fueling, you need to adjust the battery voltage compensation under the injector characteristics.
This is very easy to tune. All you have to do is disconnect the control cable from your alternator (the harness plug with the thin gauge wires, not your main battery supply cable) while the car is at idle. Then monitor your afr as the voltage starts to dip down into the 11v range. if your fuel starts to change as the voltage drops, adjust the compensation number so that you have the same afr as your did with the alternator running at 14v.

Injector dead times also play a role in consistent fueling, but not as much so as the battery voltage compensation setting.

[Supernova_6969]

hi guys,

I've been away for a while, and have been driving the car a bit lately. the AFR are still crazy.. some days are better, others, well (like today) it's like the car does whatever it wants.

I've recently checked the fuel pressure, as I suspected either the pump was faulty, or the regulator, and the fuel pressure is as steady (well relatively to vacuum/boost) as a rock. perfect. I'll be checking the injectors soon, although a faulty (dirty, broken) injector would give a pretty steady issue, not wild erratic variations, I'm guessing..

Lagos, I've seen something similar to what you mentioned in other places... it's a little bit above my knowledge level, but i'm willing to look into it. essentialy you want me to disconnect the control cable so that the alternator output becomes less stable, and then monitor the fuel levels.. if the fuel changes significantly as the voltage goes down, i'd have to change the "battery voltage correction".. I'm guessing this is a trial and error where I try a little at a time and see if it's better?
actually, can I just play with it, and see if it helps, without having to play with the alternator? as long as I can put the original figure back in if it doesn't help, right?
would that kind of issue create AFR that fluctuate multiple times per second, and by up to 2-3 AFR points (say from 13 to 14.5 then to 13.5, then 12.9, all within 1-2 seconds)?
potentially very related to this, I looked some stuff up, and from what I understand, the voltage that I can measure at the injector should be the same as the voltage that the megasquirt sees, right? I checked it out, measured between the bat negative and the injector positive (physical voltmeter)gave me 12.55v (key on, not running) but the megasquirt's gauge (computer) was telling me I have 12.00 volts. I turned the lights on and injector voltage would go down to say, 12.03 while the MSII gauge was telling me 11.5 or something. essentially, both registered a very similar drop in voltage (with the MSII loosing a little less than the injector, i'd say), but it seems the MS is not getting the same voltage as the injectors. is this related to the battery voltage compensation, or is it a different issue? and could that be contributing to the problem at hand?

as for the injector dead time, I'll keep that for another day, I think... one thing at a time.

in any case , here is a screen grab of the kind of fluctuation I was talking about (this is at iddle, data for about 15 seconds, after having driven the car for 45 minutes...afr goes from 13 to 16, rpm from 869 to 940).

erratic afr iddle.jpg

[turbo conversion]

I have not tuned B&G but check to see if this is an option in TS.

You need to adjust battery voltage correction so TS and and volt meter at battery read the same.

To do this go under tools > battery voltage correction and raise or lower the supply voltage at zero ADV count(V)
until TS and volt meter match. Do this with key on engine off.

David

[nathanhardy]

With a brief look at this tread I notice you said your grounding was different then ms for your wife band.. I had all kinds of issues till I did all my grounding correct this is something that should not be over looked!

[Supernova_6969]

Yeah, i re-did the ground from the wideband directly to the ms2 cover just 2 days ago. Now i am able to set the voltage settings to what they are supposed to be and the ms2 sees the right afr...
However, i think the ms2 is grounded by being screwed into the metal kickpanel.... Not all that securely li might add. I'm tempted to verify if a ground wire runs to the engine (a wire IS booted to the intake but im not certain it goes to the ms2. I'll have to check).
Unfortunately, grounding the O2 sensor didn't help any, i still have significant afr variation, on top or sometimes getting weird, lean afrs for hours, u til it gets normal again..

[nathanhardy]

Check the Manuel for the proper grounding once in my know it all days I tried to be smarter then a grounding loop and really had some funky issue, also I don't think the ms circuit grounds to the case it self gotta do it throw a db pin, also as a safety I don't think it can handle the O2 heater ground but I'm not 100% on that wb

[Supernova_6969]

I didnt do the set up myself, and though it was done well. I still think it is but i'm probably better double checking everything.. Ill start with the ground..... Good suggestion
The wiring loom is a mess and is right accross the engine, without enough length for me to move it... Id love to redo it, but im not confident and too lazy...

[nathanhardy]

I grounded all sensors back to ms then ran all my grounding tails to one large ground back to the engine about a foot away from my ms I removed some insulation on the large wire and soldered my ground for my wb there. Then ran a ground strap to the battery on the other side of the motor.

[nathanhardy]

Don't let wiring get you just take some time and do it right, I'm lucky cause of a little shop I know that sell wiring supply's dirt cheap. If you need help just ask! Gl

[Supernova_6969]

Thanks guys.
One thing though, to all of you about the wiring... Did you guys experience problems like mine that were caused by faulty wiring?
Like i said i think it's all done well... I dont mind going through it, but not just as a shot in the dark..

[turbo conversion]

[quote="Supernova_6969"]Thanks guys.
One thing though, to all of you about the wiring... Did you guys experience problems like mine that were caused by faulty wiring?
Like i said i think it's all done well... I dont mind going through it, but not just as a shot in the dark..[/quote

Faulty wring or wired incorrectly can cause all kinds of issues, I would check it.



David

[nathanhardy]

I would check anything that you have changed like I may have misunderstood but it sounded like you just put in a wb that grounding is just as important as a coolant temp sensor... I would say one of the reason ms get a bad rep is because of installer error and bad grounding I as well had so many issues before I redid my setup correctly granted I dthought I knew more and cause my own issues!! I'd check anything that could have changed and I'd look for grounding noise just to make sure it shows up as jumping spikes or at least mine did.. If there no noise in the signal it not a issue.. Not broken don't fix it. But I'd give everything a quick look. Better safe then sorry... Right.

[lagos]

Lagos, I've seen something similar to what you mentioned in other places... it's a little bit above my knowledge level, but i'm willing to look into it. essentialy you want me to disconnect the control cable so that the alternator output becomes less stable, and then monitor the fuel levels.. if the fuel changes significantly as the voltage goes down, i'd have to change the "battery voltage correction".. I'm guessing this is a trial and error where I try a little at a time and see if it's better?
actually, can I just play with it, and see if it helps, without having to play with the alternator? as long as I can put the original figure back in if it doesn't help, right?

You want to disconnect that cable to see how the injectors behave at a different voltage. Normally with the alternator connected, the injectors will be powered by 13-14v. With the alternator control cable disconnected you will be operating them at 11-12v.

So...
1. get a stable afr at idle with the alternator connected.
2. disconnect the alternator control cable.
3. check the afr as see if it changed.
4. adjust the compensation number up or down until you get back to the same afr that you started with.

By adjusting the injectors at these two points, you will create a proper slope for the injectors to reference. This will not only allow for correct fueling with small changes in voltage, but also allow things like your MAT correction and barometric correction tables to correctly account for changes in weather.

Its really simple to do, and I'm surprised that this information is not on the front page of the megamanual.

[Supernova_6969]

Hi lagos. Thanks for the explanation. I understand the idea now..
Seems like a thing to do except for that fact that i can't get my afr to be stable under any circumstances to begin with. A few posts up, i posted a image with a 15 sec graph of my afr at iddle. If you sneak a peak at it, you'll see that i have no stable reference point to adjust the voltage from. And all that up and down on the grap is under a constant voltage....
I was really excited by your idea, but i think its a solution for a different problem, and i'm not there just yet.. As a matter of fact, one thing i've noticed since i read your post is that the voltage falls when i'm accelerating hard, but that the afr does not get lean (it's currently tuned to be ultra-rich, due to poor afr stability, to prevent leaning out on a low spike).
HOWEVER, all that talk made me realise that i should really make the jump to msextra.... I'll possibly look into that in the next few weeks. Maybe it'll even fix my wild afr issue...

[BigLou240sx]

Definitely think it's a good move to migrate to extra code, but you may allow someone to shed some light on the sudden running change if you post the actual log file (of the screenshot you posted on wednesday) along with the running msq.