what a mess! help!
 

we had a dirty cas signal first and got that cleaned up by running a new shielded wire from the cas to the compy. now it seems the tune or wiring or cas or something is seriously wrong.... it seems like its just running on one rotor while driving at normal speed. we think thats from too much fuel. then when i give it throttle, like anything over 20% makes it ping. we tried to use my friends map and modify it for my na rotors with no luck so we started over with a na map and modified it for turbo with the same results... any ideas or help would be great! im sure he will see this post and can post up my .msq file. thanks guys

post datalogs too

Ok here is the msq. Unfortunately I don't have a datalog...

ok so i thought it might be a bad cas so i made a new one kind of. i didnt have an efficient way to cut 2 opposing teeth so i just swapped the magnets from a known working cas to the "bad cas". i made sure the gap was pretty tight between the magnets. and it didnt change a thing.. if anything it seems worse now... this is so frustrating.

Were you trying to disable closed-loop? The easiest way to do that is to change the step-size to 0.

You have your wheel decoder set up wrong. You don't have any missing teeth... In your settings you have:

"2nd trigger missing teeth" set to "missing teeth"

Set that to: "no missing"

That alone *could* be causing your problems.

You also have barometric compression turned on, but the max for baro is set at 10 kPa, effectively disabling it.

If you're stabbing the CAS as stock, your trigger angle and cranking angle are both wrong:

Trigger angle should be 65, and cranking angle should be 5 if you're using trigger-return, or with time-based you could set 0. Lower number tend to be better... If you want to use a 0 degree cranking angle with trigger return, just leave your trigger angle at 60 degrees, but change cranking angle to 0, and adjust the CAS until the timing is right.

Your lower two rows in the split table have negative numbers, which will either disable split, or rail it at 0 degrees (I'd have to review the code to be sure... ). You should probably set those to higher numbers.

Generally on a boosted engine, I set the staging parameters quite a bit differently from what you have... First, I set the RPM setpoint higher... around 4000 rpms with a delta of 300 rpms (so 40 and 3 in the dialog boxes)... second I stage on one parameter instead of both... that way if it starts boosting OR goes above the RPM it'll stage... last, I typically use a staging delay of 20-30. That helps smooth out the transition by gradually bringing in the secondaries and scaling down the primaries. There are many different ways to get staging to work well though.

The setting that's probably causing your problem is the second trigger missing teeth setting though. I'm not sure what that code will do in the case that it doesn't find any missing teeth.

Ken

i looked at your msq.. and noticed a couple of things.. but dont trust what i say.. i a really dont know crap.. anyway... when it opens i checked engine constants. it says.. 5. seems a little off. mine reads 9.4 or something also when i went to wheel decoder settings i read it that your running th 12-1 tooth setup, but then it says the 2nd trigger is on? i dont know if this is right? maybe should be 12 with 1 missing teeth and second trigger off. idk i would look into these tho.

guess i was a little late.

ok thanks guys.. all that info is very helpful but i dont know what it means and really cant utilize it till dpf gets back down here. im sure he knows what you guys are talking about

the cas is a modified cas. i took out the magnets from the upper wheel. the one with just 2 teeth and the lower wheel has 2 opposing teeth cut off. i stabbed the cas 3 teeth to the right of the missing tooth and i set timing based on the pulley mark.

ahh, I didn't know you modded the CAS... if that's the case, turn second trigger off... and just use missing.

Ken

jaa, I guess there was missing information. but yeah, modded cas. and the trigger angle is set as faq was a while back, and I leave cranking between -5 and -10 degrees. works like a charm and starts warm or cold on my car without problems or pushing the gas. and I don't have a bac hooked up at all. I will monkey with some stuff today.

Oh, and the staging is fine, I lowered it because he is high compression turbo with 550 primary and 720 secondaries.


dpf22

OK, with 550 primaries your req_fuel is a bit low too then...

Setting cranking timing between -5 and -10 while in trigger return mode doesn't actually change the timing. All you're changing is when it calculates dwell to start... and basically you're telling it to dwell later than it should, in fact, you're telling it to dwell AFTER the spark, which is probably leading to a condition that is causing dwell to be VERY long (dwell for a particular event is happening AFTER that event's spark, but since there's only a single output for leading, it just ends up dwelling REALLY long).

If you want longer dwell, set cranking timing to 0 (that's where spark is happening with your settings anyway), and increase cranking dwell from 6ms to something higher.

The way you've got staging now, if you hit 115 kPa without being above 2800 rpms (I don't know if that's possible), it won't stage until you hit 2800 rpms, even above 115kPa.... instead of making it so it stages only when both parameters are met, it would be safer to stage on when EITHER of them is met. Also, the gradual transition will help prevent a lean-spike during staging.

Ken

Actually, this has been runing for a few months great on my car with the settings as is. I may monkey with the cranking advance, and the staging delay is something I have changed also, but I prefer it to stage with both because it allows me to run a more consistent map in the top end of the fuel. While logging, I never see a lean spike while shifting and in high load/rpm. And you have obviously seen the fuel map bulge around staging to omit the lean spike while under boost and transitioning through rpm. I don't have any problems this way, though staging delay has allowed me to reduce the bulge while tuning today.

In any case, my car is constantly being tweaked a little at a time. BTW Muy, when I turned the kpa up on the barometric correction, it COMPLETELY fubard my fuel after I restarted when the engine cooled down. when I reloaded an msq with the baro correction off, it fixed the extremely lean condition i was having. It wanted to idle at 15:1 afr. Wasn't happening for very long. also made the entire map leaner than necessary.
Oh, and the required fuel is that low because I like smaller movements while tuning fuel.


dpf22

Whatever you like on your car. That's what the staging delay was added for.

Well, if you want to turn off baro correction, there's an easier way to do it... just turn it off in the constants page. It just seemed weird to me that it was on, but set to a value that would cause it to never do anything.

Fair enough, I've done the same thing, I just thought that the VE looked a bit lean in spots even though your req_fuel was already pretty low.

Ken

yeah, I guess with baro correction and the required fuel that low, it makes sense it would be leaner. I am still learning the ins an outs of the program little by little. My car runs awesome with little tweaks here and there needing to be done, I actually concentrate on drivability, then power as a secondary. Funny enough with just a front mount intercooler and not even an exhaust on 6psi on an s5 turbo motor, I can beat my friend in his s2000. He has a k&n and a titanium single exhaust.

I do have a question though. Can you explain a little what exactly the inputs on the baro correction screen do to the map. I was very confused when the car almost wouldn't run after puting in a number higher than 10. Is it just because my required fuel is soo low? or would it be something else?


dpf22

that is all on dpf's car. you know what my car is running and its failing... its very frustrating

so, did you switch the trailing plugs around and try driving to see if there was any difference like I told you, or are you still ignoring me on that. Its really hard to help you when you won't take my suggestions. I know its frustrating, but just going and shotgunning it is not going to fix the problem.

Muy, are there any possibilities of output problems with the megasquirt where the timing was too far advanced or retarded compared to the actual map. Are there any possible causes for that. Mechanical or otherwise?

dpf22

If timing has not been set properly, or the wheel decoder settings have not been changed as I asked, then sure there could be problems caused by the MS or mechanical problems.

If all the settings are right and the timing is set properly, then there shouldn't be any knocking or pinging.

Did you ever set the rows in the split table that were all negative to positive numbers?

If you suspect that the trailing plug wires are backwards, you can also disable trailing by putting a value of "21" in the split table where you want split turned off, or put "21" in for fixed split instead of -10.

Ken

i tried that. it made it worse. how am i not listening to your suggestions? i tore it down to the injectors to verify them, i made a new cas, i re timed it like 7 times... im doing everything you are telling me to.

kk, you didn't let me know you did it. I will be borrowing a dremel from lance to "modify this cas that I have from the other motor. Perhapse if it has a loose shaft or something stupid like that, then it will cause weird tach spikes mabe making it change timing drastically. Just a hunch, but we gotta search every angle.

Any suggestions muy?

dpf22

Try disabling trailing... have you verified, 100% that the timing is set correctly as well? Did you make the msq changes I suggested? It's unclear whether you did these things or not.

Ken

the dirty signal came back tonight so i dont know whats up. we havent changed anything with the map yet since dpf hasnt been down to make the changes.. how do we verify 100% that the timing is set correctly? i marked the timing when i rebuilt the motor so maybe i marked it wrong or something on the pulley

I always go by the little dimples on the crank pulley. the one on the right when they are visible from the top is 5deg atdc(after top dead center) otherwise known as -5degrees advanced. The one on the left is usually red, and is 20degrees advanced. Basically the stock computer put in 15degrees negative split at idle to smooth the idle out.

Let me know if that doesn't make sense. This is basically my interperetation of the fsm on the timing marks

dpf22

i verified spark again. its dead on and steady now.. the jumping tach came back but its all good below 2500rpm or so. it idles at 800 solid as a rock too when it warms up

modding a new cas tonight. keep me posted on any new developments.

Ok muy, here is a rundown of what I have done so far today. I have made sure beyond a doubt that the crank angle sensor is stabbed correctly. However, no matter what I do, the noise that I see while revving the engine will not go away. It doesn't ever spike up, always spikes low. I have ran a new shielded wire as far away from anything that could induce voltage as possible. I tried grounding the shield, that didn't work. as of now, the sheild is not grounded. I am also pretty sure that this problem is not caused by the computer because I put my own in the car and saw the same exactly the same behavior. So based on that, I am sure that its not the computer. I have tried reloading the code just to make sure its not a fluke in the load, no dice. I am running out of ideas as to what could be causing this. Oh and I have tried making the adjustments listed before. no dice either. Unfortunately I don't have a way to upload a log, but while looking at the log, the spikes in the rpm always just go down towards zero if not ending at zero.

dpf22