So a bad tune will pop even a stock motor. Help me learn so I can boost. K thx...
 

I have been struggling with my tune for the past couple weeks. Just when I feel like I have started to get it right, I have a problem that puts it all in doubt.

I have heard over and over about the importance of having a good tune or else you will pop your motor. I am completely stock minus a lightweight flywheel at this point, and my MS tuning experience has been less than confidence inspiring. Help me figure out what exactly is happening so I can finish my NA tune, and help me learn what I am doing wrong so hopefully when I boost I can not make these mistakes again.

First my prior issues.

I had my AIT sensor installed in my airbox which as I sat got heat soaked and thus in turn heat soaked the sensor itself which threw off my tune. I did a long drive this past weekend of probably 200 miles and throughout the drive my tune seemed solid.

However, come Saturday at autocross, my tune was all screwed up. I relocated the AIT to near the intake of the air box and started re-tuning. I think I have eliminated that issue for now. However, I am having a problem diagnosing some problems I am having.

When I first start driving VEAL seems to be taking fuel out, but once I warm up, it seems like it is putting fuel into individual cells which is causing a bog.

Also, another thing that is happening is that after my tune is working fine for 15 or 20 minutes and VEAL is off, all of a sudden it seems like the same input on the throttle that had previously resulted in acceleration now results in the engine losing power. I am unsure wither this is a lean spot or a bog. It does go lean, but I don't know if this is a result of immediately going way too rich a moment prior.

I did a log of some of my issues this morning on the drive in and I will post that along with my MSQ once I figure out why my tuning laptop can't import the log. (too big perhaps?)

Thanks!

For starters, you want the AIT to measure the temperature of the air that the engine is breathing. Try putting it in the elbow just before the throttle body. In your log, MAT is between 72.5 and 73.3F the entire time. That is definitely not what your engine is breathing.

I don't trust your wideband. Look at the nonsense around 56.3 seconds. RPM constant, MAP constant, TPS constant, injector PW constant, AFR going off the charts lean. Surely you know that can't be real. Look for grounding issues. VEAL doesn't stand a chance if your AFR is jibberish.

Here is something I was thinking... The MS3 is installed in parallel with the stock ecu for now.

The stock ecu still has control of the EGR.

Could the stock ecu be adding EGR air at just that moment?

I won't be home until late tonight so I can't disable that until tomorrow (unless I can disconnect that at the engine?)

That hot air would definitely throw off my AFRs no?

I dunno about grounding. 5v is somewhere near 20:1 so a grounding issue would show full rich not full lean correct?

Yes, EGR could really muck things up.

I wouldn't do any checking on the wideband until you have EGR disabled.

Savington has a nice block-off kit.

If you get to this point (see prior post about EGR), the concern is bad grounds. I.e., if your wideband and the ECU don't have the same ground or the grounds "float" relative to each other, the wideband can send 12.5:1 but the ECU might read 12:1 on moment and 14:1 the next.

For now, what will happen if I just unplug the EGR wires? It is a 4 wire model so I believe it is defaults to the closed position. Any pitfalls to doing this as a test?

I don't know about the EGR systems, but that could make sense. Maybe the stock computer opens the EGR valve fully for some reason and sucks too much exhaust in... Did you delete the MAF? I believe the MAF is needed for EGR calculations. Looking at these logs (on a screen I can actually see) it seems like something that isnt the megasquirt is causing this. The injectors are doing their thing and suddenly it just goes lean. This throws things way off, and it's far too lean for EGO correction to do anything about it.

But what doesn't makes sense, is that if oxygen-lacking exhaust gas gets thrown into the intake, wouldn't that cause things to go rich? I know very little about EGR...

EGR disconnected for the drive home. Seems to have helped but it's hard to say considering that the roads were wet and it was storming pretty bad. I couldn't really push it.

I did encounter the problem, but it's hard to say that it isn't my tune because I did tinker before hearing out the door.

I think this still has something to do with my IAT placement like was hypothesisized earlier in this thread. I am trying to figure out how I will mount it without cutting even more holes in my plastic piping.

I am half considering starting the turbo install, because at least then I will be mounting stuff in the location they were designed to be.

When I did my MS pre-turbo, I found that the stock intake had enough meat that I could drill a proper size hole and thread in the IAT. No issues -- just had to extend the IAT wires.

Your tune will be messed up until you do some further VEAL without the EGR diluting the mixture.

After driving to work this morning it is still hard to tell if disabling the EGR did anything at all. I definitely need to re-locate my AIT sensor though. I didn't run VEAL for the first half of my trip to work this morning but I did log and at that point I did run the VE Analyze function in MegaLogViewer and adjusted the map from there.

Total correction was approximately 100% during my drive in, and then 95% after I stopped to pick up some water at a gas station (and did the VE analyze in MLV). Once I started out driving again I noticed my Total Correction was between 94 and 96% and gradually rose to near 100% as I drove and I imagine my engine bay cooled. VEAL was running in "update controller mode" and kept fattening up my VE map during this time. This makes perfect sense because total correction was lowering the numbers but since actual intake temperatures couldn't have been changing much it was fattening up the map to account for what in essence was an error in the actual AIT vs indicated AIT.

The problem is that I REALLY don't want to cut into my stock plastic anymore than I already have. I have a cone air filter sitting around and I think I have some elbows and charge pipes I won't use so I will see about putting the AIT as close to the throttle body as I can.

Is this sound thinking?

I don't think we're cracking into the key issue here. IAT being off, which affects the VE table won't cause what you're experiencing. When I drove the car, it felt like it hit a limit of some kind when accelerating. I also heard something that sounded like knock in the heart of the torque curve. I don't know if it was knock, but you should make absolutely sure it's not knock. Check the plugs for signs of detonation, and/or run the stock ECU. If it's a vibration thing, it should make the sound with the stock ecu, right?

Regardless, that may not be related to the issue shown in these logs. The issue here shouldn't be related to VE. In steady state cruising, the car suddenly bogs. Looking at the logs, the injectors stay at the correct PW but suddenly the AFR shoots lean and the car bogs. This lean event happens before the car bogs and before the injectors change PW. Something is kicking in that you're not logging. It could be something controlled by the stock ECU (we've ruled out EGR... mostly...) Could it be VVT? Is it wired up properly? Are you able to log it? Can you try driving it with VVT unplugged? That should put the intake cam at full retard and it will be sluggish, but if the VVT is unplugged and you don't have these issues, then check the VVT controls. Don't change VVT controls or injector timing in TunerStudio yet, just unplug VVT. If the issue doesnt change, then try disabling the injector timing (which should advance/retard injection events based on intake cam advance/retard), and then disable VVT in TS too.

That same sound was there at the same RPMs with the stock ECU.

I won't discount the possibility of there being something else causing me problems. However, if I can't get a reliably accurate VE map running then it makes it very hard to move on to the next step troubleshooting wise. For example, do to the nature of my commute to work, I may only hit certain cells in the VE map when the engine is cold and correction is at 95%. This will fatten up the numbers in the nearby cells and cause me to flood out my cylinders when I come back to that same cell later when compensation is at 100%.

I have heard of people having problems with indicated lean points that were immediately preceded by extreme richness. Also, the MSQ and log posted above is still based on possible EGR issues.

I am just going to use the KISS method and fix my IAT sensor location and work on getting the best map I can. Once I get to that point I will lock my tune and start doing the things that have in the past triggered problems.

This does not mean I don't want to hear opinions or ideas - I just think that swinging in all directions will just confuse my troubleshooting process.

I made some progress but haven't solved my problem. I re-located the IAT sensor to just before the throttle and it did result in a BIG increase in indicated air temperatures especially when the car has been sitting for some time.

For the time being, I adjusted my MAT density curve so that from 50 degrees to 120 degrees there would be no air density correction (100%). My theory is that once I get the VE map close at my normal operating temperature which seems to be around 110 degrees, I will lock that VE map and start adjusting the air density correction at various air temperatures to figure out the curve that works for my engine. I realize this is somewhat of a fudge factor thing, and I am open to hearing suggestions on how better to compensate for heat soak of the intake piping.

However, I am still having my strange stumble problem. As was mentioned earlier in this thread, the pulse-widths seem to be steady, load steady, yet AFRs jump around at times and sometimes spike.

I did a short log of a segment of my commute this morning after my VE map started to feel relatively good and then all of a sudden I had that stumble problem. I glanced down at the TunerStudio dashboard and saw that my fan had just turned on. I wouldn't exactly say a "light went on in my head" because it has been suggested here and I have considered it before, but all the peices are coming together.

When things run great:
Cold
No A/C


When things go to hell:
Hot
Cooling Fan on
A/C
Raising and lowering windows

While the stock ECU didn't stumble like the megasquirt, I have always noticed that the headlights dim when raising and lowering the windows, as well as when the A/C activates.

Also, the indicated AFRs seem to be a RESULT of something else going on in the car and not just a sensor problem. The reason I say this is that at times my indicated AFRs will oscillate a full .7 yet load/map, pulsewidths, throttle position all are steady (unless I am mistaken as to how much variation can cause nearly a full point of swing).

So, to the wiring diagrams? If I was to follow the KISS principle while avoiding replacing expensive parts (since I don't have a spare wideband for example), what could be causing this. Once again, I don't believe it is on the Megasquirt end because the numbers coming out of it seem to be steady. I am open to suggestions though.

How, exactly, is your wideband currently wired? Which wideband is it? I'm especially interested in any wire called "ground".

Going off of memory here. Wideband is AEM Uego.

Power is directly off of battery (fused and on relay).

Ground is tied to a body ground on the shift boot. I am willing to move this, but what is a good ground spot? I am afraid of using the megasquirt for a ground point because if I take the megasquirt out the wideband no longer will work, and unless I am mistaken having a wideband installed but not powered up is a big no-no.

In log *06.40.13.msl, I found three instances of bogging. Look at the following:

4052.9s
4064.9s
4148.5s

In each of these cases, "bit 7" becomes Y, your PW makes a step increase, and you go pig rich (to the bottom of the WBO2's reading ability).

What is "bit 7" in your instalation? This will be the key to your issue.

Other than that, things look pretty good. I didn't see the craziness that was in your earlier log (likely from the EGR).

From your logs, power looks good.

ALL ECU related grounds should be grounded to the engine block. If your WBO2 has separate heater and signal grounds (it should), then tie both to the engine block but at separate places. Tie the MS ground and the WBO2 signal ground to the same engine block ground location.

However, like I said earlier, "bit 7" seems to be the key here.

Looks like bit 7 is accel enrichment.

I am trying to figure that out now. Being that I have a VTPS (Variable Throttle Position Sensor) I figure I should run TPSdot for the most part correct?

Unfortunately, all of the Accel enrichment stuff I find online refers to a previous version of megatune/tunerstudio etc.

The new software uses "Time-based Accel Enrichment" page.

Unclear on how it works.

UPDATE: Researching this - but if you have some special expertise I would appreciate it.
Is there some way I can turn off ALL accel enrichment while tuning? Should I be doing that? VEAL accounts for that theoretically correct? I imagine this would also have something to do with A/C and fan problems (more load from alternator, more duty cycles)?

From: MS2/Extra Tuning guide

I am going to try this on the ride home. I am wondering what the drivability changes will be though.

EDIT:

From: http://www.diyautotune.com/tech_arti...ent_tuning.htm
This leads me to believe that the change in MAP was causing me to go into AE when the throttle was still. I changed the values in the accel enrichments page to be 100% TPS, which should completely remove the Bit7, Bit8 nonsense.

Yes, if you have a VTPS, use 100% TPS accel enrichment. MAP signals can have a bit of noise and it can be enough to trigger an accel event.

In general, to tune accel enrichment, start with litttle to none and add just enough to get over any lean stumbles you detect. I think you'll be surprised how little you need once the VE table is in good shape.

By the way, I feel like I have some dead spots in my most recently posted tune. Dead as in so far rich or lean that VEAL can't properly compensate. Can you guys take a look at the overall shape and let me know if it is reasonably close?

Also, why are my fuel loads higher numerically (richer) than most maps I have seen? DIYautotune's base map is lower (starts in the 30s I think), Y8s tune which is what I used as base for my VE map starts in the same area. The VE map wizard wants to start me off with a much more lean map. What is going on here?

Can't read your msq. Is it relatively smooth, or does it have big dips and peaks?

Don't worry about the VE values. They are non-dimensional and work in conjunction with your "Req. Fuel" (multiplied together). There is a maximum value (255 (8-bit) for MS1, not sure about MS3), but as long as you are comfortably below that, you're OK. Having larger numbers is actually better because your resolution becomes finer.

Well, some of them looked out of place compared to the nearby "terrain."

I freehanded the map to lower the peaks and raise the dips. I will just be gentle with it when I drive and see what happens. I can't imagine it would be too far off, if the value is bigger than the one on the left, and lower than the one on the right, I imagine it has to be in the right ballpark right?

VE tuning is an iterative process. I usually just keep doing it until the changes being made by VEA are kind of in the noise range. There will be some areas of the map where VE will decrease with RPM. This has to do with intake runner length and getting to RPM's where flow is starting to choke. But there shouldn't be any really big dips or peaks -- just smooth those out a bit.

New log. Same crap. (actually, after looking at the log it doesn't look too bad, but it felt worse than it looks on the log, in my opinion)

I am going to work on my grounds tomorrow, but this is starting to tick me off.

AE is off (correction, was set to 250, I was getting some AE - I changed settings to 2500). EGR unplugged. Air intake temperature sensor located to just prior to butterfly.

I wish I didn't have to work tomorrow because I really want to lick this.

While grounds will be fixed for EGR and maybe even ECU, I can't see how bad grounds on the wideband can cause it to be this bad.

However, it definitely DOES seem to get worse when any of the fans are on, if I roll up or down the window, or when I turn my headlights on.

Any ideas?

With ALL correction off - No AE, no EGO, no nothing - car ran as good as it has in a long time.

Once indicated intake air temperatures got about 10 degrees above ambient I started using VEAL. I tickled each cell and let it do its thing. Whole map smoothed out, and it I was pretty happy with it. I stopped off at a gas station on the way in and burned my tune, saved the file and went inside.

Now that I had a map that I was pretty happy with, I decided to mess with stuff that I know had been giving me fits.

I did a short log attached below that shows what happens when I do stuff like turn my headlights on, turn my a/c on, etc. As you can see, it really screws with stuff.

I am half wondering if the ground from the "engine ground" tie-point (I heard one exists correct?) on the fender to the engine itself is corroded and failing. I have seen that on stock cars before where stuff gets all goofy and it turns out to be the braided ground cable.

I turned on VEAL after the gas station and by now the engine was getting pretty hot. VEAL started fattening up my map significantly with the fans on, which made it bog sputter and nearly stop. I turned off VEAL and reverted to the map I was running when I stopped at the gas station and drove the rest of the way to work on that map. Everything was great. I have a log of the entire drive but it is too big to attach here and nothing is all that interesting.

So, either it has something to do with the alternator load being so high that it throws everything off, or a ground issue on the wideband, ECU, or both.

The ground strap between the engine and chassis bolts to underneath the brake booster. That area is corrosion-prone. Easy to check and clean.

Nonetheless, you want your ECU and WBO2 grounds tied to the engine block.

You might need to mess around with coolant-related air density, although I would definitely do the grounds first.

Yesterday stuff took a bit longer than I expected. I hope I will finish this weekend, especially considering that I have to work today and tomorrow. (overtime though!)

So this is what I found. I had paper notes on my build that I couldn't find, so I pulled my jumper harness and reverse engineered the DB25 that I had installed to handle stuff like boost solenoid, wideband in, etc. I created a spreadsheet specifically for that connector to keep it straight.

Turns out I had run all five "good engine ground" lines to the DB25 from the 3x, but only two on the main connector. From there, NONE of the "Engine Ground" wires made it to the engine. The only "engine ground" connections I was using (through the stock harness) were ECM ground, PCM ground(?), and fuel injector ground (aha!?)

Since I was sharing grounds with the stock computer and I don't know how good the connections are, this seems like a good candidate for being the cause of my problems. If nothing else, it is the way it should have been done from the start I imagine.

While I am at it, I plan on running the wires for sequential ignition and pigtailing them and also bringing the wires through for the boost solenoid, the boost sensor, and oil temperature sensor. I will also be wrapping them and placing them in convoluted tubing to make it pretty.

Finished up my project today. Ran four leads to the back ground point, and two leads to the front ground point. Basically it was three from the MS3 and three from the 3x board. I also moved my O2 sensor to it's own engine ground with a quick disconnect so I can easily add additional gauges just by using a new connector. I will try and take pics tomorrow, but I also spent the time to wrap the bundle in convoluted piping so it looks all nice, and I also added a quick disconnect for both the air intake sensor and the EBC solenoid which isn't installed yet. Easier to set up the wiring in advance and it looks pretty slick. I would have ran pigtails for the coils (COP or LS2, haven't decided yet) but Radioshack was out of suitable wire and I was on my last spool. Matter of fact I burned my last spool finishing this project.

I then went out for a drive and... I have my reliable and fun car back! Ran very well, no hiccups at all. I have EGO and accell enrichment off for the time being. I even turned on the A/C just to force a load on the engine and aside from the physical load on the engine, no issues at all.

I do unfortunately have a couple pending issues but I will move that to my build thread.

Thanks for the help!

I hate grounds... They can introduce so many gremlins...glad you got it worked out though? Sounds like you're doing the wiring right! My diy wiring skills need work...