So a customer drops his car off for an AEM tune

[futurevr4man]

Another shop had "tuned" it but it was running very poorly and could not get over 16psi with out breaking up and just over all running bad. Bad idle that is high, then low, then surges up and down, etc etc.

I will spend some time on it later, building a new map from scratch of course, I never trust another shops map and won't build off of it cause you never know what small detail you will miss on one of the buried graphs etc. But like looking into what was "tuned" at this point and why it may be running poorly.

But here is what I found in 5 minutes looking at it.

92 RTTT
Forged rods and pistons
Forged crank
JDPC TD05 kit
MHI EVOIIIs
Walbro pump
680 PTE
AEM Meth kit
boostaspark
etc

I like to shoot for pump gas (on TD05) 20 degrees at redline at 20psi. It's a good benchmark to kinda aim for.

This AEM was setup with 9 degrees at 20psi at redline. Wow, that's low.

Then I thought, Let me check the base timing. This needs to be set before anything else to ensure when the AEM puts out 15 degrees, the engine will actually be at 15 degrees.

Sure enough, when the car is at idle and the AEM is putting out 15 degrees, the actual mechanical timing on the car is.......5

So at 20psi at redline the car is somewhere around 0 to -1 degrees.

Knock control is set maxed out at 1.4 volts. Very conservative.
Oh wait, if it were turned on. Yep, it's unchecked, it's off.

Looked around for a few more minutes and found some other funny things. Can't wait to dig in and tune her up!

-Chris

maybe im having a brain fart, but how is the car only at 0 degrees or -1 at wot? what should the mechanical timing have been set to? seems to me, a mechanical timing of 5, working to 9, would end up at 9. what am i obviously missing?

[DG]

To simplify:

the crank sensor on the car tells the AEM whenever the engine is at TDC. The cam sensor tells the AEM whenever #1 in on the compression/power stroke. So by watching for when the cam and crank sensor fire close enough together, you know when #1 is at TDC on power stroke and now you can work out when to fire spark plugs and injectors (STAT SYNC).

But there is no guarentee that the crank sensor is physically synched to the engine at TDC. If, for example, the crank sensor is firing at 5 deg BTDC, then the AEM (which thinks that the crank sensor pulse is is zero BTDC) will be wrong by 5 degrees. So when your ignition map says "25" what you are actually getting is 30 - 25 degrees worth of ignition map, and 5 degrees of error.

This is why the second or third step of setting up an AEM is to zero out the ignition map and every other option that plays with ignition advance and check the physical timing with a timing light - to see if "AEM zero" and "engine zero" are the same.

Then, if they are not, you can either physically resynch the crank sensor (for engines where this is capable) so AEM zero and engine zero are the same, or you can set an option in the AEM that adds a global offset to virtually accomplish the same thing.

Not doing this is typical newbie/hack mistake.

The other, slightly more advanced thing that needs to be checked is to look at RPM drift. Sometimes there is lag or error in the crank sensor so what is perfect synch at idle is off by some amount at high RPM. So you need to check timing at idle, and again at 6000 or so, and see if zero is still zero. If it isn't, the AEM has a compensation function that can correct for it, or you can bake it into the map.

DG

[DG]

Here's an AEM (actually, ProEFI before Jason sold ProEFI to AEM - same software) story from when I was first learning to tune (on the Talon).

Car would come out of the hole like a rocket, and then just fall on its face. It would get louder, but not faster. Plugs were WHITE. O2 was OK, but that wasn't confirmed by the plugs, so I was suspicious.

Put in a fuel pressure gauge and an EGT. Fuel pressure is sagging a little bit, and EGTs are off the friggin' charts. 1800+ F

OK, car has 550s and a Supra pump, plus high-flow lines and a fuel rail. More than enough fuel flow. But pressure is dropping and car must be lean (white plugs, high EGTs). Supra pump draws amps like a MIG welder and I have the OEM wiring in place. So relocate battery to trunk and power fuel pump using a 4ga wire right off the battery, with 0 ga wire from distribution block in engine compartment to battery.

Pressure sag goes away, but car still falls on face. EGTs still high. Plugs still look like old bones.

Double-check timing offset - perfect. Throw a ton of fuel at it - no change. Knock raw looks fine - just the usual engine noise with no spikes. Car runs C16 so it really can't knock anyway.

Then for whatever reason, I log KNOCK instead of KNOCK RAW - and KNOCK is stupid high, and increases with engine RPM. Hm. Check knock noise floor - and the table is zeroes across the board. SHIT! Set Knock noise table to be the measured noise curve plus a little bit, and WHAMMO - car now pulls like a raped ape right to redline and EGTs drop to normal levels. What was going on? No knock noise floor meant that all signal from the knock sensor was being interpreted as "actual knock" and knock retard was pegged to full retard. By properly setting the knock noise floor, thwe ECU could differentiate between normal noise and actual knock, and stopped retarding timing.

Easiest 100 HP I ever made.

DG

[j2k4]

That ^^^ is why I'll likely never be a tuner...

[DG]

Here's another one from way back in the day before everybody had widebands.

I was using a Bosch "enhanced narrowband"; a narrowband O2 sensor that was supposed to have a usuable knee section on the rich side. So now you are looking for changes in a hundredth of a volt.

Car is dead nuts perfect at the little end of the track, but at the top end, sensor reads way lean. Throw more and more fuel at it, still reads lean. Plugs are black. Exhaust is full of soot. O2 says "lean".

Finally, spend $1200 on a Motec PLM wideband O2 controller - and it reads 12.5:1 at the little end, and 10.5:1 at the big end - stupid heat-sensitive lying narrowband O2!

That ^^^ is why I'll likely never be a tuner...

Oh, come on - this is just basic logic and understanding how internal combustion works. It sure ain't rocket surgery.

DG

[j2k4]

It's the reps, Dennis.

I'll make the car run well enough, then call it good.

Can't see a scenario where I'll get the kind of experience I used to get, or that you guys are currently getting.

[Ninja Performance]

OK, so you are a honda shop and don't know tuning a 3S (still dyno would show it needs more timing)

The base ECU timing was never set. At AEM 15 degrees the actual timing on the engine is 5. That isn't knowing 3S tuning, that is not knowing AEM EMS tuning. This is one of the first things you do.

Had to rewire the AEM WB as it was not hooked to the AEM EMS, although O2#1 input wire on the plug was cut within 1"of the plug and stripped bare just sitting there.

So got that done and a few other safety things like turning on Knock control etc. Had to lean out cruise and richen up in the positive load area to get it safe. Nothing else was done. No timing etc. Just to see what was going on.

Did a few base pulls watching boost, AFR, and knock volts. Got log, put it into virtual dyno. 14psi she is making 204HP. On MHI EVOIIIs

Time to eat lunch and get to tuning.

-Chrisbaserun14psi.jpg

[CoreyB]

Really enjoying this thread,. I love it when people share their trouble shooting experiences.

[Ninja Performance]

I have always tried to be as open and talkative about stuff like this all the way back in my IPO days on 3si. The more info out there the better.

-Chris

[mb3000]

OK, so you are a honda shop and don't know tuning a 3S (still dyno would show it needs more timing)

The base ECU timing was never set. At AEM 15 degrees the actual timing on the engine is 5. That isn't knowing 3S tuning, that is not knowing AEM EMS tuning. This is one of the first things you do.

Had to rewire the AEM WB as it was not hooked to the AEM EMS, although O2#1 input wire on the plug was cut within 1"of the plug and stripped bare just sitting there.

So got that done and a few other safety things like turning on Knock control etc. Had to lean out cruise and richen up in the positive load area to get it safe. Nothing else was done. No timing etc. Just to see what was going on.

Did a few base pulls watching boost, AFR, and knock volts. Got log, put it into virtual dyno. 14psi she is making 204HP. On MHI EVOIIIs

Time to eat lunch and get to tuning.

-Chrisbaserun14psi.jpg