Having a cone filter mere inches from the MAF with no screen is a no-no.
Would have been better off putting it near the fuse box with a screen after that elbow coupler.

OP - Post up a datalog of your cranking/start-up. Include MAF voltage, injector PW, AFR, IAT, coolant temp, throttle position and RPM.

 

hmmm similar setup worked well for me - admittedly my maf is further away from the filter than that and inside the engine bay . I thought the setup he has would be better than inside the engine bay due to swirl effects from the turbo

 

Yes, jet air is blocked, and sealed well. I couldn't hold the rpms at 1200, the idle is to erratic in that rpm range.

Im aware the latencies are wrong. It only makes sence that MAF scaling isn't right. I'll change the intake setup today, and see what I get. What type of "screen" am I looking for? Basic window screen mesh stuff?

Will do, good sir'.

 

It shouldn't be blocked. It should be hooked up to a pre-turbo source of air with a check-valve.
Its necessary with stock injectors. Its extremely important with bigger than stock injectors.

Essentially. Get the steel stuff. The nylon tears and the aluminum isn't much better.
When you install it, don't over-work it and make it brittle.

 

I'll go to home Depot in a second and see what they have as far as mesh goes. I'll post a picture of the improvised intake once I get it mocked up.

I'm confused on the air pump thing now. I got the block off plate for the manifold from Racing beat when I ordered my flange. The physical vacuum source (the one that goes to a solenoid under intake manifold) needs to be tapped into pre-turbo source? That correct? I feel like I'm missing something. I've got a check valve to use from when I took off the V-Fad and all that jazz.

If your talking about the think that's bolted to the side of the lower intake manifold, then I haven't messed with it. I never did disconnect anything on the lower intake nor did I take off anything pertaining to the OMP. It's all still in place.

 

Air bleed is not the air pump.

Two different things.

Neither have anything to do with the OMP.

What did you do with the two vacuum lines that ran to the intake in the OE configuration?

 

/\ what he said .

Typically You have these vac lines to deal with on a turbo install .

Jet air : post maf/pre turbo with a check valve to prevent boost leak
OMP vac line : post maf / pre turbo
Oil filler neck/crankcase vent : hook up to catch can seperate from intake
Vac line that used to go to Air pump : sounds like you have this plugged so no problem

 

Ok here is how I have the intake setup now. Hopefully I'm going in the right direction.

Explanation: cone filter - pipe - 90* bend - Screen - Maf - turbo


Also reverted back to stock maf calibration and limit since my mafe tubing/intake is 3.375" inner diameter from the filter ot the turbo. Still idles like garbage.

As far as the lines on the stock intake go:

Jet air : is plumbed into the bottom side of the intake that I currently have, near the turbo, with a check valve.

OMP vacuum line: plumbed the same way (on the intake), without a check valve

Filler neck breather: goes down to the ports on the lower intake manifold. it was like this when I bought the car.

There is a third line on the intake as well- it appears to go down towards the injectors on the back side of the motor. Again, this is how it came when I bought the car. Not sure what exactly it controls.

Air pump vacuume line: Not connected, just laying around. It doesnt pull any vaccume there, but I still have a golf tee in it for reference.


Finally, here is a cold start datalog. It runs pig rich, then lean, pig rich then lean. All this while having the erratic up and down idle from about 750-2000.

Latencies are wrong, I know. I didnt change them back to "stock" from where they are now, which is bumped up a little bit, from what I remember. And is the OMP rate in the datalog even close? Looks WAAAY high, but this is the first time I'm really looking at it...

 

Seriously ? That is f'd up - can't imagine why someone would do that . Stock position was on on the accordian tube but most people use a catch can for various reasons . Try blocking those hoses and leave the filler neck open to atmos. and see what happens .


That sounds like the omp vac source line I mentioned above - which makes me wonder what the omp line you are refering to is ????

 

Mazda did it. That was their TSB fix for oil in the intake and is standard from 2005 and up.

 

oops . LOL

So that setup would have full vacuum in the crankcase ?

 

Looking at your log:

Your STFT is mostly negative and your MAF seems to be over-reporting for the given RPM, so I'd suggest that air is leaving the system after the turbo.

BTW - Take all the things I didn't list above out of your datalog. No need for knock retard or vehicle speed or ign timing.

We have a crankcase?

 

you didn't get one ?

 

[QUOTE=MazdaManiac;3473680]Looking at your log:

Your STFT is mostly negative and your MAF seems to be over-reporting for the given RPM, so I'd suggest that air is leaving the system after the turbo.

QUOTE]

So it's essentially a vacuum leak I'm hunting. I've looked everywhere for escaping air with no luck in finding anything.

 

Disconnect your BOV signal.

 

Tried that, no difference.

 

Have you verified that your BOV doesn't leak? I realize it's recirculated, but if it is leaking, the MAF will over-read. Since its a knock-off TXS BOV, I would expect it to leak.
Also, since you did your cold-side in steel, it is very easy to have pinholes in the welds.

You should cap your TB fitting and your turbo inlet and pressure test the system.

 

The cold side only has one weld. I'll check it tomorrow just to make sure. I'll also pressure test the motor again.

I did have a leak at the BOV originally, where I had missed a small portion of the flange while welding it on. Looks to be gone now. Passed pressure test and doesnt blow bubbles when water is sprayed on it any more. Going to double check that as well.

I'll cap of the recirc, and the BOV and see how it does. I thought the TXS bypass valves were pretty decent. Guess I should have picked up an HKS or similar unit, if this is the culprit.

Noticed a small exhaust leak tonight before I left the shop. Looks like the turbo flange gasket isn't seated just right. Will get that fixed up as well.

 

I count 4 just looking at that picture, two under pressure.

 

I re welded all the joints. Re pressure tested, and found nothing. Put a dab of RTV on the maf o-ring, and BOV o-ring just to make sure it had a good seal.

It's driveable, and the idle is better but still erratic. Not as bad, but still kinda funky. It idles fine and steady as a rock in open loop, but gets erratic when it goes into closed loop fueling. Looks as if my base map is off. Not sure. I'll have to get with my tuner and go from there I suppose. I don't know enough about the Cobb to really mess with much yet, although I'd like to get it straight before I give it to him.

Can any one please explain to me or point me to an article/explanation of how the stock fuel injector staging works? I am really confused with that. I found the chart that MM had posted in another thread, but don't really know how to put it to use. Still really green with the whole staged injector thing.

I'll try running it through a couple of drive cycles now to see if things settle and let the ECU relearn the parameters. Just drove it easy for about 20-30 minutes in stop and go stop light to stop light cruises. Crusing AFR looks to be good. Gets a little lean as you approach about 5 psi, so I haven't gone above that really. I'll give it several hours to cool down, and go drive it again. I'll report back.

Thanks for your help Brettus and MM! Looks like I'm getting closer. Just need a little more tweaking, rechecking, and patience. BTW, How's my intake setup looking? Should I had a screen post maf as well?

 

Yeah - don't do that.
The OE calibration won't accommodate more than a load of 90% over most of the rev range, which means anything over 3 PSI can potentially pop your motor.
Stay out of boost until you have a tune.

 

Just knowing that it works with the injector sizes you have is enough . So long as your scaling is correct you should be OK with the 380/760/840 . Although there are better combinations .
Sounds like you are getting close , let us know if it settles down after a few drive cycles .
I prefered the way you had the MAF originally , you may find swirl from the turbo will alter readings at higher level . A screen will help with that but I've seen screens tear apart so prefer not to have them in my intake .

 

Idle's still a little, eh, off. I cant get the maf scaling right either. It always idles between 7-8 g/s.

Gonna try something new with the intake tomorrow, as well as block off the BOV and recirc completely and see what happens.

 

a trick to get maf pretty close is to check your LTFT after several drive cycles and adjust your maf (up to about 40 g/s) by the same LTFT % .

IE If LTFT = -10% , adjust entire maf curve up to 40g/s by -10%

 

yeah i just did that tonight, brettus. no luck. gonna try a completely straight intake with screens to see if it makes a difference tomorrow. obviously this wont fit behind the bumper but it'll serve as a good test piece to eliminate the intake being a problem i suppose.