Ok, I've had it. My brain is borderline fried trying to figure out how to pass the OBDII plug-in test while still making real power. I want my cake and to roll around naked in it, too. And no, swapping everything to a 95 or earlier is not on the table...yet...

In short: 2001 Miata, FM/BEGi bastardized kit, running link piggyback (soon to be burned). I'm in Houston where the county, and surrounding ones, do the OBDII plug-in test and two-speed idle tail pipe test. Oh, and I'm not a tooner, so speak louder and slower for me. Also, no visual inspection/CARB EO crap that I'm aware of.

I hate the link piggyback. No idea WTF it's doing, no way to monitor it without spending bucks on all kinds of data logging equipment, which I'd rather put towards go-fast parts, and, the shop that sells Link stuff on the other side of town won't touch it. Oh, and I was told not to go over 9psi with it. Suckville.

After getting the car back on the road following 6 months of basic maintenance and trying to get poorly thought out products to fit, the car blew up on me the first time I got into boost. I was running the pre-loaded 6psi map. Half of the #3 piston is on vacation in the oil pan. Coolant all drained out with the oil. Convenient, but not good. More to follow as I do a post-motordom. Also, built bottom end on the way (stock CR, I think).

Anyway (finally getting to the point), I want to make power and pass the annual emissions testing. Since piggybacks suck, standalone is the way to go. Since standalones don't return OBDII codes, I'll have to convert the car to "stock-ish" to pass the test.

In an effort to minimize my, uh, effort, to change things annually (and still have good parts on the car), here's the work list I've come up with:
- Swap injectors
- Remove/Replace turbo/outlet with fabricated "stand-in" pipe
- Swap ECU
- Cap turbo oil feed/return lines
- "Jumper" turbo coolant lines
- Reconnect MAF/other crap standalone don't care about
- Drive around until emissions ready status is set

Things to make my life easier:
- ATI or SuperMiata crank damper using 4-tooth crank trigger
- MS3 w/ adapter harness or AEM EMS-4
- Have a local shop build a pipe that connects the exhaust manifold to the down pipe, thereby replacing the 2560/FM outlet pipe

To clarify about this "pipe", the idea is I wouldn't have to remove the exhaust manifold (BEGi cast) because, that just sucks. No wrench clearance on 3 of the studs. Also, the divided FM outlet doesn't seem to allow the wastegate to fully open, so I'd be really concerned about boost creep, even with it wired open.

Now, questions:
What do you think? Crack-pipe or good plan?
I know the 4 tooth trigger wheel means less resolution, but what does that mean practically?
What else am I missing in this thought experiment?

 

I have to do the same here every 2 years. My plan is to put back the stock injectors/IAT sensor/MAF, disconnect the wastegate actuator and wire the wastegate open, and make sure the primary o2 sensor is upstream of the cat and secondary o2 is downstream of the cat.

In theory that's all that should be required to pass. The ecu doesn't care that the turbo is on there, as long as you stay out of boost. Around here they only do part throttle sniffer testing, no WOT for any duration. The only other possible issue I can see is the turbo absorbing a lot of exhaust heat, so make sure you drive for a good long sustained time so the cat can heat up to proper "light-off" status before the test (maybe 20-30 mins of sustained highway driving).

 

Is a halfway decent piggy back not an option? Either adaptronic or MS wired in parallel should do the trick, so you can have your cake and roll around in it. However, I'm driving a 94 with no emissions to worry about, so this may just be my somewhat naive thought.

 

I will observe that datalogging equipment *is* "go-fast parts", IMHO, but I agree that the Link piggyback is a long way from being an ideal solution.

Swap injectors, plug in stock ECU, wire open wastegate and drive like a grandma for a day is the core of it. You might also need to swap the stock cat back in (depending on the test #s required by your state and the kind of high-flow cat you put in). If you installed some other kind of coils, then you might need to swap those back too, and think about uninstalling the wideband if the ECU change results in it not being powered.

--Ian

 

Ian, yes, datalogging stuff is definitely go-fast parts. However, most (if not all) standalones have this included to some degree, which would make it redundant.

Jonathan, I've looked a little at doing the parallel ecu setup. Unfortunately everyone I've seen has abandoned it. Reverant told me I'd be leaving a lot on the table trying to do this. I think the main issue is the standalone would need to keep all the sensor parameters in line with the stock ECU's expectations, which ends up limiting your window of tuning.

Ideally, someone would have a signal modifier to adjust what the stock ecu "sees". Wire it up in parallel with your stand alone, adjust the signals to what the stock ecu wants, then drive away happy. Not exactly legal, though.

 

Megastim with rpm input (TPS maybe needed to keep the ODB-II info "happy")?
All error code generating senors must be "simulated" though (at least long enough to not trigger new ones during the inspection).

Just for theoretical speculation, of course...

 

Theoretical, of course! For learning and developing a more emissions compliant car!

Hmm, I wonder if this or the JimStim can be used to "test" a stock ecu.

 

Sounds like what you want is a parallel install of a standalone and then a proper tune.

The only downside is having to keep the maf. BUT if you wanted to get really crafty you could use a output from the ecu to give the ecu a maf signal. You could also do the same thing to give it a post cat o2 signal if you didnt have a sniffer test.

 

Run ms3 in parallel and use an o2 simulator for the secondary o2 sensor. Mount your wideband where the secondary o2 sensor was, then be happy.

 

Is the AEM F/IC not an option? There is a guy on mx5nutz.com using one on a 2001+ MX5. Stock ECU is happy, car drives like stock albeit with 240+bhp

 

Honestly, I'm not sure about the F/IC. According to AEM's website, you're limited on injector size to 2x stock, which would stick me with ~480cc. However, I personally know someone who ran it with 550cc, tuned by a local shop.\

However, I don't know the "tricks" the shop used to achieve this, or why the limitation exists. I'd like to understand more before proceeding down that route.

As for the parallel ecu setup, tricking the O2 isn't the problem. Also, how would it be wired up (need a diagram)? What happens with the ecu control signals? Again, Reverant didn't recommend this setup because of the limitations.

 

o2 emulator is only necessary if you remove the cat.

I have been passing emissions for the last 3 years running an MS1 in parallel without changing out anything for the test. The only thing you need to do is emulate the forward O2 signal with the innovate and have it report a range of 14.0 to 14.7. The rear is still the stock sensor plugged into the factory computer. I am using a 300 cell metal core cat.

 

I've heard that in some states, you will be passed automatically if you show up twice with your ecu not reading 'ready.' In other words, disconnect your battery right before going in. I'm not sure how it works in other states, but in MD you get refunded the testing cost if your ecu is not in ready mode, so the only thing you stand to lose is time

 

In NY its 37 bucks pass or fail. If you fail they give you an extension on the inspection. If you go in without the computer being ready that's an automatic fail. Doesn't matter how many times you fail, they will not pass you.

They cracked down on the fraud associated with the inspections something fierce out here and a lot of shops lost their licenses so now they are by the book about it.

 

Well, according to ChipTorque, you can't run an Xede with injectors more than 30% > than stock. I'm running 550's with it. So take that limitation with a grain of salt.

 

You get one free retest in MD, not a refund. You do get 3 months to either get it fixed or spend $400 trying and get a waiver.