I have a 1995 Miata that has failed emissions, which didn't surprise me. I need to make it pass in order to re-register it. I have deleted all of the emission stuff except the cat (brand new from magnaflow). It failed because the inspector noticed that the car did not have the EVAP canister, and my HC was too high at 3.21 (limit of 0.8) and CO at 51.69 (limit is 12), but NOX passed at 1.69 with a limit of 2. At minimum I need to reinstall the EVAP canister, but my car was dyno tuned without it, so would I need to re-tune the car once I put it back on, or can I just throw it back on and it will just work? And is the CO and/or HC high enough that their could be something wrong with the car and/or tune, or do tuned Miatas just emit that much? And I noticed that my headers (and my cat to a lesser extent) have turned blue, but I don't know if this is because of the fact that I got them off ebay, or if the car is running super rich.

Mods are SpeedyEfi COP conversion, ebay long tube headers, Skunk2 throttle body, intake piping to delete the MAF so i can run speed density (still using stock airbox), and full emissions delete except for the converter, and the car dyno'd at 108hp, with my afr staying around 13 but dipping as low as 12.4 and going as high as 13.75.

 

Out of the box MS3 does not control the evap system. IIRC it uses the connector for EBC.
Can't you put the stock ECU/MAF and exhaust back in and give it a try? With a warm car, you shouldn't have a problem with the evap system, as long they don't check for codes

 

I have passed on the rollers several times over the last 10 years, but it is only getting harder with the adoption of LEV and CARB cat requirements.

Please post the emissions test graphs as well as the log you took when getting tested

Unfortunately, a dyno tune is rarely adequate WRT emissions testing. That requires detailed tuning that you simply don't get from a third party in a few hours. Deleting EVAP didn't cause your numbers to fail, but removing any easily inspected emissions equipment automatically subjects you to increased scrutiny.

Seriously though, without the data above, you are shooting in the dark.

 

The WOT tuning doesn't affect emissions. If you have shitty injectors that are batch fired and can't accurately idle at 14.7, that'll affect emissions.

You want to target 14.7 at 10* timing and retest, then start adding in evap equipment. Most tunes will idle around 18 degrees and 13-14 afr.

 

But in theory i could get MS3 to make the evap work? even if I had to retune it?

 

evap won't result in a re-tune, it is also not your problem.

There is another thread that has been active in just the past few days that will answer you evap question. The search function would take you to that thread, or simply scroll through recent posts.

 

Here's the sheet they gave me, sadly I didn't take a datalog when I was getting tested.

 

From what the tuning shop told me, they didn't do just a WOT tune, they tuned so it would idle and drive like stock. But my wideband at the time was broken as the sensor decided to take a **** a few days before I got tuned so they used the one attached to the dyno and just kind of locked it in from how they described it. Now that I think about it, this is probably the source of the problem. But I ordered a new sensor, so hopefully if I put that on and a canister, in addition to do doing what you recommended that might work.

 

gosh, my brain just automatically told me you were in Colorado because the 'CO' for carbon monoxide was in the OP. Hopefully Arizona hasn't Californicated itself in the **** like Colorado has.

Usually you can request detailed diagnostic data, in the form of graphs, that show you what the car is doing while on the rollers. This is super helpful because the aggregate numbers (HC, CO, NOx) are representative of an integration of all readings over the test. I have seen cars fail flat out cause the tune sucks everywhere, fail because AE is poorly tuned, and even fail because decel is bad. The diagnostic graphs almost always tell you right where to go from a tune perspective.

On MS3PRO you can configure logging to SD on a trigger (like TPS>5%) with a long duration, like 20 minutes and an auto-restart. This way the car just automatically takes a log when it is being tested, no PC or tunerstudio required. FW versions of 1.5 and up give you 120bytes of logging fields, and if you don't do a bunch of crazy generic sensor snit that gets you logging of just about every parameter.

If this context is over your head, either dig in and learn or find a professional to assist. If the context is over their head, find a new professional.

Edit: oh and no tune done in the absence of WB on MS3 is any good. They should have turned you away or told you specifically that the tune is going to be garbage and will need to be redone. Hard, but these are the facts Jack.

 

Just put the canister, solenoid and lines in again. It'll just sit there doing nothing, solenoid closed.
No way to use it on ms3 without ridiculous effort (see mentioned thread)

Also get it retuned with a wideband.

If you're still failing then I'd look at trying another cat.

 

I don’t want to go back the tuning shop I went to as they finished my car almost 2 months late when they initially said that it would take 5 days, and they said they were 100% ok with tuning the speedyefi ecu I had, but a few days after I dropped it off they said they wouldn’t touch them anymore, and that I had to get a different ecu (why I’m running a MS3 now). And they are the only shop in my area willing to touch a Miata, so if I get a dyno tune they are my only option. Would I be able to retune it myself and use the current tune as a more well done basemap and I would just have to mess with the fuel table a bit, or would I have to start from scratch?

 

95% of tuning should be done before taking it to a dyno (if at all) IMO. Learn to tune, stop paying others to do it for you.

Edit: and yes, use whatever is there as a start, if the car runs it is equivalent to a basemap, possibly with some timing optimization. You shouldn't need to touch timing at all really, it is like the very last thing you do to squeak out the last bit of power.

 

A recipe to pass RIGHT NOW is as follows:

make vehicle at least appear to have appropriate emissions related equipment
log everything!
Get Closed loop EGO functioning well
autotune your VE table
dial in AE, target the lowest adder that results in acceptable performance, reduce TPS sensitivity as much as possible
turn on decel fuel cut
install a new catalytic converter (new converters have a higher efficiency for a couple-few hundred miles). Be aware of regional requirements for compliant converters, some places require CARB approved
touch up VE table using autotune to adapt to new cat, but do so minimally (see above)
test car
R/R new cat with old cat (optional, so you have a high efficieny cat for next time)

 

Or you could keep the ECU stock and use magic corn juice.
Empty tank, add a gallon or two of corn. Pass with flying colors, add gas before romping.

WIN.

I had similar problems with similar mods, magic corn saves the day.