CheesePowder

I know the emissions topic has been beaten to death but I am hoping that a good tuner familiar with emissions can point me in the right direction. I want to win here. I've tested 4 different times, each time getting ever closer to success.

Car: Stock injectors, rewired fuel pump (45 psi roughly), A/P and ACV all functioning, Bonez DP, literally brand new Bonez cat, new 9 plugs all around, new coil wires, HKS megaflow intake (carb legal for testing), PFC/datalagit/LC-1, O2 sensor disabled. LC-1 is placed at the end of the DP.

Test Results:


http://flic.kr/p/8Y3jzE


Test 1: Rewired fuel pump but running with stock slightly corrected map (obvious fail because of rewired fuel pump without removing excess fuel). Trailing split is 0 in testing areas.

Test 2: Street tuning with LC-1, pulled approximately 20% fuel in the significant areas to roughly 14.7 AFR on LC-1. Ran down a tank of Guaranteed to Pass before testing. New plugs.

Test 3: 1 Gallon of Denatured alcohol (one time just as a last ditch- didn't work); Oil change. More fuel removed. Trailing split changed to 3 in testing areas.

Test 4: New gas (no alcohol), New Cat, new plug wires, used algebra to figure out ratio of fuel removed going from test 2 to test 3 and the hydrocarbons removed on the sniffer to calculate how much more fuel needs to be removed to succeed. Removed that exact amount from test areas and then smoothed out the adjacent cells a bit. Seemed to work for 15 MPH range but not for 25 MPH. The 15MPH is really good I think. It's the best result I've ever had on any emissions test. Think I should just try the algebra again and hope the NOx goes down too?

As it stands, with all the fuel I've removed, the LC-1 is reading approximately 19.5-20.0 AFR without the airpump. With the airpump on it is at 22.2.

Below are pics of the current tune in the relevant area. I've discovered that the split air check valve is missing from this car and have it ordered on the way. Should I just try to keep removing fuel in the 25MPH/3rd gear cell(s) or perhaps the ignition settings are not ideal? Or maybe the check valve is the culprit? I've passed 3 separate times without the check valve...albeit with a stock computer.

Injection (msec)

http://flic.kr/p/8Y3jzE

Leading

http://flic.kr/p/8Y3jAh

Trailing Split

http://flic.kr/p/8Y3jAS

Split should be changed? Retard timing? Just blindly go with more fuel reduction until it won't drive correctly anymore? Suggestions?

arghx

iTrader: (3)

Look at the relationship between the AFR and the emissions output from the engine:



Now that's a generic graph out of a Toyota emissions manual for a piston engine, but the overall trends hold true.

I highly doubt this reading. You really need to check the LC1. It can be sensitive to grounding and electrical noise. A rotary can't run that lean without serious misfire (a Renesis can run leaner without misfire than an older 13B). Furthermore, if you were really running that lean I would expect an NOx problem. Only lean burn direct injected engines can run at 20:1 or leaner, and they have major NOx emissions problems which necessitates special catalytic converters, sulfur traps, and converter regenerative processes.

You said your cat is a new Bonez so all signs point to a rich mixture, either due to an airpump problem or an excessively rich mixture coming out of the engine.

jeff p

with everything working properly with a new bonez cat you should pass just fine. I put the stock ecu in the car for the smog test with a hks downpipe and bonez cat and I passed just fine I even finished the test early due to good readings on the roller "so the smog guy said"... but I have all the smog stuff working on the car and the air pump conected to the cat with the stock check valve piping....I am running the stock factory plugs 7 & 9's ... so u must be either running rich or very lean not to pass. I am thinking your very lean at idle?

jeff p

forgot to add I am on stock ports with the stock injectors running the hks twinpower ignition which I think helps a great deal with smog.

Jeff

CheesePowder

I figured something was up with the LC-1 though it's never given me odd readings or fluctuated strangely. I grounded all related items to the same ground point where the ECU grounds. I will check the current calibration again.

Even with the high NOx, if you think removing fuel is the solution, I will try pulling more fuel until the 25MPH range shows the same AFR on the LC-1 as it did for the successful 15MPH range. I mean, I removed fuel and the NOx went down to 9 for the 15 MPH. That was surprising.

Any qualms about the ignition settings for this range or is it likely fixed with just fuel?

Thanks for the help

arghx

iTrader: (3)

try running the white and green wires on the LC1 to that same ECU ground. then run the blue wire, which is the ground for the heater, to a separate location. That's how I have mine. Just as an experiment, keep leaning it out until it sputters. Normally at that rpm and load (15 and 25mph) once you get to about 16:1 or leaner you start having misfire, although sometimes you need a richer mixture than that.

The timing isn't going to have anywhere near the effect that the AFR will have on emissions, unless something is WAY messed up.

wstrohm

Our '94 failed emissions in CA in 2002 several times. With 72173 miles on the car, the original OEM cat converter, original O2 sensor and 14K miles on the plugs, the failure was only at 25 mph, and only on CO, which was 0.61% (max allowed was 0.5%).

So I put a new Bonez downpipe and Bonez cat converter on it, and a new Bosch O2 sensor. The retest failed again, this time at 72461 miles, again only at 25 mph, but this time the failure was HC at 83 ppm (max allowed was 53 ppm).

So I put in four new NGK plugs and retested again. This time it failed once more at 25 mph; HC was 94 ppm!

So finally I bought and installed a Mazda OEM cat converter ($1265.62). The retest after this change passed, with HC @ 25 mph = 34 ppm.

Since then the car has been tested at 81098 miles in 2004 (HC @ 25 mph = 11 ppm) and easily passed, then again at 89249 miles in 2006 (HC @ 25 mph = 9 ppm), at 96060 miles in 2008 (HC @ 25 mph = 23 ppm), and at 101861 miles in 2010 (HC @ 25 mph = 17 ppm). That latest 2010 report shows all parameters well inside all requirements at both 15 mph and 25 mph.

My present thinking is that the OEM cat, although expensive, was the best fix in 2002, and so far has lasted 8 years and test results still look very good; and the original OEM cat had lasted 8 years also, from 1994 to 2002.

I'm not sure what the internal structure of the OEM cat looks like, but I suspect the air from the air pump is distributed within the cat more efficiently than in the Bonez unit.

jeff p

I pass just fine with the bonez cat. and no pre cat just a downpipe!. I think the trick here is the stock air piping and check valve going to the cat. I retained it and it seems to provide air to the cat at the right time to get the readings to pass. also the stock ecu turnes the air pump on at higher rpm's than the pfc so the stock ecu is a big help in passing smog.

j9fd3s

iTrader: (3)

20:1 = misfires which = high HC's.

really this is very simple, you might wanna just put the o2 back in and turn it on, the test is what its for, basically.

my FC does 10ppm HC's @25mph.

t-von

iTrader: (8)

Very nice info. FYI the water steam cleaning that most rotor heads do to help for carbon build up will actually helps clean the cat. It really burns a lot of the internal build-up after so many years. I did the water thing twice a year. When I removed my cat, it had 108k on it. I shined a light to do a visual inspection and the honeycomb was bright white. Now I never had to emission testing in my part of the country but, my car never smelled bad ever.

CheesePowder

wstrohm, thanks. I actually read through your emissions test thread(s) in my quest to fix this. It is my belief (or hope) that since I messed with the fuel pump rewire, my troubles stem from a bad tune. Thus I am putting off buying an OEM cat as much as possible.

arghx: So I was mistaken. It turns out I have the heater ground at the console with the same ground as the LC-1 gauge (separate lugs). The ECU and the rest of the LC-1 grounds are soldered in a single lug at the original ECU ground point (the bracket that holds the ECU). I did it pretty much as you did in this thread minus the AN2 ground to the datalogit and the additional ECU ground:

https://www.rx7club.com/3rd-generation-specific-1993-2002-16/my-power-fc-lc-1-datalogit-wiring-diagram-707301/

I will add the AN2 wire and see if the LC-1 gives different results when I delta them.

j9fd3s: will switch the O2 on for some street runs and if it feels right, I'll try it on the next test.

t-von: I performed a steam clean prior to test 2 but it didn't seem to help (or maybe it did but I still have too much fuel and the cat can't do anything). I do steam clean but not on a regimen- more arbitrarily.

Thanks for all the input guys!

j9fd3s

iTrader: (3)

the test is just at a steady 15mph and 25mph... 2nd and 3rd.

its basically steady state.

arghx

iTrader: (3)

I'm not 100% sure how important it is to separate the blue wire. I made that wiring diagram years ago and I recall reading somewhere that it helped, so that's what I always do. I can tell you for sure though that with the LC1 the delta AN1 - AN2 configuration gives a superior signal. Your PFC and gauge (if you have an external gauge) should be anywhere from dead on to .2:1 off

scotty305

iTrader: (2)

The air pump can skew the AFR readings measured at the downpipe, especially when the ACV is directing air into the exhaust ports (as opposed to sending air directly to the cat). It's very possible 17-20 AFR may be what the O2 sensor measured in the downpipe, and very likely that measurement would be different if the air pump was disabled.

From what I've noticed, many new OEM vehicles are closer to 15.0 than 14.7 when idling or cruising (no air pump). I'm not sure if this is primarily due to emissions or economy, but I suspect it may be worth a shot. Also don't forget that these tests are probably designed for big-torque USDM passenger vehicles... the RPM will be lower than most small engines work well at, the test will be lugging the engine in the wrong gear, and the person driving will just use more throttle rather than shift down because of the test RPM requirements.

CheesePowder

This again. Owning this car has been exhausting...no pun intended.

This time I replaced the port air and split air check valves, turned O2 Feedback on, and went to test.

When datalogging with the O2 feedback on, the LC-1 reads between 14.5 and 15.3 in the emissions testing ranges. It was a good sign: either both are off or both are functioning correctly.

So to summarize the testing conditions: PFC with fuel pump rewired, new bonez cat, DP, 4 new 9 plugs, new coil wires, Twin Power, street ported, O2 feedback on, new port and split check valves. Still fail:


http://flic.kr/p/962TxC

So the question now is: "Should I continue under the assumption that O2 feedback is good and entertain other possibilities such as weak coil, leaky injectors or should I go back to tuning without O2 feedback and just lean/richen the cells used for 25 mph, which was the only one I failed last time?"

What would you do or what would you do first?

RotaryEvolution

iTrader: (30)

recalibrate your wideband and retune it leaner, it's obviously out of calibration.

CheesePowder

Karack, are you basing this off of the 5th test failure with O2 feedback on and high HC or the progression of failures from Test 1 - Test 5?

Also, does that mean that you would test it with O2 feedback off and go using the tune only?

Thanks for the quick reply.

arghx

iTrader: (3)

We haven't had sniffer tests in NC for 5 years so all I can do is provide general advice. I think the air pump should still be pumping air to the exhaust ports even with O2 feedback off. So turn O2 feedback off. Air pump will still be on. Tune until AFR is 16-17:1 before the cat, as long as the engine isn't so lean that it misfires. Multiple owners have reported this AFR before the cat when port air is engaged.

RotaryEvolution

iTrader: (30)

basing it off the issue of the wideband reading about 3-4 points too lean.

if you want to get an idea of where your map should be for calironia emissions: write down the figures you have in those affected cells of the test range, now lean them out in about 5% increments until the car starts to buck under low loads while driving and then richen them up just a tad.

you don't even need a wideband to tune the car for emissions but it does make it quicker knowing how far off you are richness wise.

if your wideband has never been calibrated, it may be way off depending on when you set it up and if it hasn't been calibrated recently.