87FCTurboII

iTrader: (2)

Just installed my LC-1 wideband and went for a couple of test drives. I'm seeing ~13.5 afr WOT, through the full rpm range.

Car is an 87 TII. Rebuilt 5k miles ago, mild streetport, S5 turbo, 3in RB exhaust, Rtek 2.1 in MAP based timing mode, EBC, 550 primaries, 750 secondaries, re-wired FD fuel pump.

I changed the boost level and I still see 13.5 afr at 10 psi, 7.5 psi, and with the boost controller off (5.5 psi).

Cruising I see the afr bounce around between 14 and 15, which makes me believe that I'm getting a correct reading. Idle is at 12 afr. I re-did the free air calibration on the LC-1 after my first test drive and I still see the same numbers.

My Rtek was set on 550/720 when i was seeing 13.5, so I switched it to 550/550 just to see if it would richen up, and I saw 12-12.5 WOT after doing that, butt dyno felt better too.

When I did my fuel pump re-wire, I checked the voltage while driving around and saw 12.5 volts at WOT. I've also logged injector duty cycle and I've never seen over 70%.

Why am I so lean? Could my stock fuel pressure regulator be bad? It seems that fuel pressure is the only big unknown to me. I'll re-check the fuel pump voltage too just to try and rule that out.

Part of me thinks that I'm lucky to still have a running RX7 with afr's that lean while running 10 psi...

Feel free to throw out suggestions for things for me to check, I'd like to find the root of the problem before I start changing the fuel maps around to dump more fuel.

blk87Turbo2

iTrader: (23)

Where in the exhaust pipe did you install the wideband?

It doesn't seem quite right that you would be running that lean without seeing some kind of problem. I would think idle should be where you are leanest.

I had a helluva time getting my LC-1 calibrated on my ole' WRX, took forever to get proper readings...that's why I'm going with a PLX wideband on my FC.

j9fd3s

iTrader: (3)

with the Fuel Pressure Regulator, they do not usually go bad. what usually happens is that people hook the vacuum line up to the wrong nipple, there are a couple choices, but one vacuum nipple sees vacuum AND boost, the other one only sees vacuum.

so tee the boost gauge into the FPR vacuum line and make sure you've got the right one.

arghx

iTrader: (3)

That's not the case on a pre-Renesis rotary engine unless you have the smog pump hooked up. The overlap between the intake and exhaust ports require a rich idle to maintain stable combustion. The smog pump directs air to the exhaust ports (through the port air valve inside the air control valve). Mazda released a whole paper on this while they were developing the Renesis.

It sounds like the OP's wideband is reading correctly. On the stock ECU/Rtek the AFR will fluctuate between 14:1 and 15:1 during closed loop operation.

I agree that it is kind of a pain. It has good features and all but it's too finicky.

RotaryRocket88

iTrader: (7)

Do you have a log? Any fuel changes, or are is the fuel map zeroed out? In my experience, the defaults with 550/720 injectors on the 550/720 setting (stock fuel really due to the preset adjustments) were not enough to put it into an AFR range I was comfortable with. It would tend to be around 12.0. I added around 5-8% additional fuel at 10-14 psi to get the mixture to settle into the mid to high 11s range.

Switching to the 550/550 setting with 550/720 injectors means 15% extra fuel while the secondaries are in use, and that's more fuel than I ever needed to add. That should have dropped you into the 11s.

I've never had trouble with calibrating my LC-1. As long as you wait 8+ hours after running the engine to hit the calibrate button, the numbers should be just fine. When there's a problem with the values, it becomes really obvious. A couple times I've had it just start reading in the 10s everywhere, but recalibration corrected it. The only actual failure I've had with it was with the Bosch 02 sensor, which died after about a year.

RotaryEvolution

iTrader: (30)

focusing a little more on what arghx mentioned, do you still have the smog pump on the car?

HAILERS2

You have a RTEK 2.1.

ADD FUEL if you think it's too lean. Simple as that. That's why they made the 2.1 version, so you could add/delete fuel and mess up the timing.

I just don't understand.

87FCTurboII

iTrader: (2)

Well by some sort of black magic, my car now runs at 11.5:1 under full boost. All I did differently was hook up the analog output so I could pull some logs for you guys, that shouldn't have changed anything. I took it out for a drive with the 550/550 setting still on, and I was able to peg the wideband at full rich, so I switched it back to 550/720 and it ran right at 11.5 or so.

Here's a quick log I pulled.



This is on the 550/720 setting like it should be. AFR even drops down to 10.5 or so at high revs now.

My logs show higher injector D/C and pulsewidth than I was seeing before (before being every log I have, even ones from before I had my wideband), so something has obviously changed, but I have no idea what it was.

Hmm, just now I noticed that my airflow data is much higher than it was before, about 15-20% higher. Same boost, RPM, and IAT as before, but higher airflow... This is consistent on all of the logs I pulled today vs. my old logs.

So from this new revelation, it seems that my AFM was giving me low readings before, causing the car to run lean. So I guess I'll just keep an eye on it from now on to see if the problem comes back.

One thought, I did disconnect and re-connect the ECU plug while I was adding the analog output, so maybe the afm pin had a poor connection before and got re-set when I re-connected the plug.

Hailers forgive me for trying to find the cause of why my car was running lean before I started changing the map to dump more fuel...

Thanks for the suggestions guys! I'll let you know if the problem comes back, hopefully the pin for the AFM input to the ECU just had a bad connection before.

RotaryEvolution

iTrader: (30)

make a restrictor pill for your pressure sensor line or move where the line connects to the engine, that is an ugly boost map tracer signal reading.

87FCTurboII

iTrader: (2)

I have the factory restrictor pill in the MAP sensor line. I've even tried to move the pill as far away from the MAP sensor as possible in an effort to increase the damping effect of the pill, but I still get fluctuations like that in my logs.

Keep in mind that the graph is a bit misleading since it's scaled to fit.

RotaryEvolution

iTrader: (30)

well i have never understood why mazda used the port so close to the engine ports, this makes a very erratic signal reading point. i usually use the upper or lower front ports near the throttle body for TII models.

sharingan 19

iTrader: (16)

Glad to hear you got your problem figured out. I am running the same setup as you and I am having a similar issue, but worse (the secondaries don't seem to come on at all).

Afr is fine below the staging point, (11.1-12.1) but as soon as the secondaries are supposed to come one it rises to 14-15 before shooting into the 20's. I have tried using the 550/550 setting with no luck, I've added fuel to the map, and that has no affect either. I have verified that the secondaries click by switching the pins w/ the primaries at the ecu and spinning a spare CAS. Fuel pressure (both peak and hold) has been tested and there are no leaks.

Interestingly this problem only appeared after I changed my clutch/flywheel, although I can't imagine what that could have had to do with it.

satch

iTrader: (1)

Did you ever confirm whether the RTEK indicated the car was "not in neutral" when it was actually in gear?

sharingan 19

iTrader: (16)

Not quite sure what that would accomplish? In any event, if it thought the car was in neutral wouldn't that affect more than just the secondaries?

satch

iTrader: (1)

If the ECU senses the car is in neutral then it won't allow the secondaries to fire, which is something you mentioned your car was doing a few posts above and I don't recollect your mentioning one way or another if the RTEK verified that the car was "not" in neutral. Anyway, if things suggest the ECU is constantly reading a low voltage regardless of what gear the car is in then perhaps try jumpering a wire from an ECU pin which has switched 12 volt power to pin 1G (neutral switch).

RotaryRocket88

iTrader: (7)

^Satch, did you find that bit of info in the FSM or training manual? The only criteria I've been able to confirm for the secondaries to function are 1) 3800+ RPM 2) 0 manifold pressure or greater 3) Throttle position.

At any rate, there's no way to tell via the Rtek if the neutral switch is on/off. But there is a flag for the secondaries, which will show you when they come online.

Sharingan, I've attached a snapshot from one of your logs that you posted awhile back, and your secondaries do in fact come online. The thing is you're just making so little boost (3 psi max) that they come on very late, and actually go on/off at one point. For low boost, the AFRs at the transition actually look fairly normal. I'll sometimes get a momentary spike to ~13 AFR, but it evens out by the next data point. It looks like in the particular log that the mixture settles into the 12s soon after, which is perfectly reasonable when you're hovering around atmospheric pressure.

Is your car still having trouble making boost, or did that get settled?

satch

iTrader: (1)

I hope I wasn't misleading anyone, but I thought the secondary injectors wouldn't fire if the car was in neutral and "one of the tricks" to entice the secondary injectors to fire was to disconnect the TPS and Boost Sensor vacuum hose so folks could diagnose certain problems while in the comforts of their driveway and not have to worry about the car being in gear. Now when the car is in neutral the car will rev past 3800 rpm and the TPS will register the change in throttle position, so this leaves the manifold pressure which I am not sure what exactly triggers this pressure, but if the car is in neutral I'm going to guess that there is no positive manifold pressure present or the injectors would fire since two of the three criteria were already met. Thus, there must be another player which helps to dictate what conditions must be present for the secondary injectors to fire, and other than the neutral switch, what else tells the ECU that the car is in gear so positive manifold pressure can be made so the secondary injectors can fire as they should.

RotaryRocket88

iTrader: (7)

When the engine is under enough load, manifold pressure will reach atmospheric (or positive pressure in turbo cars). That's where leaving the vac line off the pressure sensor comes in: the sensor will now measure atmospheric pressure (0) instead of vacuum from the idling/free-revving engine, so the ECU will think there's load on it. Criteria #2 would be met.

RotaryEvolution

iTrader: (30)

technically unhooking the pressure sensor shouldn't make a difference, something else is at play such as AFM not pulling in just quite enough air to trigger the threshhold for the secondaries to come online. i say this because when you hold the throttle wide open you are already running atmospheric pressure anyways. perhaps it pulls it only slightly more close to 0"Hg which somehow helps push it over that wall, but i wouldn't consider that a reliable way to test them.

RotaryRocket88

iTrader: (7)

The FSM test for the secondaries is to remove the boost sensor vacuum line, disconnect the TPS and rev to 3800+ RPM. If you skip removing the boost sensor line, the secondaries won't come online free-revving (or at least not until high in the RPM range when you could briefly reach atmospheric pressure or greater).

In the various datalogs I've compiled, the secondaries have never come online below 0-4 psi.

EDIT:

Found this in the training manual. 0 to 3.5 inHg/1.7 psi as the pressure criteria.

RotaryEvolution

iTrader: (30)

i know what you're saying but i'm just trying to say that disconnecting the line really only makes a very minor difference in the signal. put a gauge on the line after disconnecting it from the pressure sensor and go WOT with the throttle, you'll see almost 0"Hg coming from the line, same is if you disconnect it and let it hang.

this is because the throttle plates are wide open and free flowing air is all that is passing through, maybe 1-2"Hg will be pulled because the runners will clamp down on the air to some small degree, that would be the only difference that makes the test valid but still makes me scratch my head.

there is also varying degrees of absolute pressure depending on altitude, people at higher elevations will get less of a vacuum signal when disconnecting that line as say someone would at or below sea level would, with all the atmospheric pressure being forced down on them.

i'd almost be more inclined to think it would make more of a difference disconnecting the pressure sensor electrical connector versus the vacuum line, as the sensor works in series with the AFM if i recall correctly.

but all that aside, it just makes me curious why disconnecting the vacuum line really affects it much at all for this test. if you go wide open with the throttle, that would basically be the same. making the assumption that mazda sais to disconnect the vacuum line just for the sake that the person performing the test may not be going wide open and some vacuum signal is still present keeping it from going into that "closed loop" state to open the secondaries.

i also don't really believe the 0-3"Hg value they state. in most configurations the pressure signal port the factory uses is actually **** poor and results in readings that bounce heavily like in the above diagram from 3psi down to 8"HG while holding a steady WOT during pulls. even with a filter, the average was still right around 3"Hg and the injectors were clearly coming on before that figure tripped while the car was moving. this figure may only be valid with the car stationary but is a vague reference as there is other things at play in triggering their operation as well.

here is an example of a pressure sensor trace in the factory configuration off the port just off the side of the engine:


here is one after relocating the line to the port under the throttle body off the driver side of the UIM on that same S5 engine:


cleaned up a little bit.. in fact it should be noted with every Rtek 2.0/2.1 that the line should be relocated for accurate tuning results. even with standalones that omit the AFM it is even more critical to use a cleaner port for accurate MAP signals. i had a haltech n/a that was using the stock port and i was having issues cleaning up transitions, simply moving the line helped tons.

the top picture gives a very good example of port opening airflow resonance, you can almost see each and every time the rotor passes the port and shuts off air which affects the pressure sensor signal.

RotaryRocket88

iTrader: (7)

I think the intention of the FSM test is primarily for you to see that the secondaries are coming online at the specified RPM. If you just wanted to see if they were coming on at all, revving with the pressure line in place should work; I'd just expect the transition to be somewhere north of 3800 RPM. I'll make a datalog of free-revving later to see if it holds true.

Also, I think the 0 to 3.5inHg spec is likely to allow for a little variance in the pressure sensor output. The acceptable voltage range for atmospheric pressure in the FSM is something like 2.2 to 2.7 volts for either the NA or TII sensor (don't remember off the top of my head). And for altitude correction, the atmospheric pressure sensor inside the car probably applies a correction factor to the MAP sensor output.

As far as the boost signal, mine turn out a lot cleaner than the graph above. This is with a home-made restrictor pill in the line, and the vac line attached to the stock location on the ACV. At full throttle, the pressure variance between data points would be at most 1 psi, but it tends to be less than that.

j9fd3s

iTrader: (3)

yes that was WEIRD. not only the signal, but the chain reaction where weird signal = weird power = weirder dyno readings.

RotaryRocket88

iTrader: (7)

Karack, is the pixel/time ratio adjusted from the default on PLViewer in that first graph? I can't see how the line could have all that static unless we're looking at about 1000 data points. In my graphs below there are something like ~100 data points, and I've cropped the ends off. If you look at my boost signal graphs below, they're very smooth where the throttle is steady. I must be doing something right, as the variance between data points is only 0 to 0.5 inHg.

Here's a little bit of free-revving out of gear, without touching the boost sensor line or the TPS. The second peak is at -0.8 inHg to -0.3 inHg from 1684 RPM to 4256 RPM. My secondary staging point is set to 3584 RPM, but the secondary flag was 0 throughout (secondaries off).



This one is cruising down the road, taking it out of gear, then blipping the throttle a couple times. The secondaries do come on in this scenario, but not until 5244 RPM & 0.4 psi of boost pressure.

RotaryEvolution

iTrader: (30)

i didn't adjust the resolution but the owner who installed the software onto my laptop may have.

the first log i posted has about 850 data points with a 1:03(1 minute, not 1 hour) log duration. each event is logged at .07 second.

which in that event i guess it really depends on how often the ECU interpolates the signal, because if my mapping is more accurate with more data points then the ECU may be using that data more dilligently than filtering/smoothing it such as say only using the signal every .70 seconds versus .07.

i will say that after changing the pressure sensor line there was a noticable difference on each car. the whole thing that alerted me to the issue was while watching the palm pilot during several runs the MAP signal was bouncing between 2psi and 5"Hg yet rarely was near 0"Hg, more often towards 3-4"Hg, after moving the line it was almost a total steady 0-1"Hg signal. note the line had been changed and had no restrictor, which may be ok for your car but even with a pill i made for this car it still only helped some, moving the port location helped the most.

keep in mind the intake runners taper down as the air enters the engine, which is a more accurate engine vacuum signal but more erratic. putting the line further up the intake path will get you less vacuum but smoother signal, a compromise i didn't really have any trouble dealing with.