RexRyder

iTrader: (10)

Got my car on the dyno today finally but having some problems with ignition issues. It free revs to redline but under load it wont go past 4.5k. It sputters really bad, the wideband shows 10.5-11 afr. It sounds like a 2 step. My mods:

13b-re
850/1600's
FFE 36 crank trigger
aem smart coils x 4
BUR9's plugs all around

thanks for any help

LOGS:
https://www.dropbox.com/s/eb4x3wc63z...20PC%20Log.daq
https://www.dropbox.com/s/1hwb84topf...20PC%20Log.daq
https://www.dropbox.com/s/b3nofmleb2...20PC%20Log.daq

MAP:
https://www.dropbox.com/s/i5uplqjmzg...EM20.01V24.cal

ryan1

iTrader: (5)

I had a similar issue, my problem was aem's fd trigger settings are wrong, and a bad alternator was messing with the trigger signal. If you look in my trigger issue thread Cody gives trigger settings that work properly.

RexRyder

iTrader: (10)

I didnt see any "crank h sens" settings in my map. I think its because I am using AEMtuner. Also I have a HALL effect sensor.

I'll check the alt, I remember seeing 15-16 volts when it had the microtech, not sure if thats normal. Thanks

C. Ludwig

iTrader: (27)

AFR shows a flat line of 10 all the time in all logs. Ignition timing shows -5 or so at full load in the logs. Appears that's what the timing map is set to as well. You'll need some more timing for one. Get the boost under control until you get the timing sorted. Turn it down to 12-14 or so. At that point you should be running 12-15* leading advance and a 10-12* split. Start there, get everything sorted so it will make clean pulls and then work you way up from there.

You're using IGN-1A/AEM coils, correct? You can also dwell them a little harder. Thought this isn't the primary issue right now. You have the dwell set at 3ms across the board. Up it to 4.5ms at 13 volts. Drop it down to 3ms at 16 volts. Run it up to 6ms at 6 volts and linearize between the points.

scotty305

iTrader: (2)

I agree with C. Ludwig, -5 degrees in the ignition map should make the engine run poorly when it hits those cells. I've seen some people do that intentionally to prevent overboosting... be careful with because too much retard can actually create antilag and cause the boost to increase.


I think you're also giving the car way too much fuel in boost, based on comparing your injector pulsewidths from the '2006-01-01 1427' log file against numbers from my car. After doing some math to account for the different injector flow rates (mine are 1000 / 1000, yours are 850 / 1680), both our cars seem to be running OK using ~400 cc/min total fuel flow around 30% throttle, 2000 RPM, 0 psi boost. My car runs OK using ~1500cc/min total fuel flow around 100% throttle, 4000RPM, 10psi boost... it's tough to estimate your fuel numbers without knowing the Fuel Inj3 Pulse, but if my calcs are in the ballpark you're probably seeing ~14ms numbers for Fuel Inj3 Pulse which would be about 4400 cc/min total fuel flow at that same RPM and boost level. My engine is a mostly stock 13BREW, stock twins but I still can't imagine any amount of mods that would make your engine want nearly 3x the fuel flow at the same amount of boost. I suspect if you have someone take video of the back of the car during these runs, you might see black smoke out of the tailpipe.

If you're worried about crank trigger pickup settings, datalog the channels 'Stat Sync'd,' 'Timing Errors' and 'Sync Errors' while doing a run. Try to log 'Fuel Inj3 Pulse' also.

scotty305

iTrader: (2)

Here's a quick log from my car including a working O2#1 reading and both the primary (Fuel Inj1) and secondary (Fuel Inj3) injectors. Note the temperatures are in degrees C, and the pressures are in kPa... 100kPa is 0 psi boost, and 170kPa is about 10psi boost.

https://www.dropbox.com/s/2dchsden9v...20PC%20Log.daq

RexRyder

iTrader: (10)

Yes the coils are ign marine smart coils, and I forgot to add Im running E85. I think theres so much fuel in there because the tuner was trying to see if that was the problem. I will go over this with the tuner during the week. Thanks for the input

RexRyder

iTrader: (10)

Ok went thru the map and got a lot of help from people here on the forum. A couple things I noticed was the 'Fuel Difference' and 'Ignition Difference' (split timing) were set to off. so fixed those, the dwell, the base timing, split timing, went thru the fuel values completely, also changed the fuel difference values. i wish the car was here to try it out, but ill have to wait until sometime during the week. heres the map. thanks

RexRyder

iTrader: (10)

forgot the link: https://www.dropbox.com/s/02dbjv1yhy...EM20.01V24.cal

RexRyder

iTrader: (10)

So I had some time to do a couple pulls on the dyno and with the revised timing and dwell it still does the same thing. I'll post a log in a bit. Thanks

ryan1

iTrader: (5)

Since he has a 36 tooth trigger wheel, shouldn't his wheel teeth,fuel teeth, spark teeth be set to 36 not 24,12,12?

RexRyder

iTrader: (10)

Those are the settings I got from FFE, the way I understand is the AEM can only see a max of 24 teeth and the settings would have to be divided accordingly.

On a side note after looking at the logs, at 5k when it hits that wall the log shows 140% duty cycle with the primaries at like 35ms and the secondaries at 11ms. So going to go back and fix those, it's still strapped on the dyno. Hopefully that's the problem.

RexRyder

iTrader: (10)

alright here is the current map i was using on the dyno same as the one i posted before but with a little less fuel: https://www.dropbox.com/s/qip20mc6sp...EM20.01V24.cal

My logs:

https://www.dropbox.com/s/v7ozc1wi8u...3rd%20gear.daq
https://www.dropbox.com/s/6wrmi9vtn8...1st%20gear.daq

When logging and it happens (the mis-fire) the laptop disconnects for a split second and then it stops logging and tells me to save.

C. Ludwig

iTrader: (27)

In the map you have boost fuel compensation set way high. It's 165% at 15 psi. That's where you're getting your high primary pulse width. The O2 still isn't logging. Shows 10:1 under all conditions. My guess, given the high injector pulse width and the boost comp map, is that you're still flooding it with fuel. It's not showing any timing errors, so, in my mind, it's down to too much or too little fuel or an ignition problem.

One other thing. Your doing this on a relatively cold engine. Highest coolant temps showing less than 160*. You're still in coolant temp enrichment at that point. So don't try to complete the tune with the engine this cold. If you do that, when it does actually warm to 180+, it will run leaner as the coolant enrichment is pulled out. Always tune and log at normally operating temps.

scotty305

iTrader: (2)

Again I agree with C. Ludwig. I'm still seeing huge fuel pulsewidths in your logs, and your fuel map combined with the Boost Fuel Trim table adds up to very large numbers compared to my fuel map. Thanks for adding Fuel Inj 3 pulse to the logs, by the way.

For instance, at 4000 RPM and 10psi, your calibration's fuel map is set to 10.0ms with the Boost Fuel Trim adding about +160% on top of that which would be commanding about 26ms of injector pulsewidth (assuming the other fuel trims are negligible)... compare that to my map which is set to 4.0 ms with the Boost Fuel Trim adding about 80% which would be commanding about 7.2ms of injector pulsewidth.


Who is tuning the ECU? Are you trying to do it all yourself without the help of an experienced tuner? I'm not nearly as experienced as C. Ludwig, but I was lucky enough to learn from a few different people who had been tuning AEM ECUs professionally for years. My RX-7 has been running without problems for multiple years, including a quick dyno session without any misfires. I've also tuned a few other (piston-engined) cars including an autocross car that we dyno tuned on pump gas, competed for a season or two, then re-dyno'd and switched to E85 and has been competing for a few more seasons. When switching that car from pump gas to E85, we did a few baseline pulls on gasoline then drained the tank (car still on the dyno). After refilling with E85, we increased the fuel map by about 30% and continued doing full throttle pulls with only very minor corrections needed to get the AFR's back on target.

Sorry to digress, that little autocross car isn't anything impressive but my point is your 13B-RE engine shouldn't need much more than 30% additional fuel compared to my 13B-REW engine just because you're running E85... your setup might want more like 30-50% more fuel than mine if you've got a much better intake/exhaust/turbo combination that lets it breathe well... but I think it's safe to say your engine does not need 300-400% more fuel than mine does at the same RPM and boost levels. If you look carefully at the logs, your engine is happy (or at least not misfiring) at low throttle pulsewidths that are very similar to the ones in the datalog I posted above. In boost, your pulsewidths increase much higher than mine and your engine misfires where mine runs clean... I don't think that is a coincidence.


The shape of your calibration's Boost Fuel Trim table is really weird, by the way. Most people who use that table set the table to subtract -100% at -14.7psi, then add 0% at 0psi, then add +100% at 14psi, add +200% at 30psi, and then the fuel map will be almost flat from 0psi and above. Your fuel map keeps increasing quickly above 0psi and your Boost Fuel Trim table is about 100% higher than the recommended settings.

One of the best descriptions I've seen for tuning AEM ECU's is on SupraForums... it includes some pretty detailed suggestions from a lot of different people who are respected in that community. You may need to register for an account to view their Dyno Tuning subforum, but it's definitely worth checking out:
Supraforums.com




This next part is rotary specific and won't be covered by that link: the Ignition Difference map from your first calibration was far less dangerous... the newest calibration you posted 'leaner down low' is commanding that the trailing coils fire before the leading coils in boost, which is really scary. This is backwards compared to what the OEM Mazda ECU does... some people online call it 'negative split' and I haven't heard anyone recommend to try it in boost.


Also, are you intending to run that much boost while you're still getting everything dialed in? It looks like you're well above 15psi in these logs... personally I'd start at the lowest boost possible for and then increase boost once you are happy with the fuel and timing. What is your wastegate spring pressure, and are you certain all the boost control lines are plumbed properly?

scotty305

iTrader: (2)

Here's a screenshot of my fuel map & boost fuel trim table:
https://www.dropbox.com/s/xa62t8bz7c...00cc%20inj.png

RexRyder

iTrader: (10)

Thanks for all the suggestions guys. First let me say that the tuner is experienced with aem, but not with rotaries, and I'm thinking he is trying to be very safe and that's why there's so much fuel in the maps.

As for the boost, the first logs I posted had a vacuum line pop off at the wastegate so it was over 15. The second set of logs were just under 15, and that's what my spring is rated at for a bar.

For the boost fuel trim map it has been corrected from -50 to +100.

The split timing map was suggested by someone else on the forum, I will read more into split timing so I can get some better values.

I'll go back to the dyno this weekend hopefully these are the problems. Thanks

scotty305

iTrader: (2)

It's nice that the tuner is trying to be safe, but from what I understand the best thing to do is remember that rotary engines don't want as much ignition timing compared to piston engines. Try starting with very conservative ignition map numbers like five or ten degrees in boost... you might sacrifice a little power but the timing should be safe enough to help you feel confident getting the fuel dialed in. A dyno run at 12.0 AFR may not be ideal for long-term use but it's not likely to hurt anything as long as you keep the boost low, especially with E85 since it runs cool and resists detonation so much better than gasoline.

In the logs you've posted, it looks like you're holding the throttle at about 40-50% and then the boost goes up to 12-15psi as the engine accelerates, have you tried doing the same thing at only 20-30% throttle in order to get the lower boost sections of the map dialed in first?

The method I was taught was to start tuning fuel by doing part throttle sweeps (assuming you have an inertia dyno like a dynojet or mustang... not a dynapack hub dyno). First sweep at 10% throttle, the engine won't make enough power to accelerate to redline, that's OK just let off where it stops. We always do this with the wideband's 0-5V AFR output connected to the ECU you can datalog AFR's and then look through the log and make corrections to the fuel map so it will be correct (or at least closer) for the next sweep. A general rule of thumb is if the car is 1.0 AFR too lean (for instance, you want 12.0 AFR but you got 13.0 AFR) you should add 10% to that cell in the fuel map (hit the 'E' key for percent change, then type in 10 and hit enter). If the O2 is 2.0 AFR too lean (wanted 12.0 AFR but the log shows 14.0 AFR), you should add 20% to that cell (remember to keep the surrounding cells smooth because the ECU is almost never using just one cell at a time). Conversely, if the log shows 0.5 AFR too rich (wanted 12.0 AFR but the log shows 11.5 AFR), you would remove 5%. Next repeat the same 10% throttle sweep with your updated fuel map, make adjustments if necessary. After you're happy with the 10% throttle sweep, move up to 20% throttle and repeat. By starting at low throttle and working your way up, you're putting less load on the engine while dialing everything in, and giving time for the engine and drivetrain to warm up before doing full-throttle power pulls.

Once you've got some experience at this, you might be able to see patterns and make some adjustments to the next set of load cells so they can be a better starting point before doing the next sweep. If you're recording power during these part-throttle dyno sweeps, you'll probably find that most properly-sized throttles will make the same power between 60-100%, so you might not need to spend a ton of time doing sweeps at 50, then 60, then 70, 80, 90, 100% throttle.

scotty305

iTrader: (2)

screenshot of my Ignition map & Ign Diff map:
https://www.dropbox.com/s/ywkndzr156...ence%20map.png

C. Ludwig

iTrader: (27)

Didn't even notice the trailing split map. Good catch.

As Scotty said, the boost compensation should be -100% at -14.7 psi, 0% at 0 psi, 100% at 14.7 psi, and 200% at 29.4 psi. Basically the base fuel map then becomes a 0 psi VE map and the boost compensation takes care of various levels of MAP. For each additional atmosphere (1 bar, 14.7 psi, 100kpa, however you want to value it) the fuel requirement doubles. This is why the boost comp map is set this way.

The settings for the boost comp map you just described make no sense.

ruddyrid

iTrader: (8)

Just now came accross this thread and I think I'm having basically the same issues. Will try to look over my cals and correct any potential issues.

RexRyder

iTrader: (10)

Mine revs to redline fine under no load, but with load it does what yours is doing. The backfiring/mis-firing was so bad it pretty much destroyed the turbo.

I'm replacing the injectors with some ID2000's, getting some NGK R7420 plugs, and pretty much going over everything. The replacement turbo will be a t04z.

ruddyrid

iTrader: (8)

I actually ended up finding the solution to my problem. It was basically the dwell settings that did it for me. It was set to 4ms across the board, in this case at all voltages from 6v to 16v or whatever.

I brought it down to about 3.2 from 12v-14v and down to 3.1 by 16v. And stepped it up to 3.3 in the lower voltages.

Another thing is I didn't have split timing turned on. I found the parameters I needed to adjust and it ran even better. Now it basically has no misfires.

C. Ludwig

iTrader: (27)

What coils are you using? I highly doubt changing the dwell from 4ms to 3.x did a thing to effect how it runs.

ruddyrid

iTrader: (8)

I'm using the stock FC coils. But it seems like it definitely did fix something. Before, it would hit 6kRPM and just like misfire like mad. After j changed that it was fine and revs past 6000 no problem.

I was even messing with dwell just idling, if I went past 4.5, idle quality would degrade rapidly.