When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Attached screenshot shows some log data from my RX-7 for comparison, similar time scale and I tried to get the y-axis scaling similar as well. My car has a stock engine, but undersized fuel injectors (250 cc/min) because I already had bigger injectors in an Xcessive LIM and wanted to see how well-behaved it could be at idle and light load with tiny Honda-sized injectors. I wouldn't expect your engine to idle at such lean mixtures with 1300cc primaries, I stand by my original recommendation of 0.81-0.85 lambda target until you are happy with idle quality. I would also start with a higher Engine Speed target, perhaps 1100-1200 RPM and then turn it down once you get it dialed in well enough it never stalls when driving around gently.
If you're not sure you trust your O2 sensor due to aggressive porting, try different mixture targets and watch manifold pressure. Lower MAP numbers (more vacuum) and higher Engine Speed for the same exact throttle and IAC and ignition timing usually means the engine is happier. If the O2 sensor is accurate, the engine should get unhappy and pull less vacuum below 0.75 lambda and you'll probably notice black smoke out of the tailpipe, and it will probably be unhappy and tend to stall when leaner than 0.95 lambda but no smoke from the tailpipe.
I agree you should try to log fuel injector pulsewidth, because the duty cycle will seem to move around a lot when the engine speed changes at idle. My math suggests your fuel injector pulsewidth was around 2.5 ms (1000 RPM at 4.2% duty), which should be in the well-behaved range for ID injectors. The same 2.5ms would be 3.3% duty at 800 RPM and 4.6% duty at 1100 RPM.
Rub20B, scotty305, Dale,and Alberto, thank you for your time and thoughts! TeamRX8, appreciate the offer... I may take you up on it yet!
I'll start off with a log update with Inj Pulse Width (red trace):
Scotty's and Howard's math is on the money: about 2.3ms. I can't find whether that includes deadtime. Again, in concurrence, this is well above the non-linear response range. Howard here is the injector linearization and deadtime data: ID1300x2
Lowering fuel pressure at idle by one bar ((35/50)^.5) results in a 17% decrease in volume. This would put me at around 2.7 PW. This comes at the cost of atomization quality. I ASSUME not a worthwhile tradeoff.
You can see the Ignition Timing Trailing Relative (i.e. Split) in the purple trace.
The Ignition floor at 10°BTDC is an Idle Control artifact. It set of the Leading Ignition (base) and increases up to the Idle Ignition Max to try and maintain the Idle Aim. I've tried the map with IC off and then on. The datalog is with it turned on.
I considered Dale's and Alberto's lead of a vacuum leak. I visually inspect the intake when I had it off. While I couldn't visually find any issues, I'll spray the intake this weekend to more aggressively pursue the lead.
I failed to mention originally the Xcessive LIM. It flows into Ihor's largest SP.
Another factor I failed to mention originally is that it has a very crisp throttle response. But the loping is still there at higher RPMs (e.g. 2000, 3000, 4000) also. Dashpot is still incorporated and the returns to idle are rock steady. Hot starts are as good if not better than cold starts.
@C. Ludwig how could I forget to solicit your perspective... would greatly appreciate your perspective/advice on my loping idle introduced about a dozen posts above?
At this point, I think this weekend plan will be
spray down the intake for possible intake leaks
decrease the Lambda
advance Ignition Timing Leading
test Howard zero split Ignition Timing Leading
Check out Scotty's injector timing lead
create a shrine to Rotary Beelzabub with an appeasement offering of apex seal shards
More then...
Last edited by Carlos Iglesias; 06-23-22 at 07:27 PM.
Reason: ADD
just for preciseness, an intake leak might cause the high idle but I think there is also something else causing the 200 rpm lope. I'm hoping (for you) that a slightly richer labda/afr will fix the lope.
you probably know this^^ already but stating it in case you are too deep...
HAPPY INDEPENDENCE DAY... a favorite holiday of this literal Refugee!!!
Since last...
spray down the intake and injectors for possible intake leaks... nothing.
decrease the Lambda... nearly imperceivable improvement.
advance Ignition Timing Leading... nope.
test Howard zero-split Ignition Timing Leading... goose egg.
Check out Scotty's injector timing lead... nada
create a shrine to Rotary Beelzabub with an appeasement offering of apex seal shards... in-works
Rechecking all the settings, I got excited when I discovered that the FFR Trigger was set up as a Mag vs Hall sensor. Change the setting, and I could hear "Murphy" laughing at the delight of his ambush and my disappointment. Alas, 'ner a modicum of improvement.
Next, I go back to the fundamentals. Recheck every single setting. Check plugs. Sleep on it some more... a go-to problem-solving fallback.
The quest continues.
Catch 'Tank'
To bolster some motivation, I decided to earn a small victory. I installed the Radium Catch "Tank". As always, Clean-ness was to be the handmaiden of Function (e.g. performance and packaging). I initially really wanted to install it diagonally on the radiator's fan motor mounts, but the >45° sacrificed too much volume. With several large spaces on the passenger side, I also eschew inordinately long plumbing. So "floating" behind the water pump was the sweet spot. I soldered up a 'floating bracket' for now. I'll further stiffen it up by triangulating it with the hood pin mount eventually. The stopcock and drain tubing are routed to the radiator fan mount. It's low enough to drain well. The end of the tubing is loosely held by the zip tie so it can easily be removed to drain the assembly. I'll call it a win and continue to lick my (idle) wounds for now!!
Engine Bay overview
Engine bay - Radium catch can position + plumbing
Engine bay - Radium catch can position
Last edited by Carlos Iglesias; 07-04-22 at 08:45 AM.
Reason: ADD
Have you tried an experiment with idle control turned off. I run without idle control and while it has some drawbacks it is easier to trouble shoot when looking for a modulating rpm situation. Also I'm not suggesting you keep idle control turned off long term, just consider it as a trouble shooting technique. When I was using my idle control settings I found the challenging thing to get dialed in were all the various PID type of settings. When not well tuned the fuel and ignition settings would oscillate similar to what you are seeing as the idle control tried to get the rpm where it wanted.
With no idle control I find that you can set the cells near the desired idle point in the fuel map and ignition map to the same value. This trouble shooting technique may involve having rows in the tables for 500, 750, 1000, 1250, 1500 rpm and also similarly whatever is reasonable for vacuum columns so you can ensure the oscillation of the rpm/vacuum doesn't go outside of the fixed value range you want to trouble shoot with. Once you have the ignition and fuel values fixed and working well enough you can search for other control items that may be causing the issue, assuming it is a control problem and not an engine air flow / porting induced issue. Also when my engine was breaking in I remember it didn't want to idle within a small rpm range and the range grew smaller as break-in continued. If this trouble shooting technique helps you then you can re-enable the idle control once you have identified the issue and then concentrate on just getting the idle control settings refined. My situation was with Haltech and I would assume some ECU are better / worst than others for managing idle on ported engines.
my unsolicited idle setup advice is as follows
so the problem with a pre Rx8 engine, is that it has lots of exhaust gasses that get pulled into the intake, think of it as having a REALLY BIG EGR valve, so the idle tuning is all working around that. the Rx8's side ports get rid of most of that, so its like having a normal EGR valve, and can thus idle at 1 lambda like every other engine...
my recommendations are based off of how Mazda did it. we know what they did, and it also works. so we have a structure.
Mazda had a separate idle "map" in the ecu, and you could do that too, but it seems like a modern ecu can do the same stuff without it, so you might as well keep it simple.
Mazda considers idle with RPM under 1100, the TPS voltage low/closed, and either in neutral or in gear with your foot on the clutch, which is again more complex than most of us need. its worth noting that over ~1100rpm the stock car is basically in closed loop at 1 Lambda, once you open the throttle the EGR effect improves a lot (they also had an air pump)
step one would be to pick an idle RPM, and make a row/column in your fancy ECU for it. 900rpm is a little high, imo, but its ok. the PP engines idle at 900rpm, so i would think your little wimpy street port would idle lower than the Lemans car.....
step two would be, if you have not already find the AAS screw, and just turn it in all the way, and then back out about 1/2 turn. can be done with the throttle elbow in place. i also think you have some extra air getting into the engine somehow, but we can move on
step three, timing. every 1985-2011 Rotary ran the same idle timing, and you should too, -5 BTDC Leading and -20 Trailing. the older engines run the same timing split, but Leading is at TDC, so its 0 and -15 instead, this is ok too.
after 1989 Mazda does use timing to control the idle speed, as it reacts faster than the air valve. i would suggest doing this, but maybe later, you want to get the rest of the setup set up first.
Mazda's chart looks like this: Mazda's chart is pretty aggressive, it has a tendency to surge and hunt if something isn't just right
Four: Fuel. Mazda's SAE papers show that the older engines have something like 2.0% CO in the intake, or to put that another way they put the emissions tester in the INTAKE and it still fails an emission test, by a lot.
the street port makes this worse.
the result? the engine needs to be run richer to try and get the mixture to fire reliably. a stock ECU runs about .82 lambda, and you might start there.
since you have a better exhaust and slightly higher idle, once its running at .82, you should try leaning it out, the final number yours is happy with really just depends i've heard of people getting it to .9 lambda
once you have that then you can start on some of the other nonsense, keeping in mind things like the idle control valve can only raise the idle, and seems to have a big non linear region just like the injectors
the duty cycles are in the manual, and obviously yours won't be identical, but it does seem to be that too low of a duty cycle, and the valve doesn't work right (like injectors possibly)
I'm surprised to hear the loping doesn't clean up at higher RPM. Does the engine sound smooth or rough? If you feel the exhaust coming out of the muffler, is it a smooth steady flow or are there hot and cold puffs that can be felt individually? If it's rough at all RPM I suspect you've got a single rotor (or injector or coil) that is not working properly or maybe firing out of time. If it sounds smooth and feels smooth but the ECU sees it as rough then I wonder if there might be some sort of software RPM calculation smoothing filter you can adjust to prevent the ECU from making unnecessary changes to fuel or timing. Worst case you might be using the wrong edge of the crankshaft signal; there's a 50/50 chance this can happen (rising vs falling edge, only one will be correct) and it can happen with either VR or hall sensors. If that's wrong you can usually see it when watching ignition timing with a timing light, the timing might drift too far advanced or too far retarded even when the ECU is commanding it to be locked at a fixed value (for instance commanding 10 degrees flat but the actual timing might be 10 degrees at low RPM and then advances to 13-14 degrees at high RPM).
Good luck, it will be interesting to hear what finally solves this for you.
@dk_davis@j9fd3s@scotty305@TeamRX8 Gents, thank you so much for your insightful suggestions. I've incorporated most of them. Once I figure out what's going on, they will undoubtedly optimize my idle. But alas, the solution remains nigh:
I did tune the map along the no-load curve to 3.5K. The surging diminished with increased engine speed. I'll take it!
Overall, the troubleshooting since my last post:
Attesting the basics, I rechecked all of the setup settings. Nothing.
I've predominantly troubleshoot with all closed loop functions off (e.g. Idle Control, Idle Air Control), Tried turning Idle Control with no improvement
Swapped out spark plug to BUR's. No change.
Tried lambdas down to .82 and stock idle timing... n'er a surge tweak.
Rechecked the compression. A little ray of good fortune, as it's only improved (especially the low front side seal):
Confirmed the REF (engine speed) trigger and associate timing. O-Scope trace and timing light check are ROCK solid (<1 degree) to 3K rpm.
Eng REF speed signal voltage trace (cranking from stalled)
I got a maps lookover from the tuner, and he seems to agree it's a mechanical issue. Mechanical... hmmmm? And then as last weekend was running out,,, A LEAD! Having spun my wheels (no pun intended) on so many "leads" now, I'm hesitant to get my hopes up. I'll chase this one, time permitting, hopefully later today. More then...
Last edited by Carlos Iglesias; 07-24-22 at 09:18 AM.
If the tuner thinks it’s mechanical, did you block off the turbo inlet / tailpipe exit and smoke test it? Easiest way to determine if it has a leak. It should be post throttlebody butterfly’s if you do.
High idle- Check weird things BOV for vacuum leaks (so vacuum line at top of BOV pulling air at idle).
Not my pics/problems as I've never used Tial BOV.
As far as idle lope.
When my stock port Mazda crate engine has the 11 heat range NGK spark plugs and the 4" exhaust it has a wandering idle when warmed up.
Not a fast lope like from 2missfires and a fire, but slowly up and down ~100-200rpms over 2-3 seconds.
With the 3" exhaust retaining more exhaust heat in the engine it does not have this wander even with the same cold plugs.
Note that I have stock sequential twins primary turbo retaining exhaust heat as well on my set-up, so it might happen for you on your single turbo with slightly hotter 10.5 plugs and a smaller exhaust size.
Short of Murphy peeing in my rice-r bowl, planning to make DGRR this year. Will check off a Bucket List if I also do SevenStock again this year also. We'll see!
The CYM has been on the road since mid-Nov, but I have a tome to write about getting there. Cliff Notes offered.
For now, only 10 PSI on the WG spring due to flex-fuel wiring issue (source of idle gremlin): 343RWHP/280Q on the Alamo Motorsports DynoDynamic rollers with Texas (E10) 93-Octane. I'm very happy for the first outing after more than a decade and looking forward to the additional 10psi in the next couple of months. Looks like an interim "W" in the Win/Loss column for I-Rotary seals.... for now.
Last edited by Carlos Iglesias; 01-10-23 at 04:26 PM.
Reason: Dumbassness
Indeed Pete, TeamRX8 has the insider scoop... Shane Tecklenburg spent about 6 hours remotely possessing my laptop, including 40 minutes pinpointing my idle problem down to an erratic flex-fuel wiring gremlin.. So far, my car has been tuned by a motley crew of Motec dudes, including Paul Yaw (my original M4) and his "brother in tuning arms," Shane (the M130). So much to add, but time has become precious this last month or so. Some cool tuning insights and a few tall tales this weekend.
That is impressive carlos
i have only heard of shane. Never crossed paths with him but his reputation and experience is off the charts. The difference of an actual professional tuner vs the clowns out there
Lots' to share but as the old saying goes, I'll "eat the elephant one bite at a time"...
Dyno Session
With an acceptable idle and free reving tune (and all the other soon-to-be-told tales of woe behind me), I cajoled a manage-a trois with Brice, (Alamo Motorsports owner), Shane, and myself in November. The day was wet and cool (about 50F-ish).
First, a little more on my "dates." Brice has owned Alamo for +30 years. By his account, he's tuned over 1000 cars across a range of ECU that he sells and/or just tunes. He had virtually no experience with MOTEC and my M1 was his first M-series. He liked it!
The tuner, Shane, was a friend of a friend. Following the axiom that "First the Tuner, then (and ONLY Then) the ECU", I’ve stayed Motec. My original Motec tuner, Paul convinced me that his colleague and friend, Shane, was THE best tuner… full stop! So when Injector Dynamics exploded and Paul moved on to bigger and better things, Shane got the RX-7’s tuning baton.
On the scheduled dyno, Brice strapped the car onto his DynoDynamics rollers. Since Shane was remote tuning, I asked Brice if he would “drive”. He obliged for the whole 6-ish hour session; it turned out to be a really good thing… Spic Racer on Alamo's DynoDynamic rollers
Remote Tune
To my surprise, I am now a newly minted remote tuning fan. With a working knowledge of the Motec software, the remote tune had the bonus benefit of allowing a direct observance of the tuner’s thinking and processes. While in the passenger’s seat with my laptop connected to the ECU and Shane remotely controlling the laptop, I literally got to watch every click, keystroke, function, and step. On several occasions, it was almost like being inside his Shane’s while he worked problems (i.e. idle, injector switchover, et). At the end of the day, remote tuning included a trove of unexpected tuning and performance insights.
It was also cool having another tuner in the driver's seat watching Shane work and discussing his process and thinking. We could deduce Shane’s intention and logic, sometimes even testing each other on the underlying first principles. You know, doing the problem-solving guy-bonding thing that’s so central to our car obsessions. Towards the end, Brice got comfortable enough to start to make recommendations to Shane. I’d type them to him over the chat window. I effectively had a tuning team at that point. It was also reassuring that Brice kept mentioning how impressed he was. At one point, he even said, "I'd heard Shane was the best tuner in the country, and now I know why."
Gremlins
I had tuned the cranking and start-up well, so the first focus was the idle fluctuation. Watching Shane tune it out was interesting. After methodically going through all of the usual suspects (trigger, fueling, timing, etc), he started validating all the sensors. That’s when he found it: an intermittent Flex-Fuel sensor. It fluctuated across the entire range. Since I’ve subsequently replaced the Flex-Fuel sensor and its harness, I know that a crappy DTM pin crimp in the flex-fuel harness was the culprit. As “Murphy” would have it, the pin’s resistance fluctuated when shaken. I had missed it because I had only looked at the ethanol content input with the fuel pump running (test @ 100%) but with the engine off. I now see how I should have looked at it with the engine running (and shaking) instead. I had no one to blame but myself… Sucks to be stupid!! Anyways, with that resolved, Shane then quickly dispatched with getting the idle humming at 950 RPMs.
A result of this was, I chose to stay on the 93 (E10) gas. The E98 container stayed unopened, so we only mapped to the 10 lbs wastegate spring and up to 9000rpm (the rev limiter w. Roger Mandeville race clearance and balanced rotors). Quite a sublime experience hearing an FD go there... even if only at 10 lbs. After an initial “fat map” up to 9000, we re-focus on refining the tune to the 8500 redline.
We then encountered the next gremlin: a lean anomaly in the 6-7K rpm range. The dip did not respond to fueling increase. Again, it was invaluable to literally watch Shane troubleshoot the issue. Increase injector demand, nope. Confirm timing, nope. Fuel pressure, nope. After those attempts, he started playing with the secondary injector contribution (i.e. staging)… bingo! He lowered the contribution point to come on earlier. Both Shane and Brice deduced that this was an artifact of the slightly long secondary port vice the primary injectors. The secondary injectors are in the LIM and the primaries are right in the intermediate housing. Effectively the secondaries had to be turned on at lower rpm/load so that the fuel path was primed when the heavy +6K fueling was commanded. You can see some of the troubleshooting progression in the graphs below.
With the lion's share of the tuning complete, we let the intake/IC/ECT cool while Shane programmed some of the other vehicle functions. We then prepared to re-check start-ups. I set the SPAL fan and DC EWP each to 80% duty cycle, and IC quickly got to ambient while I dropped the ECT to 150F. We did a couple of start-ups while cool to recheck the maps and closed out the day with a final ramp-up. It put the cherry on top with the final 343WHP/278Q pull. I thanked my lucky stars (and Murphy) and called it a day.
The engine I built survived to fight another day... until I return with E50 and an additional 10 lbs of boost. As I've already mentioned, I'm very satisfied.
And then I got the bills from Brice and Shane’s!!
Last edited by Carlos Iglesias; 02-01-23 at 04:55 PM.
Reason: dumbassness
06-22-22, 11:04 PM
pretty sure it’s Shane Tecklenburg, pretty impressive background
