I autotune idle actually, well, usually I can. I get it to ~14s then auto tune to my afr target. Usually works. With these injectors it doesn't. I have min rpm set to 1500 and min fuel load set to 40. Idle cells look a lot like Deezums, a block of 56s, then 57, then 59, then finally at 1500 I'm allowed to do whatever I want.

I'll enter in Aiden's dead time table and hit the dyno later today. I'm thinking I want my required fuel to be correct though, so I'll multiply my fuel table by 2 and make all rpms under ~80kpa equal.

 

Neither Ben@DIY or Reverant use the idle ve table, I am not sure why not. I absolutely love it and have been using it to tune idle AFR's for years.

 

I idle smoothly around 900rpm 14.5-15.5 AFR. Thats with all the idle cells identical. My deadtime calcs will only matter if battery voltage is changing dramatically. Which it might be when you rev it up.

Your sig is outdated btw.

 

Finally nailed my idle. Kris at KO Racing (highly recommended) helped. We started by increasing load to maximum except AC (cause that changes idle). So lights on bright, blower on full, etc. Adjusted the fuel table for 14.0AFR then turned on EGO correction. 60 ignition events per step, 1% step size, 10% authority. We also turned idle advance off, but idle RPM timing correction on. This allowed us to use the "Idle RPM timing correction curve", and set it to +10 for anything more than -100 rpms from target, and -15 for anything above +100 rpms from target. The ignition advance gauge moves a LOT like this, but idle is rock solid. Base timing is 25* at 950rpm.

I haven't started using closed loop initial values, I'm going to look at a few idle conditions before I enter in values for my best guess, then start generating my table.

Something is definitely up with off throttle conditions. When I let off, despite MASSIVE map based accel enrichments, and fuel cut shut off, and the fuel table being pig rich at low load/low rpm, it spikes to full lean right around idle, unless I give it massive dashpot adder. Currently I'm at 13.3%. It keeps it from dying, but it still has a lean spot and still dips a few hundred RPM lower than I'd prefer.

 

I run into this also at times. Let me know if you find a fix.

 

injector voltage compensation is set up wrong maybe? when revs drop and voltage drops, they just shut off? just a guess

 

Could be. I'm using the table Aiden posted earlier in this thread. I'll see if bumping low voltage helps.

 

Look at the pulsewidths during the lean spikes first. If they also drop to zero then its a tune issue. If you still have high PW and afrs go lean then it might be an injector issue.

Because if you have high PW and lean AFR then the ECU is trying to inject and the injectors aren't listening. If PW go down then the tune is telling it to lean out.

 

Interesting. I've been having a similar issue with my GT500 FF injectors.

Off-throttle car acts like it has fuel-cut enabled when it 100% doesn't. My MS drops pulse-widths from ~1.7ms to 1.2ms as soon as TP dips below 1% and goes full lean. I have to have ridiculously pig rich over-run row to thwart this odd behavior.

My fuel map has idle VE around 56-58 very similar to other posted above, and idles higher than 3500-5500k cruise VE's.

Interested to see what people do to compensate for bizzare off-throttle issue.

 

Could they be acting funny cause they're under vacuum?

 

Dunno. My previous EV1 stockers as well as EV6 RX8 injectors worked fine whilst under vacuum. On same Megasquirt as well. Just weird that MS dumps the pulsewidth so low off-throttle. Didn't remember changing anything besides required fuel, dead-times, as well as obvious VE changes when changing from RX8s to these.

 

1.2 ms pulse width is fuel cut with these injectors.

I did some testing at different pulse widths to build a linearity model. The equipment I was using. Wasn't precise enough to get an great linearity model, but it was clear that under 1.7ms at 12V there was basically nothing. At 13.2V it should be good for a bit lower than 1.7, but not much.

Similar to some of you guys, I went fully auto tune to get a decent tune. I just took my base map, multiplied the whole thing by 0.5 and then auto tuned for a couple hours. Idle I just guess and tested until it was on 14.7.

 

Aidan, what PW you idle at for 15.0 on 91 or whatever? This might explain why you die on low RPM CL idle issues.

For me on e85, 2.0ms=14.7 idle. 30% less is 1.4ms which is scary close to not being able to idle on regular gasoline. https://www.miataturbo.net/build-thr...4/#post1262925

When I was running 91 I'd idle 14.7afr ~1.7ms pulse which I could tell were getting a bit non-linear.

 

Nope. Not really. Closes I get is like 1.7.

I need to get more logs of it. I'll keep bringing my laptop to work and logging.

 

My '02SE idles at 2.05-2.10ms with the factory 60psi system and FF610s. Pump gas, AFRs are 14.7:1.

 

Sounds about right. Deezums and I are both rocking 1.6s so a little less fuel for the same idle.

I'm running seq fuel he's not

 

At 13.2V the dead times are about 0.2ms lower than at 12V, so they aught to run at down to 1.5ms pulse width, but not lower. My 02 idles at around 2ms pulse.

 

If I recall correctly I'm idling at high 1.4ms for 14.6 afr idles. Required fuel at 5.5, VE at 56-58 or so on the table. Vacuum referenced Fuelab FPR (43psi at atmosphere, so like 36psi at idle vacuum). Built 1.8l.

If these things pretty much shut off at 1.4ms that could explain my really really touchy idle.


How Can I get the pulsewidth up without lowering fuel pressure (already low) or running pig rich? Am I missing something real retarded?

 

Does anyone mess with this shiz?

 

No. because poverty MS2