I'll preface this by saying I know it's been discussed, I've found a couple threads on it, but not sure on how to resolve my issue.

If the car is hot, and I park it, and turn it off, and let it sit for a couple minutes, when I fire it back up it's VERY lean. It richens up over a few minutes.

I had this issue with both my MS2E and my new MS3 Basic (both rev built).

I've read relocating the IAT will help, but my IAT is mounted in the cold tank of my FMIC, directly in airflow. Which I'm pretty sure is just about an "ideal" location.

My solution as of right now is that everything is tuned fairly rich. (14 at cruise, around 12.5 at idle when everything is up and running) otherwise I end up stranded if I take the car to a store and run in for 10 minutes and come back out, the car basically won't run it's so lean (I hve to sit there revving it to keep it running, and it'll idle at like 17-18). I don't like my car idling at 12ish, so I'd like to fix this.

So far I've found these:

Megasquirt MSEXTRA / MS3EFI • Hot start running lean [Fixed] (View topic)

https://www.miataturbo.net/megasquir...an-idle-59281/

Megasquirt MSEXTRA / MS3EFI • Why use Gair (View topic)

So it sounds like I need to tune my GAIR, but all of my searching for gair tuning just shows me other people having the same issue, not how to tune it.

 

My other idea was to get it tuned and dialed in for a decent idle when hot and running, and then tune my ASE taper to be super long for coolant over like 150, and only have a small enrichment for that longer length of time.

 

https://www.miataturbo.net/megasquir...n-table-57663/

Nothing to see here. Move along...

 

I fixed mine with 10% rich idle VE under normal conditions, lots of ASE over CLT above 120F, and 10%EGO. I did it this way because my MAT did not go high at the restart, so MAT correction would have no effect. Also, long enough ASE taper so that WBO2 and EGO has time to take over. So the EGO goes to 90% at stop lights to give desired AFR, then will swing to 110% after a 10-20min shutdown with restart.

My winter experiment will be to go to a return type fuel system, but this is livable for now.

Occasionally have a longer start sequence, but that is the worst side effect I've noticed.

 

we;ve been having this discussion. its mostly an issue, at this point, with the returnless rail and the temps ******* the injector's deadtimes.

 

I was wondering if it possibly had to do with heat in the fuel system, seeing as how all things being equal, the sensors SHOULD spit the correct fuel into the engine.

I'll tinker with some more ASE and EGO as stated above. Would like to be able to idle leaner than 12.2 under normal conditions.. lol.

 

seems like the boiling point of some of the ingredients in gasoline can be as low as 100* F. When you get those little tiny gas bubbles in the injector and try to shoot them out with the fuel, I imagine you're probably going to run a little bit lean. Then you gotta go through all of the hot fuel in the line before you can get some cool fuel in there. Stupid returnless fuel system.

 

One of the reasons that the returnless systems have higher pressure pumps is to raise boiling point and prevent vapor lock. Agree on other points.

 

The only problem with this theory is that the stock ECU doesn't have this issue. I had a wide band long before upgrading to the MS3, and never had lean issues on hot re-start with the stock ECU, using the same injectors I have now. The stock ECU relying on the narrow band ECU for closed loop works fine on hot restart... so I wonder if large amounts of EGO compensation are all we need?

Keith

 

First about small vs big injectors.

Small injectors are far less susceptible to this because the change in dead-time is a much smaller % of the injector open time, as compared to big injectors.

For example. Assume a constant dead time of 1.0 ms.

Small injectors may need a flow time of 1.5 ms. So the electrical pulse is 2.5 ms. If the dead time increases to 1.1 ms when hot, then the flow time is reduced to 1.4 ms. That's a 7% decrease.

Bigass injectors may need a flow time of 0.5 ms. A 0.1 reduction in flow time is a 20% decrease.


2nd, the factory ECU probably has an additional enrichment for hot restarts - or it may increase the dead-time for hot restarts. It may also have a timer which counts how long a hot engine was off before re-starting.

 

as well as a fuel tank pressure sensor...

 

Posting to commiserate a bit, as well as note that this is my current logic also.

Hot idle at a stoplight is nice right now, high 13ish. Much better than where I had it, as you noted, in the 12 range.

Hot restarts are lean, a tiny bit hunty, but it runs. If I take off easy after that hot restart things are fine, and it's no longer upset after about 30sec drive time -- by the time I get out of that grocery store parking lot.

 

your IAT is heatsoaked, so your ecu thinks the air is less dense than what it really is. As a result, it supplies less fuel.

Once fresh air cools off the IAT sensor (20-30 seconds or so) things go back to normal.

 

What injectors is everyone with this problem using? I have RX8 425cc/min myself. Just wondering if this is an issue with a specific type of injector.

Keith

 

I'm on the same, but I seriously doubt it is related to this particular injector. It does have an additional issue around 1.500 total pulse length, but on my install, this only occurs on closed throttle decel.

I do have a theory, though, if your survey does point to the yellows. I'm pretty sure that they are used as secondary injectors. Might that mean that they never run at low PW on an RX-8?

 

See my almost perfect workable solution in post number 4. I think with a bit more tuning of cranking pulse and PWM I will be able to drop the "almost".

Again, this winter I will put in a referenced FPR. It will only increase the pulse widths at idle a little bit, reducing the EGO swings from +/-10% to about +/-8%, but will have the advantage of quickly bringing up cooler fuel.

Know, though, that a poster on a similar thread said that even with his return system, he had some hot restart issues. I don't have details on his build.

 

I am having issues and I've got FIC 650s but Im also on ghetto MS1. However, I'm lead to believe the injector heat soak theory. I see really lean conditions for 10 seconds after hot restart, recently with scary sounds that might even be det in this summer heat. After those 10 seconds things revert back to normal. This is considering overcompensation of ASE and idle VE just to try and counteract it. It doesn't help and once things are normal, I've just got super rich idle. I've monitored AITs and they arent abnormal so idk whats going on.

 

Having same issue with DW1000s on Rev MS2e. Idle is normally in the 11-12 but can skyrocket to 17-19 and not run at all when sitting between 5min-1hr.

IAT sensor mounted on cold tank and shows ambient air temp during all of this so I think injector heat soak is the issue.

I'm new to this so will try increasing ASE over 120....can you define "significantly"?
Screen slip of what the graph looks like now would be a huge help.

 

I have attached the relevant screens.
Idle Target AFR is 14.3 and Idle VE are set to be 10% rich under normal running conditions. The EGO, which is given control starting at CLT of 125*F, then takes the 10% out at that condition. Rich hot-restarts is not an issue, so the aggressive ASE does not seem to hurt on an immediate restart, even though the injector heat soak has not occurred.

The ASE time is set long enough (cycles) to allow the O2 sensor to begin working and pass control to the EGO.

Since total enrichment is ASE plus WUE, I included both. Not saying my WUE is right, just that you need both pieces of info to interpret correctly.


Priming Pulse


Cranking Pulse


Cranking Duty (PWM idle valve)


WUE


ASE


ASE Taper

IIRC, I have the rally hot tapers so long as I takes longer for the injectors to cool back down to "ambient".


EGO Authority

Hope this helps.
DNM

 

Matched my settings...or at least as close as I could (MS2 doesn't have full EGO authority table)

*crossing fingers*