Hot restart issues, solved.

[Red_Bullet]

Maybe something stupid like this: Or maybe even longer taper if necessary.

[url][url=https://flic.kr/p/FKTFeg][/url]ASE hot start fix proposal by Keith Wade, on Flickr[/url]

[NitroDann]

Yes people do that also, a longer taper after 60c or similar. You can employ this and a modified IAT curve. Also you can set the fuel pump to run with the key and just let it cycle for 30 seconds before you start it. All stop gaps, really. I bet the OEM uses the SE IAT to measure soak only and solves the issues this way.

Dann

[Red_Bullet]

Working around the problem rather than solving it, I can live with it though, it's not such a big deal.

Not keen on running the fuel pump for long periods without starting car. I've already had a lambda sensor fail once. I put it down to sitting for long periods heated before starting the engine, condensation or whatever when cold, I suspected thermal shock got it.

[NitroDann]

That's a fair call.

Dann

[bartmanftw]

An intercooler which is not heat soaked is very good at keeping IAT within 10*c of ambient. Heat soak sensor readings are typically 20-30*c higher and therefor the IAT curve matching ideal gas law tapering to rich at the very high end is a good workaround in my opinion.

Dann

I agree, I'm just saying care should be taken if adjusting VE table based on data obtained when the air temperature is in the area where the density curve starts tapering towards rich. For example I adjusted my VE table from data I logged on the track on a ~40*C day and it caused me to lean out with cold air temperatures. My fault for doing that instead of adjusting the MAT density curve though.

Here is another theory from Miata Turbo. It's to do with injector heatsoak. At 30 degrees C the injector compared to 100 degrees C. At 100 it takes 100uS longer to open than at 30 degrees C. This makes the actual deadtime longer hence cutting total fuel supply by around 10% at idle. Opening time increases because of the increased resistance of the injector coil. Apparently it goes from 15 ohms at 30C to 19 ohms at 100C.

This might explain at least some part of the hot start issues.

https://www.miataturbo.net/ecus-tuning-54/injector-current-trace-shows-slower-opening-when-hot-67534/

Would definitely have an effect at lower duty cycles such as idle where the dead times of an injector have a much greater effect on the fuel being injected.

Maybe just try to ASE our way out of it, more fuel and longer taper maybe. No doubt though you've already tried that.

I think this is the best way of dealing with it as starting with coolant above 60-70 degrees almost guarantees you have some heat soak.

The air temperature reading can also be ignored during ASE but I think that is a pretty crap workaround since it will still be ignored during cold starts too. Perhaps allowing it to be ignored above a certain temperature would work better.

[hks_kansei]

. I've already had a lambda sensor fail once. I put it down to sitting for long periods heated before starting the engine, condensation or whatever when cold, I suspected thermal shock got it.

I've actually wired my sensor to a switch specifically for this reason.

If I'm sitting in the car listening to radio etc without the engine running I can flick the switch and shut off the wideband so it doesn't heat.

[Locutus]

It would be simpler to just rewire it to power on only when the key is in the "ON" position.

[Red_Bullet]

It would be simpler to just rewire it to power on only when the key is in the "ON" position.

I agree mine is wired to ON not ACC.

I usually just start the car immediately, If I have any reason to leave it in the on position without it running (like doing a firmware upgrade) I just pull the dedicated lambda sensor fuse.

For a while I toyed with the idea of setting up an output to drive a relay to turn the lambda sensor on. Something along the lines of IF RPM>600 turn on.

But I couldn't be bothered. Sometimes simple is better.

[Red_Bullet]

Discuss. (There is plenty on this on Miata Turbo for background info) https://www.miataturbo.net/megasquirt-18/what-does-your-mat-air-density-table-look-like-89026/

Braineack style MAT Air Density Correction by [url=https://flic.kr/p/2455eD8]Braineack style MAT Air Density Correction Temporary for VE tuning by [url=https://flic.kr/p/24miNrQ]Ideal Gas Law MAT Air Density Correction by [url=https://www.flickr.com/photos/132829053@N08/

1. I've been persuaded to use the Braineack (Picture 1) style curve with good results, VEAL tuning is far more consistent. Also no risk of going lean when it matters at higher temps. I see the sense in VEAL tuning with picture 2 for consistent results.

[GR124]

Whilst we are on the subject of the ideal place for the IAT. A lot of the stuff you guys are talking about is over my head however I do understand the possible issue with IAT heat soak.
Currently building a track only locost/clubman. For now I'm running the stock NB8A ECU.
After a bit of research and asking about positioning the IAT on the ozclubbies forum I was about drill into the rear/firewall end of the manifold/plenum and rubber grommet the OE ATS. I was concerned about heat soak but I was told it should be as close to the inlet runners as possible.

Thoughts?

[NitroDann]

We cant be certain what the OEM ecu is doing with the info so you should put it in the factory location.

With an aftermarket ECU the sensor is typically used to increase fuel or decrease fuel by the same amount the air is getting thicker or thinner due to heat, so in that case you need it to read air temp entering the inlet manifold accurately.

The issue being discussed here is that screwing the sensor into a huge heat soakable item such as the inlet manifold will result in inaccurate readings.

Your car has an airflow meter which directly measures the amount of air entering the engine so no 'air is hot-thinner/air is cool-thicker' adjustment needs to be made. The fact that the ecu is measuring the sensor and that its in the intake air doesnt necessarily tell us what the factory ecu is doing with that info.

For example they may know that it heat soaks and are using it only to measure heat soak in the engine bay. They may take the air flow reading from the air flow meter to determine fueling and then use the air temp sensor to decrease ignition timing advance to protect from potential detonation 'just to be sure'. We simply do not know.

Dann

[hks_kansei]

It would be simpler to just rewire it to power on only when the key is in the "ON" position.

removing dashboard bits and unwrapping the ignition barrel loom/wires and finding a wire live at IGN, vs tap the cig lighter right next to the gauge.

[GR124]

Cheers ND for your response. I was reluctant to fit the Air Temp Sensor into the manifold due to heat soak. If it ends up in the wrong spot I't will be an easier fix if I fit into the 76mm ally tube just after pod filter for now. I'll do that.
Thanks again for your knowledge.