Hot start issues - Unfixable?
 

Hi all,

Years of lurking the forum & there is finally a problem I can't find a fix for...

My cars been turbo for about 18months now and in that time my biggest problem has been hot starts. I have thrown money, time & specialists at it but it still remains.

Unfortunately I am very much still a novice when it comes to tuning but I am learning.

As the car didn't do this NA and I have since replaced or taken apart & cleaned every mechanical part that could be affecting the hotstarts/misfires I am starting to believe there is an error with my tune. (Plugs, leads, upgraded to COP, taken apart and cleaned ICV, cleaned throttle body etc etc)

I brought the car back to the garage who tuned my car but they claimed they "couldn't replicate the error" which is interesting as I can do it on demand.

If I drive the car somewhere and get it hot, then get out & do what I need to do for say 15-20 minutes, when I come back to the car it will splutter & more often than not fail to start without throttle. Once throttle is removed it will yo-yo and hunt an idle very strongly.

Are there any MEITE experts on here? Car is running an ME221 and I'm at the point I'm tempted to sell the car it's that irritating.

Thanks in advance!

(I have videos of the car dying and photos of the tune if anyone wants them)

Attach a tune and log of it happening.

Most likely, you need to substantially increase your post start enrichment above about 180 degrees coolant. If you're on Flow Force's injectors I've found its much worse, not sure why. Again, I don't have your tune, but most MS type ECUs you can change ASE counts from cycles to .1 seconds. .1 seconds are much longer than cycles, so you can hold the post start enrichment longer with seconds rather than cycles. I'm in the 400s for .1 seconds (so 40 seconds) for ASE taper.

For sure fixable. Just need to diagnose the problem, then you can decide how to solve it.

My advice to determine the cause of the problem.

Buy a can of CO2 spray for cleaning dust off like this: onn. Electronics Duster Compressed Gas Cleaner, 10 oz - Walmart.com

You said you can reproduce the problem on demand, so heatsoak it so you know it will run rough.

Then pop the hood and empty and entire can of the duster stuff on the fuel injectors and the fuel rail. This will remove a lot of heat from them.

Fire it up and it will run fine.

Heatsoak affecting injector opening confirmed. Then solutions can be

1. Return style fuel system that circulates fuel through the rail.
2. Endless tuning to atempt to mitigate the above issue.

My 2 cents.

Log
 

Thanks for the reply,

I've attached a log,

I will say the log attached is a "mild" bad hot start, often it is much worse than this with the car dying out rather than starting on its own.

I've also noticed cranking the key & running the starter for far longer than feels appropriate seems to help with the hot starting, for whatever that is worth?

All help appreciated, I haven't implemented your crank fueling adjustments just yet (tuning noob etc), let me know what you think after looking at this.

Also the forum didn't like the tune format, how should I save it to allow uploading?

Tune should be a .msq, should attach fine.

GRR ME221, forgot. I tried to look at the log on MLV, couldn't even find AFR, unless it's "lambda curr AFR 1" and running at 10.36. That would make sense as your AFR error is 4.34, so you should be targeting 14.7 at idle? So maybe fuel isn't your issue. Although running at 10:1 wouldn't help. Ignition advance might, as well as a calming down the idle valve, which I think I looked at? it's moving similarly to your RPM.

Thanks for mentioning the AFR and timing advance, I actually went back to my tuner & said does this look right a few months ago but they never replied & simply took the car in but didn't find the problem.

Given everywhere on the internet is saying I should be targeting 14 AFR & have advance 10-16 degrees I'm not sure I agree with the tuner. I'll look into saving the tune down in another format but it doesn't look like I can via ME221's tuning studio.

I've attached a photo of the software showing the indicated advance & AFR at idle. This can't possibly be right surely? 20+ degrees advance? The car runs & drives perfectly when it's not heatsoaked. Even when hot it'll do a full pull just fine.
https://cimg5.ibsrv.net/gimg/www.mia...96df732dbb.jpg

Just change your tuner. Looks like they only tuned cruising a bit and WOT and thats it.
What a shame

edit:
What the hell is going on with them.... 25° at idle.
Even IF base timing is like 10° off, all other cells would be totally off as well, this is pure junk

Yeah not really sure what's happening here looking at that timing even as a novice it looks busted,

It was tuned by the guys at Mtech (UK well known turbo conversion & partners of the people who make the ME ECUs), so I'm giving the benefit of the doubt that it's the tuning software displaying wrong?

Wouldn't it be pinking/detonating like crazy at that much advance?

​​​​​​I unfortunately can't save the tune from the ME221 as a format the forum likes. Best I can do is a photo of the cell table.

Any UK folk on this forum that can recommend an independent to look at it?
https://cimg1.ibsrv.net/gimg/www.mia...d849329398.jpg

MBT is in the mid-twenties around idle cells. I assume that the majority of modern ECU solutions offer an idle advance curve or table that is used when idle conditions are met. Using a primary timing table that provides MBT everywhere results in a much nicer off-idle response as well as better start-up transient conditions.

It is significantly easier to control idle speed via air-metering with the motor power degraded through timing at idle. The engine is much less sensitive to variables in the de-tuned state. Additionally, through idle RPM correction features you can more effectively stabilize idle when you provide the algorithm space to work with in both the advance and retard domains, so say 14 degrees base +/- 7 degrees for RPM control.

Hi all still working away at this cars brainmelting hot start problems...

I convinced myself some injector thermal sleeves would fix the issues but naturally this did nothing.

Please can someone provide their load tables from MEITE?

I'm just going to take it into my own hands now and mess about with fueling at cranking rpm till I have the problem fixed but would like to have a "good" tune to centre myself around
Thanks

See post 3. You never responded to it. Diagnosing the cause of the problem is a good way to know how to fix a problem.

Thanks I thought I had replied my mistake, I did take your advice and that's why decided to fit some thermal sleeves to the injectors themselves. Can you elaborate on the tuning idea to fix it?

Like you say I can guarantee its heatsoak & starts with throttle(more air), but your post made me consider that actually more throttle will also mean more fuel (cooling the injectors) so could be either

You cant' tune away a hardware problem. I wrote how to diagnose a hardware problem. If it's a hardware problem, you can attempt to mitigate it with software, and this can help, but it won't 100% fix it.

Thermal sleeves will do almost nothing if the rail and lines are all heatsoaked.

I've spent countless hours myself trying to tune around a hardware issue. If that's the route you insist on going, GL.

If applying the air duster cleared the issue fully, you're diagnosed!
You added heat shields...
Problem still exists?
Can you again clear the issue via the air duster method ?
Then the heat shields are either ineffective or there is more heating of the fuel in the rail than just shields can solve...
The next easiest way to increase cooling is to have a return style fuel system in which the fuel in the rail is not dead headed.
Works even better if you energize the fuel pump for 15-30 seconds before you attempt a start (pushes the vaporized fuel particles down the return line).

Years ago, many fuel injected OEM vehicles had dedicated fans and ducting directly on the injectors to combat exactly what you are describing.
1981 280ZX turbo had a big setup over the center of the valve cover. It ran for 10-15 minutes after shutdown.
It worked great
Until the timer module fails and it takes the battery down overnight...

Pat's post 3 was EXCELLENT advice if...
The application of the air duster always clears the issue.

Decades ago, when no-one was aware of the environmental impact, we used R12 instead of a duster (they didn't exist back then either).
R12 freezes the shit out of anything...
You didn't have to use much.

I would NOT try to clear a true heat soak via software. I don't see how you could.
Pat's a wizard at the software end of things and maybe he would have a chance.
Personally, I'd bet against him fixing a heat soak issue via a keyboard...


If Pat's diagnostic procedure doesn't always clear the problem then something else is going on.

One thing no-ones touched on is what is your static fuel pressure at the 15 min after shutdown point.
Most FI systems have a check valve at the fuel pump that prevents backflow. A properly functioning system will often show static pressure 24 hours after shutdown.
Any liquid boils at a lower temperature as you reduce the pressure. Water boils below 0c at full vacuum.
Your static rail pressure should be 60psi (NB?) which is worth maybe a 100 degree INCREASE in boiling point for fuel over atmospheric.
Might be a bad FP check valve.
A bad check valve would exhibit longer than normal crank times for all starts, not just the hot ones.
The hot ones would have a bunch of vaporized fuel to deal with and they would be far worse than a cold start.

Both the air duster test and a fuel pressure check should be done first, and mechanical solutions be employed if any trouble is found.

Thanks for the feedback I will try that & see how I get on. My car is a Mk1 if that helps regarding fuel pressure.

The problem I'm confused with is the setup I have on my car is fairly common, Bosch injectors & a safe power output of 240hp, though I mostly run 180hp on the local roads to me. I also live in the UK, at best it is 18degrees Celsius ambient at the moment, why would I have greater heatsoak than others?

You doubled the power which doubles the heat.
You now have a large metal heat sink in the engine compartment.
Add that to flat fuel pressure at start up and it may well be your problem.

It's worth testing before you "melt your brain" or melt others trying to figure out the ME221 before verifying you have liquid fuel at pressure behind the injector.
That's the FIRST thing to check regarding heat soak in anything.

So, a MK1 has an early system with a return line. Correct?
Have you changed the fuel pump?

I'd check fuel pressure first before using the KBDuster.
If pressure was OK the KBD is the next step for me.
I'd freeze the injectors/rails, coils, any temp sensor that feeds the ECU.
If the KBD didn't help anywhere I'd look to the coolant temp sensor for the ECU next.
This is the point that you go to software based solutions.

I might, as a work around, make a FP switch temporally to allow FP run before start to allow for system priming/cooling as you should already have the return line.
30 seconds of fresh fuel should clear any vapor from the loop.
A fuel pressure priming circuit negates the requirement for a check valve, you just have to wait a bit between on and start.

I have a sedan that has a big engine. It's known for heat soaking really bad. The manufacturer programmed the car to intermittently circulate fuel through the rail every 20-60 seconds or so, for several hours after a hot shutdown. Pretty clever as it keeps the rails/injectors cool despite everything around them being so hot.

Thanks for the advice sounds like a good area to try, however wouldn't I have stuttering etc under load if I had issues with the pump/fuel pressure?

Read it again, with special attention to vaporized fuel.

@technicalninja do you have a recommendation for a low-cost, in-line check valve?

DNM