MikeT 88IROC350

So for all you TPI owners, i have a simple question. After you shut down a HOT engine, and let it sit, the fuel pressure in the rails should bleed off slowly. So, where does that fuel go? Back to the tank, via the lines, and regulator? Or out the 8 injectors, eventually, into the cylinders? Mine does bleed down slowly, and i have serviced everything on top, injectors and all. After i get my answer, i will elaborate more why i am asking.

Tuned Performance

iTrader: (92)

It drains back into the tank

Kingtal0n

It should be draining to the tank. But if the fuel injector O-rings are bad it can drain into a cylinder and hydrolock the engine if the leak is very bad.

Some fuel pumps have anti-drain back valves which slow or stop the draining (walbro generic 255)
Some fuel pumps do not (walbro 450)

even pumps that have anti-drain back valves still bleed some pressure slowly over time. It is rare for any pressure to be left in the system after an hour for most systems but some will still hold 5-10psi. Just depends....

MikeT 88IROC350

Thanks, that is the best scenario of course. Mine takes well over an hour to bleed off. I have a FP gage mounted permanently on the rail, so i can watch it anytime. What i was thinking, that the fuel is going out the injectors, into the cyls. But i have had everything serviced, many times since i have owned the car. I did have a leak at the back of the rail, where i had a plug for the cold-start injector. The O-ring failed there, (ethanol likely), and it was leaking.

So my main issue, is that i have long engine cranks after the car sits for a few hours. Starts real fast cold, and after 9 hours of sitting starts decent. But after 3 hours i have long cranks. I can give it a little throttle, during the long cranks, and it helps it to fire. So more air, and less fuel? I do my own chips, and have been working the crank fuel tables. Also messed with the IAC warm park steps table. Just havent been able to nail this issue down.

Kingtal0n

sounds right

pretty easy to overfuel the hot starts, Because sometimes when the coolant is 180*F the engine is still 'cool' and sometimes the coolant is 185*F but the engine is 'very hot'
When the engine is 'very hot' it takes much less fuel to flood a cylinder

Modern OEM efi computers have something called 'soaktimer' to help with this, which adjusts fuel pulse for recent engine shut-offs (so when the coolant is hot, and the engine was running recently, it will inject much less fuel than if the engine is hot but its been off for a while, coolant can stay warm quite a while)

MikeT 88IROC350

Yup, makes since. After 3 hours of sitting in work parking lot, my temp gage is down to 130, but who knows what the engine is at. That's when i get long cranks. On the other hand, you can start a cold engine, only drive for 10 mins, and things are certainly not HOT yet...... i have been backing off on the Crank fuel PW vs Coolant values, but still cannot make headway. I am determined to nail it down, however!

Kingtal0n

I'd suggest maybe using a similar value at 120-135*F that you use for 150-175*F
For reference most idle injector PW is around 1ms (just the normal)
So if you are cranking around idle PW for hot engine, 190*F 1.1-1.2ms of fuel usually
So for 150*F I would only be around 1.3-1.4 if that
Basically dont start with the "3, 4, 5ms" of fuel until the engine is TRULY cold like 60-75*F ambient numbers,
the reason its so much harder to flood when the engine is 'truly cold' is because the fuel doesn't evaporate well. So you shoot 5ms in and 2ms of it evaporates which helps the motor get started and 3ms of it hangs around as liquid on the valve, cylinder walls, intake ports, whatever

If its not enough fuel the engine cranks too much with the throttle closed

If its too much fuel it cranks forever until you "clear flood"
If there is some prime pulse or some initial pulse settings this can also have dramatic effects. It depends when the pulse occurs.
For example if there is a prime pulse 0.250 seconds after you key on. Well, after 3 hours, when you turn the key the fuel pressure in .25 seconds is probably only 30psi at most. So the prime is small. But if you turn the key after the engine was running 20 minutes ago the prime pulse (same pulse) is "large" because the fuel pressure didnt bleed out.

The way it can make you crazy is if you try changing the prime pulse without realizing its pro-rated based on fuel pressure at the time.
Anyways, if there is any prime pulse, use a small pulse until you figure things out with the main cranking pulse tables first.
Then once those are set so the engine fires up pretty fast you can try adding prime to get it even faster.
I usually set clear flood around 65% throttle so I don't have to fully floor it

MikeT 88IROC350

Good stuff here, i will try to keep up with it all........I dont have a prime pulse option in my bin...running a 1227165-6E-WB adx, with a modified ARAP based bin. I have been messing with the crank fuel for long time, but never took good notes on what works and what doesnt. Plus, you have to have a solid TPI system before can fix things in the chip. I have a crank fuel PW multiplier table, vs ref pulse. Maybe that is what you are thinking of a prime pulse? Mine is set for a 1.00x for the first 19 pulses, then i back it down to 0.50 to pulse 24. I think it stays at the pulse 24 value, if you havent started by then.

For my main crank fuel table, starting at 68F, pw= 6.25, then 90F, pw = 5.47, then 111F, pw= 4.69, then 133F, pw= 3.91, then 154F, pw= 3.13, and from 176 on pw= 2.34. Those are current values in my car right now. Already got a new bin started, lowering the values more, from 111F on up. For my crank PW vs TPS, i have 1.00x up till the last 2 entries, where it goes to 0. I dont ever have to use clear flood to start the car. But how do you know it really is flooded? Or not enough fuel? I have cracked the throttle a little, during long warm cranks, and that helps for sure. So more air is needed??

Understand what you are saying about fuel pressure. But the way i start my engine, no matter what, is i hit the key on for like a half a sec, to fire up the fuel pump, then energize the starter. That's just how i roll. It does take some time to build it up.....

Kingtal0n

I think flooded originally used to mean that liquid fuel had puddled/collected on parts enough to prevent the spark. Like a fuel soaked spark plug. Back when carbs would do things like that, especially when over-choked.
However now I believe we also use the term to describe fully vaporized fuel that is simply too rich of an a/f ratio to fire.

There are two ways to 'tell' I can think of atm.
One is if adding air (throttle valve position) or reducing fuel seems to help the engine start. More air, bigger 'gulps' during cranking helps move fuel into the exhaust and out of the way (it can puddle in the exhaust and cause a backfire when the flame finally starts in the chamber and reaches the vaporized fuel in the exhaust).
The other is to check the plug during these times. For example say you crank the motor cold for a couple seconds, and it does not start. So you pull a spark plug and its covered in liquid fuel. Yeah, it flooded.
You could also crank for a couple seconds, then check a plug and find it bone dry.

It's just visual indication of fuel, plus whether it smells strongly of fuel or not.

Anyways, I guess I would start reducing 111F through 154F if thats where the car has trouble starting. Use percentages to keep from having to 'guess' new numbers, and always try less before trying more seems like good practice. For example simply multiply .85 or .80 into 111 through 154 and see how it does. Then gauge your engine's 'reaction' based on that percentage. For example if you pulled 15% fuel and the engine starts 50% better you know you probably need to pull another 10% or whatever

Kingtal0n

another way to tell if its 'flooded'

You can set the clear flood (0% injection) around 25% TPS temporarily for diagnosis for this to make it easier, I just feel like im ripping the throttle body out of the car by stomping the pedal all the way down, you shouldn't need to touch the pedal to start at all

Start by cranking normally, engine does not start, not sure if flooded
now clear flood position and crank again, no fuel goes in now
if the engine is flooded, it will crank a little then suddenly fire and try to start even though no fuel is being injected
it means there was too much fuel sitting around, and that cranking the air pump got rid of enough fuel to have successful spark-ignition

MikeT 88IROC350

I like your idea of pulling a spark plug, when i get long engine cranks, to see if it is wet. Certainly a wet plug means too much fuel. We have all looked at plugs on small engines, to see if is flooded. Just might be hard to time it on the IROC, when it actually happens. I will work on lowering my PW again, at the warmer temps. I use Tunerpro RT, so you cannot just put in any value, from a percentage. After you save the bin, TP seems to put its own values in the table, fairly close to what you ask for.

Not sure i will change the crank fuel PW vs TPS table, to a lower value for clear flood. certainly a good idea for troubleshooting.....maybe if i still cannot nail this down.

Kingtal0n

hah. Use a calculator to get the number then let it make its own new number.
The "new number" is because of memory blocks in the computer have specific sizes or because memory allocation is done in specific sizes.
.
For example some might require multiples of 16, so 32, 48, 64, etc... if you enter 56-63 it corrects to 64.
To figure out what its doing, simply try every number, 1, 2, 3, 4, etc... each time look at the corrected number. the changeoverpoint between 1 and the next step should be about halfway between both.
If it isn't then the relationship wouldn't be linear which can happen if the software is programmed to present data to us that way for some reason (usually to account for the visual aspect of non-linear data)
for example in the small electronic device "super airflow converter" the rpm-maf data is presented as being linear base 10 familiar to us, but the actual voltage compensation for typical foreign maf sensors is parabolic in nature.
In other words the software is working in base 10 linear adjustments to appease human viewing/setting but the hardware is in reality converting that information into something it can actually use inside it's own memory blocks (perhaps base 64 or base16) and then using that to interact with a parabolic sensor.