Twin-TB, Twin-Supercharged 408 with external fuel system nightmare

[Pekka_Perkeles]

Engine details:

- 408 stroker
- some custom cam
- twin 90 mm TB?s with external controller for E38 PCM
- twin superchargers
- air-water-air intercooler
- 1050 cc injectors (Bosch-based I think)
- external twin-pump fuel system next to the engine (single pump for non-boost regions, second for boost)
- base pressure either 40 psi or 58 psi, boost/vacuum referenced
- stock fuel pump, acts as a transfer pump, return line from twin-pump system back to the rear tank

Problems:

Without 2 bar COS, I couldn?t get MAF to fail. So I switched to 2 bar COS. Now with it, VE numbers are abnormally high. Small part of it was that I was running only 40 psi of base fuel pressure. Now it is 58 psi or so.

Also, Cylinder Airmass is very high at idle. MAP sensor settings are a bit off, but even if I fix that, I doubt it will help much.

After succesful start and stop, when I?ll try to start again, it won?t. It?s like fueling has changed dramatically. This may be related to the stock fuel pressure sensor which does not see any useful amout of pressure, since external fuel system has return line back to the tank in the rear of the car.

I?ve tried to disable things that could alter fueling that are related to the stock fuel pressure sensor, but I think I haven?t succeed to do so.

I also think at least one of the plugs in the right side is not ok. I will replace all plugs as soon as car starts consistently. Now it does not.

Stock tune: 01 camaro stock settings.hpt

Modified tune: 125 muutetaan spark taulua.hpt

Log when car runs: 113 runs now high cylair.hpl

Log with next start: 114 does not start.hpl

[GHuggins]

Well lets see and this is just the very few things I looked at. Your injectors aren't set up correctly. You missed a table. The vacuum offset table needs to be one value all the way across. I honestly didn't check to see if all injector data was put in - if not it needs to be - can't cut corners on this. Injection timing really needs to be 0ed out on the rpm setting - this will help with the high VE numbers. You have the baro torque model tables, so at least you can make it drivable. ETC scaler needs to be maxed since you're technically running a 180mm tb now. Min air will need adjusting. Min fail needs to be set to 8100 - this will help with the hot restarts along with everything else being fixed.

I really hope you didn't plan on having a heavily modded setup like this and just plan on leaving most things stock. With a setup like this you're literally going to have to dial everything in. Just hope you can get the throttle under control to the point to where it moves easy otherwise she's just a death trap with sudden snappy over-opening movements.

[cc-rider]

Damn. What a crazy build! Following this thread. Tuning nightmare Im sure. But what a build.
In Finland too. Need to get it ready soon to consume that cold fall air and let those blowers run!

[kingtal0n]

Start with fuel pressure as low as possible. I would go 40psi baseline. This will keep the pumps and voltage happy in boost. Use a reference so it will be even lower at idle and cruise. This is how factory turbo cars such as nissan skyline and toyota supra are from the OEM baseline at idle around 36psi fuel pressure (44psi minus reference 15 to 18"Hg of vacuum). This will reduce fuel temperature. This will be safer and promote lifespan for fuel components. It will also most importantly keep the pump flow rate as high as possible in boost.

Like Ghuggins says if you dial the injector stuff and phase in properly it will help make the VE table sensible. Remember you can adjust the Injector settings to control the shape and feel of the VE table. I think As long as the torque model is close and commanded airfuel is matching actual air fuel in the VE table then it is very close to where it should be.

[blindsquirrel]

Sure thing brah, just tune it like it's a Skyline, you'll run 9's no doubt.

You have less than no idea how these Gen 4s work and what happens when you try to finger-fuck injector data like that.

[kingtal0n]

Sure thing brah, just tune it like it's a Skyline, you'll run 9's no doubt.

You have less than no idea how these Gen 4s work and what happens when you try to finger-fuck injector data like that.

What are you talking about. Where did I say finger fuck injector data. I said to make it correct. I said match commanded and actual a/f. How is that fucking anything.

Quote me where I made a mistake. I do not see any mistake.

[kingtal0n]

The 95-2002 Skyline engine, 95-2002 Toyota Supra engine, and many of the 2002-2007 LS style engines share almost every detail in common besides plug placement, external feature and displacement.

They all have:
1. computer designed combustion chambers which mimick efficiency that approaches calculable limits i.e. they respond to identical timing per power to displacement ratio e.g. 22-24* btdc naturally aspirated peak cyl pressure
2. computer designed main girdle support which guarantees the weakest link is the girdle and not the internals, e.g. The crank will walk and mains will fail before the pistons do
3. Oil pan main girdle support built into the design of the block
4. Modern, reliable Oil system based around front mounted oil pump
5. Coil over plug with modern reliable high mileage capable coils that provide enough energy forced induction
6. Sequential Port injection, injector phase, and with appropriate injector placement and manifold runner length with suitable design for forced induction in street cars
7. Factory OEM pistons all made of the same brittle materials and similar cyl-wall clearance for reliable daily driving and rapid cold start performance

I challenge anybody to find a difference in the make up design engineering or tuning efforts between these engines in a hobby installation.
They will all go 200,000 miles with a turbocharger at hobby level output figures.
With port injection the fuel pressure and spray atomization is not an issue. OEM always deposit fuel to the closed intake valve in a puddle. IF we change that to spray later it is at wide open throttle when the air velocity is helping to mix and vaporize the fuel, which is generally not recommended for a daily driver IMO but could be done the same way for all these types of engines.


I believe you may be reading into this statement I made

if you dial the injector stuff and phase in properly it will help make the VE table sensible. Remember you can adjust the Injector settings to control the shape and feel of the VE table.

Pay very close attention now. Read carefully. It says "If you dial the injector stuff and phase in properly it will ... make the VE table sensible". That is my recommendation, to dial it in properly.

Then , I said this:
"Remember you can adjust injector setting to control the shape and feel of the VE table"
I am not recommending that you adjust the injector setting to change the shape and feel of the VE table. I am saying that if the VE table looks all fucked up then clearly then the injector stuff must not be dialed in properly. They need to correct and to do that requires setting up the injector stuff to change the look and feel of the VE table properly.
In other words, I am reminding people that when they set the injector stuff they are changing the shape and feel of the VE table- which is absolutely true. This has obviously been overlooked previously. They need to know that this is possible so they can identify the problem: Injector is not setup properly.

[Pekka_Perkeles]

Ok, got it finally running quite well.

I re-checked injector information from customers receipt and it stated flow matched set of EV-1100-S injectors, about 1044 cc at three bar (45 psi). While it still does not make any sense having VE with that high values, I decided to reduce both VE and Injector Flow Rate at 50%.

Then I fixed MAP values and it now shows 100 kpa KOEOff. I also made suggested changes (but not yet lowering base pressure) and got it running but with 3000 rpm idle, even with lowered RAF values. So I changed Min Fail back to stock. It did run then nicely. Also the Cylinder Airmass values are much better aligned.

I made some other mods as well, to fine tune it a bit more. I also made first drive, although very slowly and not far as customer said I would need to check the auto trans oil level. I also noticed fuel leak at the back of the car. And yes, it has new spark plugs now as well. Throttle response seems to be stockish, at least when using less than 50% pedal.

First good run (stationary): 125 long run ok.hpl

First drive: 131 first drive.hpl

Latest tune: 152 vikakoodeja pois se?k? idle alemmaksi.hpt

I will probably lower the base fuel pressure to 3 bar sometimes later after driving it a bit more.

[edcmat-l1]

What are you talking about. Where did I say finger fuck injector data. I said to make it correct. I said match commanded and actual a/f. How is that fucking anything.

Quote me where I made a mistake. I do not see any mistake.

Start with fuel pressure as low as possible. I would go 40psi baseline. This will keep the pumps and voltage happy in boost. Use a reference so it will be even lower at idle and cruise. This is how factory turbo cars such as nissan skyline and toyota supra are from the OEM baseline at idle around 36psi fuel pressure (44psi minus reference 15 to 18"Hg of vacuum). This will reduce fuel temperature. This will be safer and promote lifespan for fuel components. It will also most importantly keep the pump flow rate as high as possible in boost.

Like Ghuggins says if you dial the injector stuff and phase in properly it will help make the VE table sensible. Remember you can adjust the Injector settings to control the shape and feel of the VE table. I think As long as the torque model is close and commanded airfuel is matching actual air fuel in the VE table then it is very close to where it should be.

Sure thing brah, just tune it like it's a Skyline, you'll run 9's no doubt.

You have less than no idea how these Gen 4s work and what happens when you try to finger-fuck injector data like that.

Few weeks back he was in a guy's thread about a C5 with return to idle issues. He was telling the guy to use the PIDs to fix it (completely wrong) then a couple weeks later he's in another thread and said "I don't tune drive by wire". WTF

[kingtal0n]

If you reduce the fuel injector size in the fuel injector size table, the ECU will think the injectors are smaller and so for any given VE value it will inject more fuel. So then you cut the VE in half I guess which counter acts that, sort of.
However now the computer thinks the injectors are half of what they really are and consequently the VE table is half of what it really is, which means when the computer estimates torque it will estimate half the torque of what it really is. If the transmission goes to shift and the ECU gives only half of the required line pressure for half of the torque it may slip and ruin the transmission.

I would put the actual fuel injector size in the file. And add a channel for torque (engine torque and/or transmission torque) so you can determine whether the airflow model is accurate within some common sense reasoning. Unless it is a manual transmission in which case the impact might be dealt with in a minor way elsewhere. (It is set to auto in the file so I assumed automatic)

Did you actually verify the fuel pressure with a mechanical gauge on the fuel rail?

When inputting injector flow rate it needs to match the flow rate at that pressure. So for example if the injectors flow 1000cc @ 45psi then you would not put 1000CC into the injector size table when the fuel pressure is higher. This might be common sense to you but I am making sure to cover each detail anyways.



Here are some question I would be asking if I see this file:

-Is there some reason you can't put the actual flow rate of the injectors into the fuel injector size table? (what is the actual problem when you tried this) Just the g/cyl? Did you look at engine torque model yet?
-Cylinder charge temperature bias seems crazy bananas to me
-Not logging accel pedal position for DBW seems like a mistake
-Timing map seems insanely aggressive when trying to accelerate at low rpm. For example: 6:58:45:749 there is 45 degrees of timing at 1700rpm 55KPA. The main spark advance table is crazy advanced to my eyes. That region should be closer to approx maybe 25~degrees ish I would normally expect.
-200*F coolant temp is good
-why does the VE tables(s) seem like its in the 1,000's? What are those units? I understand the VE table is not actual VE of the engine, its just a fuel table and subsequent torque predictor for some ECU, but why are there literally 10 times what I normally see in a typical VE table? Is this part of the problem? Do other files look like this? I am curious if this is typical or not, or what sort of values you are expecting.

-Why does the VE table appear to stop at 175KPA when it seems to have a 300KPA map? Or even a 2-bar OS (which normally stops around 207KPA). I'm curious the reason you did not load a 3-bar OS instead? Not available?
-EQ ratio for power enrichment and Boost enrichment is not tuned, its just a solid same number everywhere
-Commanded air fuel is slightly rich but commanded eq ratio is 14.7, how can there be two different commanded air fuel ratios?
-Why don't I see a wideband in the channels, are you trying to tune a forced induction engine with aftermarket injectors without a wideband?
-Fuel pressure channel is reporting extremely low values of like 1 or 2psi. Not sure what this will do if anything
-I am curious why the tune file is opted to select "supercharger not fitted" when there is clearly superchargers fitted
-Why does map estimated maximum say 105KPA when it will rise much higher than that?
-Timing spikes like crazy sometimes 6:59:43:943, I do not allow engines to have this behavior, maybe the overspeed/underspeed tables are too aggressive (they don't look *that* bad) but I would investigate this to help smooth idle quality.


I don't tune Gen4 ecu. I am just curious and since all engines and computers basically work the same way you can take the behavior of a carb engine and apply it to an EFI engine so stuff like bouncing timing and power valves and transmission pressure and fuel pressure is intuitive to investigate. And this file and logs seem to have alot of needed investigation.

[kingtal0n]

Few weeks back he was in a guy's thread about a C5 with return to idle issues. He was telling the guy to use the PIDs to fix it (completely wrong) then a couple weeks later he's in another thread and said "I don't tune drive by wire". WTF

I don't tune HPtuners ECU. I didn't design it nor did I code it so I have no idea how it works other than intuition based on how all computers in the world work. That is why I am able to tune any computer even if I never tuned it before. I am willing to say I don't know.

If the rpm is changing there is a derivative in play which is the D in PID. Therefore PID is always a concern and always effective in any frequency changing situation to prevent overshoot and undershoot. The return to idle problem is an over or undershoot problem which is a D Derivative problem. There are many work arounds besides PID of course but it is a valid suggestion when you are 20,000 miles away and cannot physically access the vehicle for something to trial and error.



If you spent more time helping people instead of chasing me around trying to make me look bad for being honest and intelligent you would not have so many useless garbage posts which add nothing to threads.
I have 0.1% Error on this forum and many others. You cannot refute statistical interface nor provide proof otherwise. Your efforts are futile and only serve to derail threads and populate threads with garbage empty posts.

[edcmat-l1]

I don't tune HPtuners ECU. I didn't design it nor did I code it so I have no idea how it works other than intuition based on how all computers in the world work. That is why I am able to tune any computer even if I never tuned it before. I am willing to say I don't know.

If the rpm is changing there is a derivative in play which is the D in PID. Therefore PID is always a concern and always effective in any frequency changing situation to prevent overshoot and undershoot. The return to idle problem is an over or undershoot problem which is a D Derivative problem. There are many work arounds besides PID of course but it is a valid suggestion when you are 20,000 miles away and cannot physically access the vehicle for something to trial and error.



If you spent more time helping people instead of chasing me around trying to make me look bad for being honest and intelligent you would not have so many useless garbage posts which add nothing to threads.
I have 0.1% Error on this forum and many others. You cannot refute statistical interface nor provide proof otherwise. Your efforts are futile and only serve to derail threads and populate threads with garbage empty posts.

Then why are you here????

You're a complete and utter fool. The joke of this place.

You don't understand how this stuff even works.

You can't tune drive by wire. You don't know the difference between PIDs and airflow adder tables.

You don't understand what stoich is.

You don't know how to calibrate a wideband.

You don't know what VE means.

Nearly everything you post is wrong.

You're a hack. You're talking about manipulating injector data right here in this thread. That's hacking shit.

Your own tune file out of your own vehicle is a complete hack job. It's so bad it almost has to be a joke.

[GHuggins]

OP - what size blowers are those? I thought they might be 1.9's but looking closer are they smaller? The sudden kick and tb size should be all you have to worry about tuning in but all of that will require entire remapping of open loop tables and so on. You'll need to increase the baro tables for the tb size for throttle control. It'll feel stock, possibly even powerful until you put it into a light traction area and then you'll pray to God that you can control the throttle cause she'll open up on her own at that point. This is why the baro tables being tuned in are so important. Make sure manifold volume is right too.

[kingtal0n]

Yawn

Nobody knows anything. This is the higher attitude one must take when attempting to invent, understand, and progress.

If you don't understand something that doesn't make it wrong.

If you think you understand something you are wrong.

Anybody that thinks they know anything is fooling themselves.

The mature thing to do in public venue is admit you do not know anything.

[Pekka_Perkeles]

They are 1.9 liter Eatons.

I think that the reason why the idle tune seems to be ok with rather stock settings in E38 is because there's an external throttle controller.

It is tunable as well, but at the moment I'm only playing with E38 settings.

http://www.torquerush.com/x2.html

[Pekka_Perkeles]

By the way, I'm not familiar with tuning those baro tables. I guess I have to do some googling here.

[GHuggins]

Look for post by myself or Jslic on here since he's the one that had them added in. Basically the way they work is they define airflow based off of throttle opening, so bigger tb's with a blower = more airflow and necessary to raise the numbers. Will make a 1000hp build drive just like a stock fully controllable setup. Hjtrbo also has some post on them where he fixed some off idle airflow errors.

The sudden punch from both 1.9's with both tb's will be the main thing to control. You're running some bigger pulleys but they're still just providing air for 4 cylinders. I assume there's a cross bridge to keep air the same to both banks? Definitely an interesting build

[Pekka_Perkeles]

Started to dial in VE in low throttle, low MAP, low RPM areas and I think I got it pretty good. But the end-result now with good fueling was a very bad idle.

It was oscillating so much that without touching throttle pedal, it eventually died. I noticed typical symptoms, i.e. idle airflow PID controller was not in sync with idle spark control and maybe RAF was wrong as well.

150 idle seilaa.hpl

I think I went too far by desensitizing idle air control so I went back with some stock values (not all). I also changed some idle timing values and made RAF smaller. Then it started snowing so I didn't take log anymore but it felt like a perfect now with the latest tune. Very stable idle with park/neutral and with reverse and drive as well. I will continue tomorrow, if weather permits.

180 palautettu kaasul?p?n asetuksia.hpt