When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
You can increase the sensitivity on rear rotor until it becomes limiting, so that you know that the engine is not knocking, and than creep on with the front to the same. That way, you know at least a little bit that it is not in knock with to little sensitivity on the front one.
Are you sure its not as well Traction correction ?
I recently spent a whole day trying to get some kind of sensible reaction from the knock tables but eventually decided I was just chasing a phantom. Every stock ECU tuner previous to us has just disabled the knock system. We need a much better understanding of not only how it works, but what we need to pay attention to and what we need to ignore before we can have it do anything useful for us. IMO
One of my old friends is number one in tuning cars, as he said " he is in the back of all projects, but still no one knows that he is" he gets paid top money by hours to do fine-tuning,
very average-looking guy, come, take out laptop doing his own crazy thing, get paid and get lost.
Having said all of that, he is looking outside of the box, I remember a long time ago we had a conversation late at night drunk before sleep, about camshaft timing, as he said to me "No need to make them ALL right, a few wrong timing would increase flow for the other "
Indeed he is right, and proven on dyno he is! Not always putting all at the right time means you get the most performance.
He was able to hear knock by ear as accurately as the sensor well had developed spacial aids to make him work in the way he does.
So, if a genius doing something doesn't me an average/ amateur shall follow.
Back to the knock sensor, they should be mounted in very specific way, there is another thread way back between me and Team about it.
@motodenta is till right. The closed loop is the average of the two. so no, you can't be aware of any mismatch between two injectors.
I am glad it was depicted by someone. well done :D
A closed loop is the worst thing you can get and the most useless thing for pro tuners,
The O2 sensor as it is named, only can say how much O2 is in the exhaust, is very dume, and can not differentiate if the missed fire caused high O2 or leaking in the system introducing O2, if one rotor is lean while the other rotor is ******* rich and... etc.
ULESS have two O2 sensors for each bank/ rotor
Also as said need two separate EGT sensors.
Have more or less fixed KR issues. Getting some random 1-2 degrees pulled here and there under higher boost. I'll leave that be.
I've been losing boost past 6500-7000 though, so I've ordered a 4 port EBC solenoid to help with control.
Busted another coil - this time an OEM one - on trailing plug rear rotor. With how my setup is, I can replace it without taking out anything else. I am done with aftermarket coils - at best they work as good/long as OEM.
I've been losing boost past 6500-7000 though, so I've ordered a 4 port EBC solenoid to help with control.
.
Have you looked at how you might reduce backpressure ? I seem to remember you fitted a catback exhaust.... but 99% of those aren't free flowing enough for the boost levels you are now running.
I'm not doing anything else to the setup that will trade flow/power for noise. Not my kind of car. That being said the whole exhaust is 3", has a resonator in the middle and a (crappily welded) racing beat catback. I don't know how much more "freeflowing" it can get except for a stubby exhaust out the side of the car.
I'm not doing anything else to the setup that will trade flow/power for noise. Not my kind of car. That being said the whole exhaust is 3", has a resonator in the middle and a (crappily welded) racing beat catback. I don't know how much more "freeflowing" it can get except for a stubby exhaust out the side of the car.
understand. I thought my old setup was free flowing too till I had a custom system made which reduced backpressure by 5.5psi.
is it clear 3" or with smaller connectors ? those reduce the flow A LOT,
if its straight, you can help it by having the mid section 3.5 or even 4" as that is the part where the gases are hottest and need most space to create least backpressure. as they cool down the required cross section gets lower.
It is 3" from turbo outlet till the tips at 2* 3". No connector smaller in diameter, no tight bends. Car is really quiet and I intend to stay like this. On top of that I found that adding 8deg. of timing at idle kills some more noise and aids in idle torque for stuff like AC and heavy electrical loads.
Found it difficult to get more boost with this new setup using a 10psi wgt spring and a 3 port solennoid. Even cutting all boost ref. to the bottom port of the gate I found that I can't keep up boost to 85kpa past 5500-6000 rpm. Would not want to fit a tougher spring because I need that for trackdays and for when I can't get 100RON gas so I need to run as little boost as possible. So I moved to a 4 port MAC solenoid but that thing is plain nasty when it comes to control. I've been doing all my boost control on my own now, and with 3 port solenoids the control is easy to predict and linear. For this 4 port one I found I have to scrap my entire boost control strategy, pretty much turning my 5-10 lines of code into a cesspit of base target boost, base wgt dutycycle feedforward, a bunch of corrections based on rpm, tps, pre-tb boost, post-tb boost, KR, then everything is to be glazed over with a PID loop but whose parameters seem to require constant readjustment based on boost error.
Funny how I used to look at OEM stuff saying why do they need that, nowadays looking at myself needing at least 2/3 of those "useless" compensation tables.
New setup is indeed better. Who would have guessed. It feels good enough on 82-85kpa that I don't feel the need to push it higher.
Integrated a hard boost cutoff at 7700 rpm - wastegate reverts to spring pressure and BOV opens. This works wonders - I get to feel the torque just going away well before the ECU fuel cut-based rev limit hits at 8000.
I'm still tweaking the EBC to spool faster, now that I've got a 4 port solenoid that can boost till the moon I have to be careful. Log attached.
Last edited by ciprianrx8; Oct 12, 2024 at 12:20 PM.
Made my full time closed loop(FTCL) tune for this boosted build, yet to try it completely. If it works as expected then I feel I've achieved some great peace of mind. Maybe I'm a terrible tuner but found fueling the most challenging part in a boosted tune. It doesn't help one bit that the common approach so far was to mash the VE table(essentially just a fuel scalar) to get the desired OL AFR, when the wiser approach would have been to use the estimated manifold pressure table + injector flow vacuum multiplier to get that, and the VE would be left for minute changes(as it was meant to). Well too late for that, I've already spent hours on my MAF/VE tables and I'm not going to spend them again trying a new approach.
My experiments on an NA car have given me enough trust to believe that these tables shall be followed to the letter:
The large table - the small table = FTCL AFR target. Simple as, no other BS involved.....
The only gripe I have is the limited rich-side measurement of the factory WBO2. It won't go any lower than 10.89 AFR and for reliable FTCL I would want at least a 0.2 AFR separation between the richest point on the map and the richest the sensor can read. Asking for a richer mixture than you can measure is a surefire way to set your car on fire. Trouble is, I don't have the foggiest on where to start this. Probably by reverse engineering the ECU schematic and seeing if anything can be done hardware-wise to extend the domain of the WBO2 measurement? Maybe patch the input to accept an analog voltage from another O2 sensor + controller ?
If you find the AFR sensor calibration table, you could perhaps eliminate the internal heater controller circuit and use an external one. That would be my aproach since i have 0.0001% of your programming skills.
how does the BMW you have calculate fuel = VE or ?
Dug up 4 more useful tables for any boosted builds. I'll use these in place of me watching gauges in the summer not to beat on the car too bad when the weather doesn't allow it.
Is oil pressure supposed to be this high? I've never seen it go past a little past half way in mine.
Originally its either yes (7-8psi or how ever much the switch is.. ) or no pressure. Ciprian made it so that the gauge actually shows pressure in engine.
Original gauge is dummy one ... Oil pressure needs to follow engine oil temp (as temp goes up - pressure goes down) and and engine speed, higher rpm-> higher oil pressure.
His position on that gauge shows 6bar.
The "yes" pressure position is to low even for idle... So you have no clue with original if engine has enough pressure
The function basically wants to take a float (a number like -324.45, -45.4, 0.453, 23.42) and turn it into a number in the range of 0-255, like 0, 32, 43, 124, 200 and so on.This function is used a lot in the ECU, for lots of things, and for the most part it works. BUT IT DOES NOT WORK FOR NEGATIVE NUMBERS. Give it any negative number, it will return a 0. The 0 I always saw when trying to log knock retard. I am beyond pissed that the devs overlook this when making the firmware, they never ever used the PID either since they would have noticed that it doesn't report KR.
So now I shall not sleep until I make a patch for this function and allow myself to log knock retard... at long last.
I tested it out today. I feel I've achieved something important for this build and not just for it. Knock retard logging is now possible and I shall make good use of it.
I also get to define the resolution of this PID, the factory value of 0.31 isn't as great as I'd like it to be. On top of that I do my own logging setup too where I can account for this.
I have also validated the 2 sensitivity tables for the knock sensor readings, but I am still unsure if they really are per rotor or something else. Decompiling is hard. :/
10k kms post re rebuild stud insertion whatever. Been a while. I had to redo the stud install somewhat as I didn't know I had to put sealant on the stud nuts at the end. I was leaking coolant badly from there... to the point where the rear iron looked like Hulk.
BOV failed somehow as it is no longer sealing well into high boost. This I figured out as my ECU reported abs eng load went well into the 220% range at just 75kpa - this didn't made any sense; I wasn't having any vacuum leaks under vacuum, and my WGDC also shot up a lot to compensate, my STFT went -15% where it once was +- 5% (FTCL rocks). All points to the compressor dumping charge air back into the intake, from which some made it back into the compressor and some just went out the air filter. MAF reads both ways though(maybe not that great the other way around, but it reads). Have to see what I can do, sounds like an easy fix though(prob. torn diaphragm)
Seems I got away with the rattle of death... insert stays in place for now. Hopefully forever.
Nope, just chuck testa. Son of a b* insert shattered again, about a year and 9k kms of boosted abuse. Heard the all familiar rattle and knew 101% sure what it was and what happened.
This time the turbo didn't make it - 3 turbine blades were damaged. One is chipped and cracked, and 2 are just chipped. The damage is not deeper than 1 mm so I figured that the damage is light enough not to warrant a new CHRA... so that is that. Seeing this I said I'm not taking anymore chances again, so I took out the siamese insert out entirely, cut off all parts that prevented the rotors from spewing hot gas at each other and back in it went.
The theory behind these failures is as such:
1) lots of boost, lots of heat. No porting done. This creates a lot of heat in the middle iron, which is forced to flow exhaust gas continually w/o any break.
2) the first thing the exhaust gas hits is a wall that forces them to take another 90deg bend... and this wall is hit from both sides, all the time, non stop.
3) the insert just melts away, starts to chip apart, some bits come flying out through the turbine
4) eventually a large chunk breaks off which rattles inside which causes the rest of the insert to start rattling which makes an audible sound - game over, you lost.
For my next magic trick I trimmed the insert to remove anything that might break apart again.
This obviously is playing with fire. Literal fire. The rotors are now spewing hot gas at each other - this is obvs not good. I fear the side seals + springs, corner seals + springs will get cooked and lose tension or break apart. However the rotors are cooled with oil at least, and judging by the literal rubber seals inside the oil control rings, I suspect they don't get much hotter than oil does(in that area). So, by wishful extrapolating, the rotor tips don't get any hotter than 250-300C... or so I hope. Without the insert, I expect that temperature to raise to 400C. After a brief exposure to incoming exhaust gases, the seals then rub against the middle iron face which is coolant-cooled, and so, the temperature of that surface doesn't exceed 100C. In theory they would have plenty of time to cool down back again. We will see how long this lasts.
Don't know why this failure isn't more common - maybe others don't push this engine hard and for longer periods of time? I didn't bother with a turbo to use it for a few seconds now and then...
For the moment the car ran 3 fuel tanks w/o that middle wall in the insert, but I've dialed boost down from 85kpa to 75kpa and will leave it there.
