Torque Model: Normalized Torque, Optimum Torque Table and Torque to Load Discussion.
There is a lot to cover here including lowering normalized torque to increase power and associated effects (e.g. on transmission and tcu tunes).
The formula/heuristics of how the ecu internally calculates torque.
The internal torque calculation versus actual output of torque and associated errors P061A and P061B.
Reducing normalized torque will increase the actual torque, increasing will decrease the torque. You only have about 50 torques to be off from the actual or you will get p061a or b. but for example you could reduce your normalized torque to 855 to get 30 more.
I believe normalized torque is multiplied by optimum torque times .01 (e.g. optimum torque is a percent). This number is the amount of torque being modeled at any given cell. If you exceed this or go beneath this by more than 50 I believe you get an error and limp mode.
Torque to load is poorly named. I believe this is the commanded desired load at given cells, e.g. the number in the tables are engine load. this is more important table than many realize and potentially more important than spark tables...
The normalised torque sets the scale of what represents 100% load. The torque is calculated based on the torque to load conversion. If the normaliser is 1000Nm and the vehicle load is 100% then the torque to load reference will be looked up and there will be a percentage that the normaliser is multiplied by to give the torque.
You need not modify the Normaliser unless you plan to exceed it, often it is set with headroom OE. You need to make sure your monitors are within what you make with appropriate headroom.
if at 50% Torque the load is 93%, which is 48% of the normaliser value, if the normaliser is 1000 Nm then the torque is 480 Nm.
Attachment 116981
This is interesting, I'd like to hear more.
From what you said it seems like I could simply lower my Normalized Torque and not touch any other torque data and it would simply boost power across the board?
I've only been tinkering with HP for a couple weeks, and only two test loads to the car, so very much a noob. I understand most all of it, but multiple torque tables & values and the Normalizer is very confusing.
What 4wheelin said makes it even more so: 50% Tq & 93% Load = 48% on the Normalizer, but if the Normalizer was 2000 would it not = 960? And 960 what? Torque? Is that supposed to be the torque the engine then provides? Or is it a table stating what it thinks it's providing? How does the 960 value make less actual torque at the tires?
Other confusing torque items:
Wheel Torque vs VSS, which seems obvious but the data in the table says otherwise.
Vehicle Mass Max Tq?
Max % Torque vs RPM?
Testing Max Tq?
Torque Monitor Maximum?
Max Torque in Emergency Mode?
Indicated Torque Limits?
Optimum Engine Torque (Monitoring)?
Internal Torque Max?
What do they actually represent, and what is their function in life?
I've searched the forum and online pretty much all day but only netted a headache.
Thank you
Good Clifftoo! You start this :) I am lazy and i will copy my posts from your thread.
"Torque management is pain.. I don't still get it. Normalized torque is like theoretical "100%" my car has 600Nm. The car will never go over 80% of engine torque (i don't know why... TCU?). If i change Normalized torque, torque to load table will be out of everything, because it will follow 0-100% of engine torque. I have get best solution by leaving NT 100% to original 600Nm and i only edited torque to load table. But, it can get funky also..."
"if i raise the NT100% (Normalized Torque) from 600Nm to 800Nm and try to calculate whole optimum torque, torque to load and torque limiters to match the "new" 100% maximum torque (NT100%)
so, how i think it works
0-600Nm = 0-100% (torque to load table)
0-800Nm = 0-100% so all in torque to load table is wrong. We have to make calculation to the whole table *1,33333 (600Nm*1,33333x=799,998Nm)
It should be correct? Then i lowered torque limiters, because they are much more higher now (limiter based to 0-100% not to Nm [0,75*limiters]). It should go to close to the default NT600Nm, IF i don't want to use that extra torque, but no... It will get very odd. Shifting more aggressively and not like full throttle."
We have "sameish" understand of this. Optimum spark will change the torque calculations and it can be a mess. The base spark is like 0% and optimum is like theoretical 100% or some sort of multiplier between NT0-100%? What is the correct way to add more air/boost and how it will alter our torque. speed density? Someone trick MAP to show wrong pressure to gain more, i don't like that. Same with lowering the NT100% i think it's wrong way to do it. Maybe this because TCU will limit? Sorry for many many questions.
Keep the Normalized TQ as it is because thats the 100% torque percent the ECU makes calculations based of it, only change it if you are going to hit that limit.
All the other tables you mentioned including the Optimum TQ are calculating the percentage based of the Normalized torque.
I like to look at Optimum TQ as a lookup table for the Torque to Load map.
The picture 4wheelinls posted is the right way to look at it but you wont reach those desired percentges of torque unless you raise the other torque limiters.
Don't over think it, its straight forward.
I did leave the Normalized Torque alone, because of this thread, but did tamper with most of the others. It was a stab in the dark because it's not straight fwd. Maybe to you because you know it, but I can only guess what each of those torque charts and settings mean. Sure I got it working nice, but just luck and I could not say why it works and there's no way it's ideal.
I sometimes write work instructions at my work, and they always come back because what seems obvious to me, because I've done it, is not to someone else who has not. I've used the following example to describe the process, an instruction to drive a car: Turn the key to start it, apply the brake, shift into gear. The user: Standing outside the car, holding the key and turning it various ways waiting for it to start.
Edit:
Ironically, I just had a perfect example: I had to explain how to wire up a device to a data collector at work. I'm at home. Describing via text was a fail, despite a simple description, then a detailed one. I ended up drawing a simple diagram in photoshop and sending it.
So even an accurate description can easily fail if the person doesn't know the nomenclature or has a good grasp of electronics. The picture did the trick and instantly made sense.
So far, from what I gather, Throttle position feeds Torque to Load, which then points to Optimum Torque, which then points to the Normalizer. Ok, now what. Lets say all that pointing ends up at 400 on the Normalizer, what does that do? Does the 400 go to another chart that says 400 = xx mass air flow and the throttle then drives to do that?
If I raise the values in Optimum Torque by 10%, which ends up 10% higher on the Normalizer, will gain 10% more torque at the wheels?
If instead I raise the Normalizer value by 10%, which would give the same #, I lose power?
If I/we knew wth the Normalizer is and what it feeds, it would make sense. As it is, it does not. This is what I meant about standing by the car turning the key. It makes sense to you, but without more info I can read what you say a lot of different ways and there's a ton of possibilities for how it all plays out.
Surely there must be a chart somewhere that shows what each item in my previous post is, or maybe all the items in the HP Tune Editor, what feeds into each one, and where it goes from there? I'd imagine if such a chart exists, it would make everything so much easier for everyone.
Even if it were just an example for each chart, like; if you raise/lower all the values in this chart by 10%, and no other changes, X will happen.
How about we start with the example chart 4wheeling posted.
It looks to me like the Engine Torque % (wherever that comes from) in the Torque to Load chart ends up as a value in the Optimum Engine Torque chart that is basically the same #. So 50% Engine Torque % equals a value of ~93-100 depending on RPM. That value plugs into the the Engine Load % column of the Optimum Engine Torque chart and nets a value of ~50. So 50 in and pretty much 50 out.
It seems the only thing happening between these charts is a very slight change in torque based on RPM. It appears to me that one of the charts could be eliminated and you'd lose nothing.
You could also go a step further and get rid of both charts since the slight variation in torque based on RPM is very small and I don't see how it matters. So whatever feeds Engine Torque % in the Torque to Load chart could be plugged directly into the Normalizer. So once again; 50 in, and 50 out to the Normalizer.
So help me understand if I'm correct, or what is wrong with my image of what's happening here.
To make matters more confusing for me, the Engine Torque % in the Torque to Load chart on my car is 1.5-80, not 0-100. I can only assume Engine Torque % is driver demand, but the Driver Demand chart on mine outputs values of 0-100.
Based on working with these platforms here how it goes;Quote:
Originally Posted by chevota
Throttle position (Driver Demand) -> Optimum Torque -> Optimum Torque (monitor) -> look for the percentage on Torque to load map -> ECU asks for the load (which calculates the boost needed)
All these should be within limit of [Normalized Torque + Governor + Max Vehicle Mass Torque + Indicated Torque limiters)
Thats why I said don't overthink it because too many limiter over there and some of limiters are not even included in HPtuners but you can find them using Winols.
On your car if you are aiming for more than 80%, you can change the axis of Torque to load and Optimum Torque.Quote:
Originally Posted by chevota
Torque to load is the inverse of Optimum Torque
Thank you, that does help in a way. The problem is it also conflicts with other info on this same thread so I have no clue which is correct. Is there really no source of info that just lays out the how and why?
I wish I could make something that doesn't make sense and not tell anyone anything about it and get $700 for it. Even a toaster comes with instructions.
Also, no clue what is meant by Torque Load is the inverse of Optimum Torque. My understanding is one chart simply points to the other, no inverse involved. Which one points to which is the aforementioned conflict. 4wheeling said Torque to Load feeds into Optimum, you agreed. Then you said, what I read as, the reverse. The way I see is it the Torque to Load numbers can't point to a % of the Normalizer because the values go up to 200. So 200% of the Normalizer is the request? 885ftlbs x 2? I don't know the answer, which is the whole problem.
It appears I'll most likely be forced to learn this on my own, from scratch, with a blindfold on. So how about this: Since all my tunes go into limp mode, I need to get past that so I can see what my changes actually do.
Yes my first tune, total shot in the dark, worked really well, but now it won't. All my tunes go limp mode on just a little gas, that one goes limp on decel. No clue why and all I have to go on is the following:
Post #2, clifftoo; "You only have about 50 torques to be off from the actual or you will get p061a or b."
I have no idea what that means. 50 off from what compared to what? I'm only asking for one torque value, which is the Normalizer. Is that not correct?
"Actual" to me is crank, and it's much further off than 50 now, so actual must mean something else? What is meant by "actual" and what exactly has to be within 50 of it?
I meant by inverse is that Optimum Torque (percentage) points to the (Torque Percentage) axis in Torque to Load map not the other way around as 4wheelinls posted and No there isn't any where you could find a clear explanation only you figure this out by practicing and testing.Quote:
Originally Posted by chevota
Optimum Torque = Torque Percentage requested
Torque to Load = Torque Percentage convert to Air Load which then goes to other monitors which then makes the request
Regarding your tune issue with limp mode, attach your file and will try to pin point the problem.
That is very kind of you, much appreciated.
I attached the one I made last night and tested this am. This one worked for a bit, meaning ~ one mile, but limped on decel. I gave it maybe 75% throttle for two seconds at 3K rpm at 60mph (traffic limited) and about two seconds after letting off the gas it limped.
When it limp modes it's not a limp I'm familiar with, you can't drive it at all. If I try I get maybe 5mph tops and it's bucking and kicking hard.
Some tunes I've made limp the second I lean on it even a little, like normal acceleration from a light. Just a little throttle and low rpm, boom, limp. Very frustrating...
I was curious about the Normalizer, so I bumped it up by 25%. I lost a small amount of power, not 25% worth, but I also got limp mode...
Attachment 117913
You may be hitting limp due to reducing the EGT tables. I'd suggest returning those to stock and re-testing.
I had assumed if you can disable them, it wouldn't matter?
Plus they were holding back power. My first experiment was a bunch of changes but zero power gain, it just felt a little different. I didn't touch Ex Temps.
So I started from scratch and the only mod was reduced Ex Temps, and it helped a lot. So if I moved them back to oem I can only assume I'd lose everything again.
Plus it's just a reference chart and I'd imagine it has only one purpose, which is to limit power when the chart is referenced and the value is too high? Just an assumption, and so far assuming has triggered limp mode many times.
Doesn't everyone lower Ex Temps?
I put the Ex Temps to normal, still limped.
Test this fileQuote:
Originally Posted by chevota
Attachment 117938
Thank you for taking the time to make that! Very much appreciated!
It didn't limp on me, and pulled decent once it gets going, but overall 0-60 is still slow at ~5.2sec. If the bottom end wasn't so soft I'd imagine it might hit or beat the stock 0-60 # which is apparently 4.8sec.
This is why I was pushing all the numbers so far, because even where I had them it's still not cutting it, but I have to stop it from limping first.
I have made some that look pretty close to what you did, but fail fail fail.
T̶h̶a̶t̶ ̶t̶u̶n̶e̶ ̶I̶ ̶p̶o̶s̶t̶e̶d̶ ̶w̶a̶s̶ ̶t̶h̶e̶ ̶l̶a̶s̶t̶ ̶o̶n̶e̶ ̶I̶ ̶h̶a̶d̶ ̶t̶r̶i̶e̶d̶,̶ ̶w̶h̶i̶c̶h̶ ̶s̶t̶a̶r̶t̶e̶d̶ ̶a̶s̶ ̶o̶n̶e̶ ̶I̶ ̶g̶o̶t̶ ̶f̶r̶o̶m̶ ̶t̶h̶e̶ ̶f̶o̶r̶u̶m̶.̶ ̶I̶ ̶l̶o̶o̶k̶e̶d̶ ̶w̶e̶i̶r̶d̶,̶ ̶a̶n̶d̶ ̶r̶a̶n̶ ̶e̶v̶e̶n̶ ̶w̶e̶i̶r̶d̶e̶r̶,̶ ̶b̶u̶t̶ ̶i̶t̶ ̶d̶i̶d̶n̶'̶t̶ ̶l̶i̶m̶p̶.̶ ̶S̶o̶ ̶I̶ ̶t̶r̶i̶e̶d̶ ̶t̶o̶ ̶s̶m̶o̶o̶t̶h̶ ̶i̶t̶ ̶o̶u̶t̶ ̶a̶ ̶b̶i̶t̶,̶ ̶a̶n̶d̶ ̶f̶a̶i̶l̶.̶.̶.̶
That tune is a whole other story...
I simply got my tunes mixed up and the one I posted started as my OE Tune and ended up with all kinds of changes, then I keep changing this and that based on things I see other people doing. When it works, meaning I can floor it first, before it limps, it's not bad. Not great, but better than it was before I tinkered. Looking at what you did I don't see how the power down low was lost. I'll try it again tomorrow because the only thing different was I normally test at 5 AM and it's ~50F out, but I tested yours as soon as I got home and it was more like 90F. Heat never bothered it that much, but we'll see tomorrow.
I see the stuff you put back to oem, or close to it, so I'll leave those be and see how far I can push the other things.
Fyi the HP Scanner shows that with the oem file loaded it reads 399hp & 433Tq, almost right at the oem 402hp 440Tq. So it appears the ECU thinks it's doing what it's supposed to, it's just not actually doing it.
Like that Normalizer, it could simply be something reading wrong and it thinks everything is good.
Your file said it was making 501Hp and 503Tq, and % wise it feels about that much over the oem file.
This is why I was working my way towards 1100ftlbs, so I'll actually end up in the mid 6-700 range? And that's why I have questions about the Normalizer etc. Like can I somehow ask 125% of Normalizers 885 value, or do I really need it to be 1100? But if raising it lowers power... Very confusing.
Thanks again for that file, I will try to follow what you did and simply increase values. Maybe boost the low end. Wish me luck!
Glad to hear that it worked !!
for the low end you have to throw some mods on to your car, add timing and fuel in a way to compensate any torque changes and to avoid limps.
If it was a stock 63 you would be able to achieve 600 - 700 without mods but then its risky, yours is E550 so I wont advice you pushing it that hard without the right mods and tune.