Learning to write maps
- Thread starter mcleansc
- Start date
why would they do that when they can just charge you for the tune
looking back at how i started....first, learn how to read logs properly
lots of things are cause and effect so you will need to learn how to differentiate whats the cause and how it affects everything else
then.....you should be able to tune the car with the below...
- basic knowledge from google/ebooks/hp academy course
- basic n54 tables from rfp videos
- n54 specific stuff from v8's spreadsheet/dme document/shitloads of googling
You're not going to find anyone amongst the popular tuners that will be able to explain how to correctly tune the N54 becuase none of them know themselves. I've seen maps from most of them and there's not a single person going further than what you'll see in the old COBB maps. But that's pretty normal in the tuning world as those podcasts explain.
I think the only change from COBB is that most of them will use 3 cell MAP scaling and the MHD+ boost ceiling.
The podcasts also a do a good job of explaining why accurate modelling is important. I'd say if you listen to those podcasts and scroll through the COBB tunes and compare the values to stock vehicles you'll have more knowledge of tuning cars than someone like Doc Vu.
I think the only change from COBB is that most of them will use 3 cell MAP scaling and the MHD+ boost ceiling.
The podcasts also a do a good job of explaining why accurate modelling is important. I'd say if you listen to those podcasts and scroll through the COBB tunes and compare the values to stock vehicles you'll have more knowledge of tuning cars than someone like Doc Vu.
Glad the thread has a bit of life and some of the way more knowledgeable guys have pitched in. From what I've been learning it appears the difficult part tuning the N54 with Tunerpro is more about understanding what the many tables mean and how they are interconnected than the knowledge and understanding of timing, AFR's etc. That could be the case for an aftermarket ECU right enough (I've never actually tuned anything).
Thanks for those podcasts carabuser, HP Academy has some great stuff, I'm working my way through their podcasts currently but I'll jump to those straight away. I've actually just bought their reflash tuning course.
Thanks for those podcasts carabuser, HP Academy has some great stuff, I'm working my way through their podcasts currently but I'll jump to those straight away. I've actually just bought their reflash tuning course.
From the little bit that I've messed with some of the aftermarket ECU's (very limited), they make things much easier to figure out. You don't have multiple tables doing calculations in the background that affect AFR's, Boost, Timing, etc, etc.
And yes...you are 100% correct, the part that throws me for a loop on the Tunerpro stuff is all of the different tables that go into calculating load, and how if affects all of your other tables. It's not like you can just say "I want 20psi of boost at this amount of throttle opening , and use this amount of timing". That's how some of the aftermarket ECU's I've messed with in the past have been test up. Maybe not quite that basic, but you get the idea. The timing table controls the timing. The boost table controls the boost. etc, etc, etc......you don't have all kinds of other tables in the background affecting things. I'm sure it's great for OEM BMW cars, where you want things as nice and smooth as possible, in all conditions, in all situations........but to get that end result, there is a lot of things going on in the background.
Last edited:
Yeah understanding, for example, what the values in the table of the P-Factor actually represent is what Im struggling with. On the V8Bait spreadsheet he states he likes to half it, I like to understand what I'm changing before I change it if that makes sense.
On that note of boost control hopefully someone can answer this in here. I understand our cars are load based boost control: the DME has a load target and uses the turbos to attain that target. To increase boost we increase requested load. Now obviously the WGDC table is set up for 8psi from factory. The PID system has some authority to intervene when off target due to barometric pressure, IAT etc. but within a small scope. Most tuners and tunes (I've compared the MHD V6 maps to stock) I've seen seem to increase the adder and ceiling, leaving the base table essentially totally stock. Am I correct in assuming by doing this we give the PID a larger scope to increase WGDC on order to hit the load target? But lose resolution?
So going off what I've said (I don't like changing things I dont understand) can I simply leave the PID stock and fine tune the WGDC base table? Why is it most tuners opt to make substantial changes to the PID side of things but leave the WGDC base table essentially totally stock?
Thanks man. Theres so much I don't know. It seems the more I'm researching this the more I discover how little I actually know lmao
hp academy has a section that explains pid control
this youtube vid makes visualizing what pid actually does a lot easier
the way i would explain it is that
wgdc table/adder tells you how much wgdc approximately is needed to reach a certain load/psi
pid controls the rate it gets there at
scale the map, adjust the load tables, half the pid tables and play with wgdc base/adder, then play with pid tables again...just remember to reset adaptions
Boost Control Strategy
https://bmw.spoolstreet.com/threads/explaining-wgdc-values.3679/
https://bmw.spoolstreet.com/threads/wgdc-calculation.2835/
Last edited:
Here's a graphic example of how P, I, and D impact the output of whatever process they're controlling.
A simplified way of thinking about it:
P: affects how quickly you hit set point
I: affects offsets between actual and set point
D: dampens the response between actual and set point.
A simplified way of thinking about it:
P: affects how quickly you hit set point
I: affects offsets between actual and set point
D: dampens the response between actual and set point.
Attachments
PID changes are only needed if you are changing the turbos though. Most issues that people see with unsteady PIDs are caused by the base table being incorrect.
Touching the airflow or PID adder tables is not a good idea. The airflow adder can be rescaled for higher MAF levels but is largely redundant since the base table also accounts for target airflow.
The PID adder is also not something you want to mess with on stock turbos. For a large single you need to increase it in the lower airflow ranges but just blindly raising it is just a band-aid for a poorly set up base table.
The best approach is to just set the car up without any PID adder at all and get the base table as close to perfect as possible. Then enable the PID and make adjustments if necessary. Outside of spool, the PID should only be adding or subtracting to compensate for inaccuracies in the altitude and temperature correction models and maybe some wear in the turbo components over time.
Touching the airflow or PID adder tables is not a good idea. The airflow adder can be rescaled for higher MAF levels but is largely redundant since the base table also accounts for target airflow.
The PID adder is also not something you want to mess with on stock turbos. For a large single you need to increase it in the lower airflow ranges but just blindly raising it is just a band-aid for a poorly set up base table.
The best approach is to just set the car up without any PID adder at all and get the base table as close to perfect as possible. Then enable the PID and make adjustments if necessary. Outside of spool, the PID should only be adding or subtracting to compensate for inaccuracies in the altitude and temperature correction models and maybe some wear in the turbo components over time.
Some great information in here not just for me, so appreciate everyone taking the time to post. This is exactly what I wanted from the thread, not a spoon fed how to (not saying that wouldn't be great ha).. but really informed people showing the way and helping everyone understand the process a bit better so thanks for that guys.
Finally decided to bite the bullet and start 'writing' a map. Now I see theres MHD and MHD+. What am I best starting with? From reading posts on the forum am I correct in saying you need to create your own WGDC base table on the MHD+ as it works off the boost by gear rather than the load tables?
My advice would be to start from the stock file. Start by just increasing the load targets then work from there.
Avoid the MHD+ WGDC tables. Load by gear and the MAC solenoid control are great but the rest can be left alone.
Thanks for the info mate I'll do just that. I'm currently using the latest XDF that has the MHD+ custom boost tables etc, but not active. Really daft question but I take it that's ok to use as I'm assuming it'll use the normal tables when the custom boost tables aren't active? Or should I be using the legacy XDF and bin?
As an aside I'm considering making a video series of my attempts. Not instructional (other than learning from my mistakes ha) but just cataloging the changes I'm making and what they're doing. Would anyone be interested in that? Although it would be akin to watching a monkey use a typewriter
if i increase load past 180, do i have to scale the timing table load breakpoints further then? or as i understand, anything past 180 would just use the 180 timing. is that okay?
or say for my custom tune that i assume is using the 3d boost ceiling, i'm at 19psi/full boost and it's under 180 actual load. with requested load out of the way. what is it that's changed to make that possible? that being changing the load/boost relationship from what's in the v8bait sheet. boost limit multiplier? the single cell boost ceiling?
apologies my knowledge is super surface level i'm just trying to get an understanding of some mhd tables while im doing a general efi tuning course, not actually driving on anything i've messed with lol.
Last edited:
Correct if you go past the last breakpoint then you'll just stay on the value at the end of the table.
If you're reporting less than 180 load at 19psi then there's probably map scaling being used. That is the biggest problem you should fix. You only need the scale the map abobe 21psi. Ideally you don't tune the boost curve using the boost ceiling, that is going to cause problems, just set the boost ceiling out of th way and use the load target to get the engine load where you want.
thank you! I wont use the 3d boost ceiling at all, just set the normal ceiling to 1.5. and use load target the way it was intended to be used. that sounds better to me.
I guess I was wrong, looking at more logs now actually i see my load up to 190-95... so i dont think the map is scaled? regardless tune is locked so ill be starting fresh anyways.
on the timing table again, so if i'm targeting say 195 load, it using the 180 values is fine then? but what if you go up to say 205 load? would standard practice be to then scale the breakpoints further? and adjusting the values correspondingly. i guess same question for vanos as i see it only goes to 170.
your help much appreciated
Good stuff in this thread.
I started with a stock IKM0S file and did the following:
-Increased load tables to 183 straight
-Removed the polular limiters
-Smoothed BRO in the area of WOT to ~92%
-Flatened both BPTM to 1.01
-Zeroed Load Target Offset
I'm still experiencing issues with the boost target not being straight or proportional to the load target. Could this be related to the Torque Reduction Factor (RPM) kl_kuehln? Would it be a good idea to set it to 1.0 straight above 2000 RPM?
Also, is there any other table, aside from the ones already mentioned in the public XDF, that influences the boost target without having load on the axis?
I started with a stock IKM0S file and did the following:
-Increased load tables to 183 straight
-Removed the polular limiters
-Smoothed BRO in the area of WOT to ~92%
-Flatened both BPTM to 1.01
-Zeroed Load Target Offset
I'm still experiencing issues with the boost target not being straight or proportional to the load target. Could this be related to the Torque Reduction Factor (RPM) kl_kuehln? Would it be a good idea to set it to 1.0 straight above 2000 RPM?
Also, is there any other table, aside from the ones already mentioned in the public XDF, that influences the boost target without having load on the axis?
Last edited:
