View Full Version : Mapping questions
I originally sent this message to chip, but he said it would be a shame not sharing with other forum members. All you people are very welcome to add your opinion/experience to this discussion. :wink:
Hi there my friend.
Hope you can answer me some questions.
Engine mapping. Been reading quite a lot about it lately. Particularly Dave Walker's book on engine management. I have now a better understanding of engine mapping and standalone management systems.
What I would like to know is what's your opinion on Emerald M3DK system and IIRC you said it's kind of old and slow to work on?
How do you map a car on the street?
How can anything be decently mapped without rollers with brake?
Do you left foot brake the car while you adjust the injection map at given load and speed site(otherwise RPM's would instantly rise and you can't hold the engine, I mean you can but only at one load which gives a given RPM)?
That will be enough for now. :lol: Still have many questions tho.
Anyway thanks for your help in advance.
Cheers
David
PS: I may have written some things a bit clumsy, hopefully you'll understand.
Hopefully now its on the forum you'll get input from other people who map too not just me. :)
Yeah I am not a fan of the emerald, I've mapped quite a few cars on it now and the reason I dont like it is that I like to make lots of small changes while mapping and with the emerald when you want to edit a cell you have to do so, then press to say you want the ecu to be updated with the value, then it says are you sure, then you hit enter, then it goes and does it and then comes back and tells you it has done it and asks you to hit return to acknowledge that it has done the thing you entered for it to do, then set it to update, then confirmed you wanted to update.
Just far too many keypresses just to make a simple change IMHO, feels very cumbersome compared to Autronic or Mtech or Omex which I map all of often and much prefer.
Mapping a car on the street I use det cans to make sure that there is no detonation occuring and I use a wideband to see the fuelling, sometimes I also use a phormula knock detector as well.
It depends what ecu I am using wether I use the datalogging or just do it from memory as im running it up and down, Autronic SM4 for example the logging is so detailed and allows you to revist the map exactly as it was live at the time, so I make a great deal of use of that.
Yes I use the left foot to hold load sites, or to slow progress through load sites, and I also use hills to slow the cars progress through the rev range if its a powerful car (as otherwise if its something like my 500bhp or so nova its so quick through 3rd gear you dont have time to really see what is happening and in 4th gear you are repeatedly doing 130mph which isnt good for your license), not really an issue with an N/A car with only a couple of hundred bhp though as they are so slow to accelerate anyway.
If you get fast ford magazine from a couple of months ago I did a few pages in there on DIY mapping, will save me typing it all out again, lol
sideways danny
26-06-2011, 04:14
emerald is clunky. Not nice to use on the fly at all. Most other systems allow you to use a single key press. Omex uses plus and minus buttons, canems uses page up and down
I've cut and paste this from paddys ITB build thread. Canems is what I've chosen to use on mk1's and similar.... as it does everything you'll need without having dozens of unused features and at what I think is a decent price
This car has had the very first Canems ECU with the new software version (v2.10)
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/availablenow.jpg
More control features
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/setup.png
More accurate accel fuel control
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/accelfuel.png
and bigger main maps
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/210ignmap.png
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/210fuelmap.png
I've had some new firmware specially written to allow use of the standard mk1 flywheel/trigger pattern
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/renix2.png
And it runs.....
http://www.youtube.com/watch?v=1rMFRwRz1F4&feature=related
@chip: So am I right thinking that with Emerald takes more time to map a car? In his book Dave says it takes somewhere around 3 hours to map a car or even more.
How much time do you usually need to adjust one cell, don't you "cook" your brakes in the meantime?
Do you use laptop/keyboard or do you use trim box/wheel?
If you get fast ford magazine from a couple of months ago I did a few pages in there on DIY mapping, will save me typing it all out again, lol
That would be great. Do you have it scanned or can you tell me where to get it?
@sideways danny: Yeah, I heard few good things about Canems already.
How do you map a car on the road Danny?
Laine_16v
26-06-2011, 10:28
Ktec fully mapped my car on Canems in just over 2.5 hours
sideways danny
26-06-2011, 14:34
I've been at Emerald when wobba had his car mapped, and had the conversation with them about how un user friendly the software is. They said a lot of people say the same but they dont really "get it". They have made several newer versions of the software but havn't released anything to the public yet.
Dave Walker maps his way and the software is designed around this. When you've learned another way you can't make it do what you want.
How long it takes depends on the car and spec really. It's aways a far faster process on the dyno though. We did a 172 from scratch in 1 hour 20 mins the night before FCS '09 and it remained with that map for over a year until it had some major changes made. 3 to 4 hours is about right for a customer car though.
sideways danny
26-06-2011, 14:35
oh, and i regard road mapping as a temporary measure only. it's "good enough" until you cn get on a dyno and do it right
@chip: So am I right thinking that with Emerald takes more time to map a car?
Yes, its more cumbersome than most other ecu's so as a consequence it takes longer.
In his book Dave says it takes somewhere around 3 hours to map a car or even more.
Depends what you class as mapped TBH, an hour is normally enough for me to get an N/A car to be what most people would consider mapped, but you could spend 100 hours and still find more things to improve if you are using a fairly sophisticated ecu with good control over transient fuelling etc, not to mention there are temperature corrections to map accurately etc. Getting a car to drive nicely in one set of conditions is easily, getting it to drive perfectly in all conditions is anything but.
If you really want to do the job properly, the 3 hours dave mentioned would just about map the voltage compensation table perfectly for example, but most people arent going to care about that the way that a car manufacturer would for example.
A manufacturer spends literally thousands of hours working on a map, especially the backup tables.
So it really is impossible to give a "one size fits all" answer to that question.
But yes 3 hours is a reasonable ballpark on the rollers to get most of the map done pretty well though in most cases.
A lot of the time mappers spend far longer than that sorting out minor problems though with home built cars, if you have some wiring fault or sensor fault you can be paying for expensive rolling road time just to find that.
And it even catches those of us with a lot of experience out too. Once I spent about 40 mins of paid time on a set of rollers I rented fixing a water leak on my own car for example!
How much time do you usually need to adjust one cell, don't you "cook" your brakes in the meantime?
You dont typically hold a cell totally still on the road, just slow progress enough for you to see what is happening, the only time I spend a few seconds on one cell is when I am on the rollers.
Do you use laptop/keyboard or do you use trim box/wheel?
Laptop
You can only really use a trim box on the rollers and even then I see no real advantage over a laptop anyway, even emerald has a live adjustment screen you can use once you are held on a load site on the rollers.
If you get fast ford magazine from a couple of months ago I did a few pages in there on DIY mapping, will save me typing it all out again, lol
That would be great. Do you have it scanned or can you tell me where to get it?[/quote]
This issue:
http://www.fastfordmag.co.uk/2011/05/26/july-issue-on-sale-now/
oh, and i regard road mapping as a temporary measure only. it's "good enough" until you cn get on a dyno and do it right
Depends on the application, if you have a turbo car which is knock limited on ignition and EGT limited on AFR, the rollers are less useful, if you want to optimise the ignition on an N/A car though, or do the off load sites on a turbo car to the Nth degree then rollers are useful for that to.
I always do the final checks to the mapping on the road or track or runway though even after I have spent time on the rollers.
Rollers utterly cannot accurately replicate the way that the airflow characteristics of the engine change with roadspeed if you are talking 150mph pulls down the back straight at bedford like some of the cars I map are capable of, so the only way to know what the fuelling will really do in those circumstances is to do that speed and check, it can be quite surprising sometimes what happens to AFR's due to aerodynamics and road speed.
Ive mapped cars for example where the AFR gets a little leaner as speed increases to a point (you'd kind of expect this anyway due to "ram air" type effects) but then as the speed increases further it goes completely back the other way and goes very rich again, so it might be leaner in 3rd than 2nd, but then richer in 4th than it is in 3rd.
Depends what you are trying to achieve though, generally I am chasing 95% of the available power but with a nice big margin of safety as it tends to be trackday or road cars I am mapping, where as if its a competition car chasing the last few bhp can by far the most important factor.
Also with car type too, on an N/A car safety isnt generally much of an issue, the most powerful tune is normally a very safe one anyway, so its more on big boost turbo cars that I am talking about where I will dial in extra safety at the expense of power.
^^Thanks.
New question. Can car with ITBs be mapped with MAP sensor as main load sensor? Ofcourse with hoses from every runner attached with T pieces together to one main hose which would go to MAP sensor. Is this possible and if it is, is it better than Alpha-N mode?
^^Thanks.
New question. Can car with ITBs be mapped with MAP sensor as main load sensor? Ofcourse with hoses from every runner attached with T pieces together to one main hose which would go to MAP sensor. Is this possible and if it is, is it better than Alpha-N mode?
Yes its possible but IME the problem is that joining the 4 together doesnt provide an accurate enough vacuum signal as you have 3 cylinders that are at atmospheric pressure and one that isnt, and you lose the resolution of how far into vacuum it goes as the other 3 drag it up towards atmo.
So I find ITB's work far better with TPS as main load.
zmaster2k
28-06-2011, 16:54
isnt that how Rstuning managed to get the 172 to run on the stock ecu with ITBs?
ran a vac feed from each inlet to a "vacuum chamber" (looked like a swirl pot to me) then ran that off the map sensor?
zmaster2k
28-06-2011, 16:55
So I find ITB's work far better with TPS as main load.
so having two load sites is of no benefit? would of thought it provided more data?
or is it over complicating things?
isnt that how Rstuning managed to get the 172 to run on the stock ecu with ITBs?
ran a vac feed from each inlet to a "vacuum chamber" (looked like a swirl pot to me) then ran that off the map sensor?
That is how they got a map sensor signal as far as I am aware yes, but wether they have MAP as main load (the question that was actually asked) with TPS as secondary or vice versa I am not sure about, you would have to ask them.
On my ITB corsa (and how my ITB clio was too actually) I run both MAP and TPS and I can select which one to treat as the main load and which as secondary, and the percentage of influence that each has on the transient fuelling conditions etc
I have tried both ways round on several different ITB cars (including my clio) and like I have already mentioned, in my experience they work better with TPS as the main load rather than MAP as the main load.
So I find ITB's work far better with TPS as main load.
so having two load sites is of no benefit? would of thought it provided more data?
or is it over complicating things?
You seem to be getting VERY confused there mate.
Firstly, load sites refers to the number of individual cells on the map, typically this will be a couple of hundred.
Secondly you seem to have confused me saying "MAIN load" with "ONLY load"
I find that MAP values can be useful as a percentage of the input used as a combined load value to determine transient fuel control on an ITB car for example.
Just because you have selected TPS as your primary load, doesnt mean that you have to ignore MAP.
zmaster2k
28-06-2011, 17:17
yes very confused was refering to "load sites" as inputs not the actual cells (my bad!)
that actually makes much more sense now pmsl! proberly should learn to use my eyes LOLOL
So just to clear up, you still ultilise both inputs (map and tps) yet choose which one has priority?
yes very confused was refering to "load sites" as inputs not the actual cells (my bad!)
that actually makes much more sense now pmsl! proberly should learn to use my eyes LOLOL
So just to clear up, you still ultilise both inputs (map and tps) yet choose which one has priority?
It depends on the ECU, but typically yes, you will have one as a main load, and the other will be a multiplier or a secondary table.
And then for transient fuelling conditions you will often be able to set the relative input values of each, so if it sees an increase in TPS it will add more fuel during the change based on the rate of change (im not talking about the fact that the new cell in the load table has more, I mean during the change itself and related to the speed of change, like a throttle pump on a carb), likewise if it sees a decrease in vacuum (or increase in absolute pressure to be more relevant to turbo applications as well) it will add an amount of fuel based on the rate of change of MAP.
Some ECU's will only do one or the other transient change, and some will do both with you being able to set the percentage of each that takes precedance in determining that a transient situation is occurring.
Transient fuelling of course has utterly nothing to do with making peak power, but does have a notable effect on drivability, especially at small throttle angles and low rpm, and also economy when driving under a varying load.
To get the transient fuelling spot on like most manufacturer cars are takes far far longer than the time spent on any map that you will get from any normal aftermarket tuner, you can easily spend several months full time working on the transient fuelling if you have suffient controls to do so, standard ECUs tend to have very detailed transient fuel control but often on aftermarket ECUs its very basic.
Megasquirt is one of the better ECUss in this respect, it runs rings around most other aftermarket ECUs
A modern standard ECU though (ie not a clio mk1 one, lol) will have many many tables dedicated purely to attempting to get transient fuelling correct.
The main thing that people need to remember, is that if you are paying 500 quid for someone to map an aftermarket ECU, you really cant expect the same thoroughness in the mapping as a manufacturer spends potentially as much as millions of pounds on the calibration for.
Aftermarket ECU's are far more basic than modern standard ECU's and aftermarket mappers are only aiming for a car to drive what seems ok to the driver and to make decent power, if you actually accurately monitored the transient fuel AFR's on pretty much any aftermarket mapped ECU you would find it falling short of the same car when on standard management, partly due to the relatively crude controls in most aftermarket ECUs and partly due to the fact the customer doesnt want to pay for a few thousands hours of mapping and the use of a climactic chamber etc!
So on ITB'd car having both MAP and TPS sensors helps good drivability(MAP being main load sensor at low revs and TPS being main in upper revs)? How do you map a car like that? Two injection maps?
So on ITB'd car having both MAP and TPS sensors helps good drivability(MAP being main load sensor at low revs and TPS being main in upper revs)? How do you map a car like that? Two injection maps?
It can do, although the gains are less than on a shared inlet.
It depends on the ECU how you do it, you cant normally just add an extra fuel table in, you are normally constrained by how the ECU is structured n the first place, although some very high end motorsport ecus have a lot more choice.
To add a bit more detail to my last comment, if you look at the screens Danny posted for example:
Here is your transient fuelling:
http://i133.photobucket.com/albums/q59/sidewaysdanny/Canems/accelfuel.png
Its very mickey mouse compared to the sort of control you get within a modern standard ECU or even when compared to much more sophisticated aftermarket options like megasquirt with its detailed transient control and X-Tau algorythm to predict inlet manifold wall wetting etc.
All the CANEMS does it is just lets you set 5 different threshold values, then an enrichment time for each one, then has a second set of settings that then tweak these values based on RPM that allow you apply only a percentage of the enrichment rate for only a percentage of the enrichment time.
Properly "idiot proof" stuff that just lets you do some very very basic control which gets you into the right ballpark and thats about it.
If you showed that to someone at a car manufacturer compared to a modern ECU they would fall about laughing at you trying to actually get a car to fuel right with just that sort of over simplified algoryhtm to work with, but despite how very crude it is and the fact you will obviously never get the sort of results you can on more sophisticated management systems, the bottom line is that it will be MORE than upto the job of running any car that anyone on this site is likely to build for trackdays etc, just accept that its not a very sophisticated solution, cause when you are actually driving the car you really wont notice the difference.
The trick is to not got too obsessed with how things should be perfect, because like I said, unless you are talking mega sophisticated ecu and thousands of hours of mapping they never will be, aftermarket mapping is a very approximate science in reality when compared to manufacturer ECU calibration!
Powered by vBulletin® Version 4.2.5 Copyright © 2024 vBulletin Solutions Inc. All rights reserved.