Author Topic: Mapping - Who\'s tried it themselves?  (Read 15527 times)

Offline Prof

  • Frustrated Inventor
  • MRC Full Club Member
  • Major Contributor
  • *****
  • Location: Camberley, Surrey
  • Posts: 4,749
  • Karma: +51/-0
  • Rotorhead
    • View Profile
Mapping - Who\'s tried it themselves?
« Reply #15 on: June 18, 2009, 08:49:44 PM »
I don\'t know how good the timing maps are on the basemaps for PFC - I think if you\'ve got a basemap for most of the same mods it\'s going to get you close, but even a little bit out on timing makes the engine quite unresponsive and if it\'s advanced rather than retarded it could land you in trouble :(
 
At the very least make sure the timing at idle is set correctly as per the service manual - at least you\'ll not the starting point is spot on.
"Plus this engine is probably one of the most volumetric engines going! It passes more air than a herd of cows eating vindaloo\'s" - courtesy of AtomicRex
 

Offline chrisJDM

  • Major Contributor
  • ***
  • Posts: 3,242
  • Karma: +8/-0
    • View Profile
    • http://
Mapping - Who\'s tried it themselves?
« Reply #16 on: June 18, 2009, 09:56:03 PM »
this is a really interesting thread, ive always wanted to get into mapping but dont have a clue at mo and dont really know where to start, great thread guys :Thumbsup!

lee mills

  • Guest
Mapping - Who\'s tried it themselves?
« Reply #17 on: June 18, 2009, 10:06:56 PM »
hopefully some of the more experienced guys can chip in with some info to help us less experienced guys

Offline chrisJDM

  • Major Contributor
  • ***
  • Posts: 3,242
  • Karma: +8/-0
    • View Profile
    • http://
Mapping - Who\'s tried it themselves?
« Reply #18 on: June 18, 2009, 10:12:25 PM »
yep would be great :yes

Offline ianelvar

  • Contributor
  • **
  • Posts: 407
  • Karma: +20/-0
    • View Profile
    • http://www.virtualtechie.co.uk
Mapping - Who\'s tried it themselves?
« Reply #19 on: June 18, 2009, 11:03:13 PM »
I\'ll do my best with what I know.

Lets start with timing:
We have a matrix with RPM on the X axis, and manifold pressure on the Y axis. I believe the apexi has a 20x20 map, but my example uses 12x12. In the screen grab below the RPM is represented as usual, and manilfold pressure pressure is represented in KPa, or Kilo-Pascals. 0 being a complete vacuum, and 100 being atmospheric pressure (or thereabouts) Anything about 100 is obviously boost, which you\'re probably used to seeing stated in bar. Bar uses the same units as KPa, but is normally stated as pressure above atmospheric. So running .8 bar boost will give 180 KPa.
In our map below, (from the \'rocco turbo btw) the figure in each of the bins represents ignition advance. It\'s that simple. So the idle area of the map is in the bottom left, at low RPM and KPa. We don\'t need much advance here. Moving to the right you\'ll see the cruising area. Here you\'ll see that we\'re building advance quite rapidly up to 3000 rpm. After this we level off the advance through to peak torque, then add maybe a couple of degrees after that up to redline. Moving up the map towards atmospheric, or full throttle on an NA, and the \'transistion\' area on a turbo, you\'ll see that that we lose some advance overall, but we keep roughly the same curve up and down the RPM. Moving into the boost area, we lose timing fairly rapidly. THis is dependant upon compression ratio, combustion chamber shape, and a whole bunch of other factors I won\'t go into now. You\'ll find some great info on making a good base ignition map for the rotary in the posts Prof gave us earlier.

As it goes there\'s no black magic or voodoo here yet. The trick is obviously what to put in the bins, but the rules are fairly simple once you get to grips with them!
The megasquirt tuning software also gives us another neat way of changing the map. You can get it to generate a 3D map, with advance on the Z axis. The little red X can be navigated around the bins with the arrow keys, and the advance value can be modified up or down with the Q and W keys, and when running a green blow moves around the map with RPM and load, so you can tune realtime. I\'ve no idea if the apexi software gives you something like that though. Maybe someone can shed some light here?
I\'ll cover fuelling tomorrow when I get some time.

Offline chrisJDM

  • Major Contributor
  • ***
  • Posts: 3,242
  • Karma: +8/-0
    • View Profile
    • http://
Mapping - Who\'s tried it themselves?
« Reply #20 on: June 18, 2009, 11:08:11 PM »
great info, my boost controller reads in kpa, i just let my mapper set it up, i never bothered touching it as i dont understand it

Offline ianelvar

  • Contributor
  • **
  • Posts: 407
  • Karma: +20/-0
    • View Profile
    • http://www.virtualtechie.co.uk
Mapping - Who\'s tried it themselves?
« Reply #21 on: June 18, 2009, 11:17:39 PM »
Thanks. Lol, just realised how many typo\'s I made, and can\'t edit them now. Sorry about that!

Offline slingo

  • Contributor
  • **
  • Posts: 246
  • Karma: +10/-0
    • View Profile
Mapping - Who\'s tried it themselves?
« Reply #22 on: June 18, 2009, 11:47:13 PM »
Quote from: Prof;617759
I don\'t know how good the timing maps are on the basemaps for PFC - I think if you\'ve got a basemap for most of the same mods it\'s going to get you close, but even a little bit out on timing makes the engine quite unresponsive and if it\'s advanced rather than retarded it could land you in trouble :(
 
At the very least make sure the timing at idle is set correctly as per the service manual - at least you\'ll not the starting point is spot on.

cheers prof i will try and find the thread its got a complete apexi guide with a datalogic

Offline BlitzBoy

  • Major Contributor
  • ***
  • Posts: 21,023
  • Karma: +41/-0
    • View Profile
    • http://
Mapping - Who\'s tried it themselves?
« Reply #23 on: June 19, 2009, 01:58:19 PM »
Hi Ian link for the Megasquirt forum on the US rotary site
http://www.rx7club.com/forumdisplay.php?f=153
"Piston Killa" 3rd Gen\'94 \'Oldone Racing Full Bridge Ported REW 13b engine with SCR seals\'. HKS T51 KAI turbo BBM mapped Motec ECU

JDS/HKS Street Class Championship Runner up 2011
Rotorstock winner 08,09
http://pistonkilla.blogspot.com/
old spec HB T51 470 @ wheels @ 1.2 BAR, 555 @ fly 10.46 @ 130 @ 1.25 Bar
new spec FB T51 482 @ wheels @ 1.15 BAR, 570 @ fly

New Project
"el fénix" RX2 by Mark & Mark

Offline ianelvar

  • Contributor
  • **
  • Posts: 407
  • Karma: +20/-0
    • View Profile
    • http://www.virtualtechie.co.uk
Mapping - Who\'s tried it themselves?
« Reply #24 on: June 20, 2009, 12:50:18 PM »
Thanks, I haven\'t been on there for a while and it looks like there\'s some interesting stuff going on. Anyone here played with Megasquirt? I think that\'s the way I\'ll go for my project, but I\'d like to see what the Apexi or Emanage are like to tune. Would any of you guys who have the software and live reasonably close mind letting me have a quick look over it sometime?

Offline ianelvar

  • Contributor
  • **
  • Posts: 407
  • Karma: +20/-0
    • View Profile
    • http://www.virtualtechie.co.uk
Fuelling
« Reply #25 on: July 06, 2009, 10:55:24 AM »
So, belatedly on to fuelling. On the FD we have a MAP (Manifold Absolute Pressure) sensor for measing airflow into the engine. It doesn\'t do this directly - we have to tell the ECU how much air is flowing at a given RPM and manifold pressure, and we do this by building a map. Just like the spark advance table, we have a Volumetric Efficiency (VE from here on) table which is used for calculating how much fuel to inject.

Cribbed from Wikipedia: "Volumetric efficiency in internal combustion engine design refers to the efficiency with which the engine can move the charge into and out of the cylinders. More correctly, volumetric efficiency is a ratio (or percentage) of what volume of fuel and air actually enters the cylinder during induction to the actual capacity of the cylinder under static conditions. Therefore, those engines that can create higher induction manifold pressures - above ambient - will have efficiencies greater than 100%."

So, to walk through the fuel map as we would when \'mapping\'...
Starting at idle makes most sense, so we\'re in the bottom left of the map. Low RPM and low manifold pressure - the throttle will be shut and the engine has to work hard to move air through the tiny gap left past the butterflies, and we\'re left with a partial vacuum in the cylinders, or chambers in our case, and hence a low volumetric eficiency. It depends on the engine - compression, porting, air cleaner, exhaust and even ignition timing all have a big part to play in determining the VE. Still, a low value here is sensible.

On to the no load range of the map. A slight crack open of the throttle and we\'ll start to build RPM. Volumetric efficiency won\'t be much more so the map will remain fairly flat as we move up the rev range, but will increase slightly if the throttle is opened more.

Time to give it some load. We\'ll look at the cruising area of the map next. This means 2500 - 4000 RPM and low to medium pressure. You won\'t find the VE increases much with RPM under cruise, but the VE map will start to ramp up sightly as the pressure builds towards atmospheric.

The fun part is always Wide Open Throttle. On an NA motor it\'s pretty simple - WOT yields 100KPa, and close to, but not neccesarily complete cylinder filling - 100% VE - and this will be the highest line on your map. Generally you\'ll find it increasing with RPM up to peak torque, and might drop off a little after that.

On a turbo motor it\'s a little more tricky. The 100KPa line is called the transition area - the point at which you actually start making boost. The volumetric efficiency obviously ramps up from here as we\'re pushing more air in than you\'d get if it was at atmospheric pressure. More boost = more air = higher VE.

So, the great question is, how the hell do I know what the volemetric efficiency is to start filling in the bins??? The answer is a wideband lambda sensor. If the lambda is showing lean (greater than 14.7:1) then you\'ve got too low a value in that area of he map, and you\'ll need to increase it. Too rich, and you\'ve overestimated cylinder filling and you\'ll need to lower that part of the map.

I should mention here that target Air Fuel Ratios (AFR) will be different for the vairious parts of the map I\'ve listed above. At idle and cruise you\'ll want to keep it pretty lean for good fuel consumption, so look for the magic 14.7. Some engine will run leaner, others won\'t run lean at all and you\'ll have to keep it slightly rich to keep it smooth. Under higher load up to 100KPa look to make it richer, maybe down to 13 or 12.5 - this is where you\'ll get best power. Under boost I normally go for 12 to 11. Any more and you\'re just wasting fuel. Further than 9.5 will yield a massive cloud of black smoke! If you have to run this rich under boost to stop detonation something else is wrong. Don\'t get tempted to use the fuelling as a band aid for other problems.

I should mention over-run too. Coming down form high RPM with throttle shut will yield the lowest manifold vacuum and lowest VE. If you fancy some lovey pops on over-run, aim for over 16:1 above 3000RPM. Trust me, it\'s good. Th E30 drift car has a straight-through side exit exhaust, and convincingly scares the bejezus out of anyone lucky enough to be on that side of the car. It\'s like a shotgun going off. Looks mighty impressive at night too :Giggle

Offline BlitzBoy

  • Major Contributor
  • ***
  • Posts: 21,023
  • Karma: +41/-0
    • View Profile
    • http://
Mapping - Who\'s tried it themselves?
« Reply #26 on: July 06, 2009, 11:04:25 AM »
good write up there Ian
"Piston Killa" 3rd Gen\'94 \'Oldone Racing Full Bridge Ported REW 13b engine with SCR seals\'. HKS T51 KAI turbo BBM mapped Motec ECU

JDS/HKS Street Class Championship Runner up 2011
Rotorstock winner 08,09
http://pistonkilla.blogspot.com/
old spec HB T51 470 @ wheels @ 1.2 BAR, 555 @ fly 10.46 @ 130 @ 1.25 Bar
new spec FB T51 482 @ wheels @ 1.15 BAR, 570 @ fly

New Project
"el fénix" RX2 by Mark & Mark

Offline Prof

  • Frustrated Inventor
  • MRC Full Club Member
  • Major Contributor
  • *****
  • Location: Camberley, Surrey
  • Posts: 4,749
  • Karma: +51/-0
  • Rotorhead
    • View Profile
Mapping - Who\'s tried it themselves?
« Reply #27 on: July 06, 2009, 11:34:19 AM »
Quote from: ianelvar;618251
Thanks, I haven\'t been on there for a while and it looks like there\'s some interesting stuff going on. Anyone here played with Megasquirt? I think that\'s the way I\'ll go for my project, but I\'d like to see what the Apexi or Emanage are like to tune. Would any of you guys who have the software and live reasonably close mind letting me have a quick look over it sometime?

If you are coming to RS6 I can let you have a look at the Emanage stuff. :Thumbs-up
 
If not, you\'re not so far away.
 
The biggest issue with the Emanage is finding out what you\'ve got in the stock or chipped ECU - as you can only tune on top of that as a starting point. Getting a log of the ECU before making changes let\'s you see what the ECU is doing as regards timing and fuelling - and then you can tweak the maps about a bit to compensate for the modifications made.
 
Because of that, you pretty much have to get things logged and build replacement maps for stuff like the airflow meter before you take them off, so using an Emanage on a peripheral port will be virtually impossible without some big guess work. To use it on a street or bridge ported motor where you can take off the AFM once you\'ve built a replacement map is more doable, and the map\'s available will allow a wide range of adjustments to be made allowing for quite a lot more boost or using a MAP sensor for speed density tuning for instance.
 
I tried the AFM replacement and whilst it ran reasonably well, the car was always a little lumpy compared to running with the AFM - a lot of that is down to the more accurate fuel metering at low air flow volumes (where the AFM is changing it\'s signal the most). It would also appear from some research I\'m doing into replacing the AFM with a MAF (hotwire) type sensor, that the overswing of the cone/flapper in the AFM is used to provide enrichment during rapid acceleration - it\'s not possible to do that with the AFM replacement map, so that leaves you having to use the table in the Emanage provided for that.
 
My plan is to program a little PIC microcontroller to spot the rapid Aiflow change and temporarily spike the output signal, where it would normaly just provide a straight lookup from the MAF signal output and translate to the expected voltage from the AFM.
 
That would leave the only other problem being the air temperature compensation. The stock AFM has a thermistor to let the ECU know how hot the air is so it can calculate air mass. The hotwire sensor adjusts it\'s output automatically for air temp - colder air is more dense and cools the wire more. The big question is would the MAF provide the correct adjust ment if a fixed resistor were used to let the stock ECU assume the temp was say 20 C - or would the air temp need sampling as well to provide additional correction - these are unknowns at the mo ;)
"Plus this engine is probably one of the most volumetric engines going! It passes more air than a herd of cows eating vindaloo\'s" - courtesy of AtomicRex
 

Offline ianelvar

  • Contributor
  • **
  • Posts: 407
  • Karma: +20/-0
    • View Profile
    • http://www.virtualtechie.co.uk
Mapping - Who\'s tried it themselves?
« Reply #28 on: July 07, 2009, 02:46:12 PM »
Thanks, I hope it\'s informative.

Cheers Prof, but I can\'t make RS6 unfortunately, already booked up :( Would be good to meet up sometime though, they\'re some interesting ideas you\'ve got. I\'ll give you a rundown of the MS stuff too if you want.
What do they give you on the Emanage for acceleration enrichments? Megasquirt uses the TPS, and you just have to set up an injector pulse addition for 5 x volts/microsecond bins. Implementing that in a PIC should be pretty easy though.
The hot wire idea is particularly intriguing. I\'m tempted to say that running a fixed resistor to lock the temp signal would work fine with the self compensating nature of the hot wire system.
A trick I\'ve used on MAP systems with lumpy idle vacuum (throttle bodies) was to add a little canister before the sensor. It averages the vac a bit and makes it much more manageable. Doesn\'t help if you\'re not making much vacuum in the first place though ;) What setup are you running?

Offline Prof

  • Frustrated Inventor
  • MRC Full Club Member
  • Major Contributor
  • *****
  • Location: Camberley, Surrey
  • Posts: 4,749
  • Karma: +51/-0
  • Rotorhead
    • View Profile
Mapping - Who\'s tried it themselves?
« Reply #29 on: July 07, 2009, 04:24:03 PM »
Emanage enrichment map uses Throttle opening % against RPM, and then has a seperate time to decrease row (back to standard map) along the bottom (default 20ms).
 
The engine makes about -0.5 ~ -0.7 BAR of Vac according to the gauge (on idle), but I think the issue is that the 3 BAR sensor is only operating over 1/3 of it\'s range in the vacuum part of the map, whereas the AFM covers 70% of it\'s voltage swing in that area of the map, with the top end of boost resulting in hardly much change in AFM signal - in fact the stock AFM bottoms out at around 11 psi - so additional fuelling has to be done on the MAP sensor from then on.
 
As for the hotwire trick - I am going to try a fixed resistor first, but will run with both AFM and MAF on a dyno to get the table setup, then I can just change the plumbing over - slap in my little PIC box and it should work straight off (famous last words).  I\'m hoping for slightly crisper throttle response and spool up from the turbo.
"Plus this engine is probably one of the most volumetric engines going! It passes more air than a herd of cows eating vindaloo\'s" - courtesy of AtomicRex