Apexi Powefc - Ljetro Vs Djetro

[paulr33]

Apexi PoweFC - Ljetro VS Djetro

For some time this debate has gone on with various opinions and ideas with no real concrete evidence/proof or logic to back it up.

There are two distinct differences between the LJetro and Djetro PowerFC (and presumbably other LJetro/Djetro ECU's) and we will discuss them as this is a hot topic.

Having since worked out the mathematics behind the Load axis calculations for both the Ljetro and Djetro versions of the PowerFC I can show you how they are both calculated.

First we should look at each base model and how it works;

Apexi PowerFC Ljetro Version

Uses standard airflow meter

20x20 Load Map

Other addtional supported airflow meters (selectable on menu)

Apexi PowerFC Djetro Version

Uses optional supplied Apexi Map sensor

20x20 Load Map

Other addtional supported map sensors (selectable on menu with correct voltage and offset)

Both units are scalable upto 1100bhp with supported modifications, sensors, injectors etc.

i suspect the ljetro version was created first, and then Apexi following demand added the Djetro version for some select cars as they found customers were migrating to other ECU's to map sensor support.

Ljetro Vs Djetro Load Axis:

Both have arbitary values for Load indiciation, Djetro has higher RPM as its from RB26 Djetro

Airflow Meter Ramp:

Voltage Ramp for Airflow meter load signal. Infinitely scalable, as we can rescale any given airflow meter.

The most commoin airflow meters are supported out of the box, Z32, Q45 and Apexi Power Intake.

Seemingly all the Apexi Power Intake does it alter the ramp values, which should yield slightly higher airflow load values, which should drop them on the map 1.3rows if that, ever so slightly.

Map Sensor Ramp:

Voltage Ramp for Map Sensor. Each sensor has a single Offset and nothing more. Fixed offset value

Optional Map Sensors include GM5bar map sensor and Denso unit's also. The Apexi one is in fact a Denso unit. The only map sensor supported out of the box is the optional Apexi one (denso rebadged).

GM 5 bar and other sensors need their scale and offset placed into the PowerFC to work.

Apexi Ljetro Load Calculation:

LOAD = CORRECTION * (16384 * Airflow lookup(MAFSV) / RPM)

Airflow is proportional to RPM so this is why we are divisible by RPM.

We obtain MAFSV from our running environment.

We lookup the Airflow value from the Airflow ramp table (viewable by datalogit) and use this value.

We lookup the CORRECTION value from the Airflow corretion table, for 4.48v and higher my correction is 84%

See ljetro-data.txt

LOAD = 0.84 * (16384 * 3598 / 4950)

The answer gives us 10003.57 and we can see for load point 14 we have an Load value of 9642.

The load value for point 15 is 11571 so we place our load axis on 14 with a value of 10003

As we can see here the load will always increase as the engine speed increases.

More and more air comes into the system and thus airflow load will always keep increasing until redline.

There have been doubts over this as many people see bog stockish R33 GTST's drop to load points 15 and shoot across.

This is due to the fact they are maxing our the airflow meter at 5.10 or very close to.

The reason it drops to load point 15 is because the majority of them have airflow % correction somewhere near 80% for 4.48v and higher.

Apexi Djetro Load Calculation:

LOAD = Map Sensor Offset * Map Sensor(PIMV)

We lookup the Map sensor offset and use this value (viewable by datalogit).

We obtain PIMV from our running environment

See djetro-data.txt

LOAD = 3289 * 1.95

The answer gives us 6413 and we can see for load point 8 we have an Load value of 5938.

The load value for point 9 is 11571 so we place our load axis on 14 with a value of 6681.

As we can see once pressure is fixed or doesn't increase any more, the load value doesn't.

This will result in a horizontal shoot across method once target boost pressure is reached.

This is because the Djetro version uses air pressure and runs a guesstimation method to calculate airflow.

As the engine speed increases more and more air comes into the system, pressure remains the same and the load axis remains the same.

The Djetro has two additional features which are;

IGN vs TPS correction

INJ vs TPS correction

These are there presumably to assist in correction for the throttle position sensor.

These will have an affect on the output figures but don't appear to adjust the load bearing.

These are much like Airtemp correction or water temp correction, they affect values but don't adjust load bearing.

So that's it in summary, a few myths cleared up. Im glad to have finally worked it out.

ljetro_data.txt

djetro_data.txt

[paulr33]

I've done a quick pro's and con's of each (my comments are marked as PR). These are the most debated ones

Ljetro

Pro's

-> Out of the box, works great, plug and play nothing more to add/change

PR: This is suitable for the majority of customers

-> Supports factory airflow meters and larger models

PR: This is suitable for the majority of customers and tuners (less fiddling the better)

Cons'

-> Some tuners/owners believe the airflow meter is a restriction

PR: Turbocharged cars have a restriction after the compressor outlet, not before.

-> Once you max out the airflow meter it can't show any more useful load to the ECU

PR: Once you max out a given airflow meter, move to a larger unit with higher resolution.

If you stay with an airflow meter maxing out (stockish GTST customers) then you basically have the same

load bearing as Djetro customers. Once you peak or reach max airflow (the most the AFM can show) then Load doesn't increase anymore.

This is dead obvious for stockish GTST customers who do a map trace.

Djetro

Pro's

-> No air intake restriction

PR: Turbocharged cars have a restriction after the compressor outlet, not before.

-> No airflow meters in pipework

PR: This can be helpful for large twin applications. It may also help avoid chuffing or compressor shuffling.

With correct pipework and bends you should be able to run twin MAF's or even a big single MAF relocated in the after FMIC pipework

-> Can't max out airflow meters

PR: This is a valid Pro. But once you max out any airflow meter simply change to a larger unit or more scalable unit.

Con's

-> You loose load point resolution once you reach target boost (horizonatal shoot across, as pressure never changes)

PR: This would be my main caveat of moving to map sensor system, the lack of tuning points, as the map isnt a 20x20 useable map anymore.

[justin911]

i have no clue in any of this pretty much, so i cant argue or agree... but speaking from a practical point of view hehe wouldnt it be better to dyno test them both for 2 seconds and then you'd have a definite answer as oposed to the answer on paper ???

So that's it in summary, a few myths cleared up. Im glad to have finally worked it out.

[paulr33]

there is nothing to prove here. all the information is in what i have posted and how the load axis calculations are different from the normal and djetro versions.

do you honestnyl think if i change my PFC to a djetro version i would magically make another 20rwkw on my 194rwkw gtst? of course not.

this thread simply puts down the exact method used for load axis calculation. there is no 3rd dimension for the djetor load axis. even I thought TPS was the 3rd dimension, but its not.

[MerlinTheHapyPig]

interesting read paul. good work

(where did you get all the data from?)

we've talked about MAP verses AFM alot recently

though i don't really think the argument of "loose load point resolution when reached target boost" plays that much importance in the scheme of things. For 3 reasons

1- I've noticed when running map traces that the stock ecu is reading more or less from the same load column when at full boost, so just becuase the load *can* show differences in air flow at full boost, it doesn't really need to (because load cell doesn't change with afm reading anyway)

2- Keep in mind at full boost, even if load axis won't change, RPM axis still changes, so the progression in cells across the rev range allows you to make adjustments under these conditions.

The only time this would present a problem is when holding a constant RPM and load is changing, I can't really think of a situation where this might matter unless say, you select 5th instead of 3rd and car takes a while to build up revs

and lastly

3 - Motec and other top ecu's use MAP as the primary load sensor. Infact Power fc's are the only ecu's i can think of that use afm as standard.

So, my belief is that there is nothing wrong with using an ECU that uses a MAP sensor, (mainly due to reason 3, but I have put some thought into this), and whilst the powerfc is probably one of the best value for money ecu's on the market for all-round performance, there are many others to choose from which are better (though mightn't be as easy to install)

And on another note, noone has any hard evidence that MAP sensors are bad anyway, it's getting a little annoying seeing all this anti-MAP stuff going around!!!

Edited June 17, 2006 by MerlinTheHapyPig

[brn8r]

i find that the djetro version will be more fiddly to tune as it won't compensate for climate changes as much as an AFM version would....e.g on a cold night your afm will show increase in voltage as the cooling of the hot wire is greater.....on a map ECU this is probably why most import tuners are mucking around with the tune between drag runs. For a street car i'd go maf and for a race spec car i'd go map.

Edited June 17, 2006 by brn8r

[justin911]

umm where the hell did my last post from 10 mins ago go........?

[KamikazeR33]

from memory map sensors use a temp sensor aswell?

also paul i remember sydneykid saying map sensors work worse with multiple throttle bodies (gtr)

[paulr33]

map sensor tuning becomes more difficult on multiple throttle bodied engines such as the Rb26 and the SR20 GTIR version as they have individual throttle bodies.

the djetro doesn't utilise an extra air temp sensor. the RB26 has a factory air temp sensor in the plenum for air temp correction, but this won't affect load placement on the load map.

on my R33 when i hit target boost in 3rd gear just near 3000rpm my AFM reads 3100mv and slowly increses to 5100mv just near 6000rpm. so for the AFM vesion the load keeps increasing until we peak the engine's efficiency near 6 grand. on the map sensor version my load would have remained static from 3000rpm until redline as the air pressure would have never changed.

[MerlinTheHapyPig]

on my R33 when i hit target boost in 3rd gear just near 3000rpm my AFM reads 3100mv and slowly increses to 5100mv just near 6000rpm. so for the AFM vesion the load keeps increasing until we peak the engine's efficiency near 6 grand. on the map sensor version my load would have remained static from 3000rpm until redline as the air pressure would have never changed.

I know that, my point is, who cares? revs are still increasing, so your ecu is still progressing through cells on the map.

Sure, there are more areas on the map where you can "theoretically" use at max boost with an afm based computer, but the real question is, do you actually need them?

I haven't seen anyone on this forum post any hard data that demonstrates this is a problem. And the fact that many aftermarket ECU's use MAP sensors, obviously demonstrate that the designers of the ECU don't see this as a problem!!

[paulr33]

this point of this thread wasn't to prove anyone right or wrong. it was merely to clear up a few myths and post up the actual mathematics behind it.

ive often wondered why most stockish gtst only max near load point 15 and ive also often wondered how you strectch or bend the afm load points, now we know, use the load ramp table and rescale, which in turn affects the final load bearing

[RB30-POWER]

so does the djet use the rb26 air temp sensor for injection correction though?

the density of the air is going to change a fair bit with temp.

[paulr33]

all apexi powerfc's use the airtemp for airtemp correction, that is

IGN correction for airtemp

INJ correction for airtemp

this applies for ljetro and djetro versions (and pro).

they influence more or less IGN/INJ for given airtemp values but won't affect load axis placement.

the R33 GTST is one of the only PFC's to have broken airtemp implementation, that is the factory car does not come with an airtemp sensor so it doesn't work. SW20 customers can purchase an additional sensor and connect it up to the PFC loom for working air temp sensor support depsite the car not having a factory air temp sensor

[MerlinTheHapyPig]

ive often wondered why most stockish gtst only max near load point 15 and ive also often wondered how you strectch or bend the afm load points, now we know, use the load ramp table and rescale, which in turn affects the final load bearing

most ecu's will allow you to change the load and rpm points (stock ecu can do this)

Here is the equivalent of a "map trace" on a powerfc for my car (i think this was third gear but may have been second). This is with the stock ecu, standard tune, standard boost, fairly stock r32 gtst

http://i2.photobucket.com/albums/y19/merli...o/tim_dyno1.jpg <- dyno with boost

anyway, i've found, most of the work is done in load point 11-12 and it tends to stay in the same load point for the rpm range that matters (2,800-6,400 rpm it's in the same load band). In fact, it's interesting that the stock ecu has the load points more spaced out higher in the "load" axis.

Anyway, what i've interpreted from these results are that under high-load conditions the "load axis" is less important than rpm (because it's remains in the same column anyway, even using an AFM)

[paulr33]

yes the load points are not linear, that is they are not 20 eqaul load points. as load increases so to do the scale of each load point