Better aftermarket computer - RX Turbo L .

[Outback bloke]

We could go around in circles with this for hours.

Power means poor fuel consumption , you better go and tell the international race teams that one given that pit stops are being allowed less and less .

That is right, they are trying to make their cars more economical with the maximum amount of power possible. Pretty much the same as it appears you are. Which brings me back to the question of what you are trying to build, a powerful engine or an economical one. You are talking about using top line engine management to achieve this economy with injectors twice the size of the factory ones.

Thinking about that, straight away says that, you are planning on using more fuel not less. If it is a safety/reliability issue so you don't lean it out when giving it a hard time then that is fair enough but there goes your economy.

I honestly don't understand people that try to build 1000hp cars to cruise around trying to get the best litre per hundred figures possible. If you want power then go for it by all means. Then when you are done with that drive your hybrid to work or the coast and leave the monster in the garage. It really is the only way a high power vehicle is economical.

[discopotato03]

No , thats not true at all .

The injector reasoning works like this . Many but not all of the aftermarket engine management system have injector drivers to suit saturated injectors - or high resistance ones if you like . The better ones have high current/low resistance/peak and hold injector drivers , the Vipec V44 dosent have these .

Now most turbo cars from the 80's used peak and hold high current injectors and drivers because they had to be larger to inject more fuel than the NA engine needed . The pintle valve in the PH injectors gets heavier as it gets larger so there is the need to have fine control over them for clean AFR's at idle and light loads . The reason why they are called peak and hold is because the drivers switch about 4 amps to open them and drop off to about one amp to keep them open for the desired pulse width . It made for "snappy" injector response with the technology of the day .

With low current or saturated injectors the available sizes in a form suitable for EA82T's are few .
I should also add at this point that sequentially fired injectors need to be larger than group fired ones because they only open once per four stroke cycle - in other words they have to inject all the fuel required in one pulse .
Group fire inject normally twice per cycle so they can be smaller because they get two bites at the cherry .
Group fire injectors inject with total disregard to inlet valve timing so at low to medium revs fuel control can never be as accurate as a sequential system .
The injectors I'm looking into will only cost me a few old school Skyline parts I have kicking around so no dollars involved . They are high resistance Mazda turbo RX7 O ring type but they can be modified with a little machine work to take an EFI hose and clamp . I think size wise they are either 440 or 460 .

Now to E85 fuels . The general feeling is that for a given output you need to have the ability to inject about 1/3 more volume to get acceptable AFR's .
Now lets just say that I manage to get my engine to crank out 160 flywheel horsepower which is ~ 50% increase . We'll make the assumption that it'll need 50% more fuel capacity and if the apple juice is used multiply that by 1/3 again . Is a picture forming ?

I know for a fact that an engine that has greater volumetric efficiency everywhere will get better consumption at any comparable power output than the std one .

The real challenge is to have and engine that has better throttle response/more part and full throttle torque/better consumption when driven in a normal manner .

There will be life after this RX Turbo and it'll almost certainly be parted out to get anything worthwhile back from it . No one would pay a lot of money for even a really flying example because they have low perceived value .
A well set up aftermarket computer and loom comes out of a car quickly and easily because it wasn't integrated into the body looms like in OEM one is .

Then when you get serious about AWD turbo cars and buy the Lancer Evo you stick the engine management system in that and go again .
You don't try to run sophisticated engines with basic computers because the lack of electronic smarts means dumbed down engine control .

To each ther own , cheers A .

And BTW the reduced pit stops means the teams are forced to get maximum consumption if they are to finish long races . It's not about full load only fuel usage , its everywhere .
A few years back Nissan ran one of the Europen endurance races with a (I think 390Z) turbo car and it never went richer than 13.5 to 1 AFR for the whole event . It didn't go bang or cook pistons or run red hot .

A .

[AndrewT]

A decent ECU should allow you to run multiple tunes and switch between them easily.

So you could have one tuned for economy (best as you can get) and one tuned for ball tearing power. Probably can't get a huge difference between them but it's something that can be done.

Also remember heaps of it comes down to driving technique. You can very easily get many more kms out of a tank if you just stop pushing the go-fast pedel so hard and drive off-boost as much as possible.

[discopotato03]

I agree with the bit about the foot but not necessarily about staying out of boost . OEM's often tune their engines really rich on boost and it isn't aimed at performance .
With the std header and turbo my engine would go richer than 10 to 1 AFR and it probably would have performed better at 11 or 11.5 AFR .

They mainly do it for long term reliabilty ie injectors getting a bit dribbly etc etc .

I do things differently to the manufacturers by having low restriction exhaust sides , it's also why I'd never run a std Subaru supplied turbocharger given a choice . My way is to slightly upsize the turbine and insist on a ball bearing cartridge because they give you same or better turbine response with less turbine/housing restriction . When you uncork the hot side you can often run more ignition advance because you don't get as much reversion/charge pollution/charge preheating so the detonation threshold drops away .
I call it the "windmill" effect where the turbos rotating group is spinning fast enough to provide a bit of extra airflow without inlet manifold pressure being positive or above atmospheric .
I mentioned this on the Nissan forum and a few came back saying how they could feel the difference with a properly specced BB turbo .

The Subaru spec ones are about the manufacturers ideas and engine characteristics .

If you can tune an engine properly it never uses any more or less fuel than is required to get the desired AFR's . When your stuck with std computers you don't get a lot of choice unless you can change Rom chips or flash the later ones .

A .

[discopotato03]

I'm looking into GM LS1 coils ATM because these critters have an in built ignitor so you pretty much just mount them off brackets near the spark plugs and use short HT leads to the spark plugs .

The V44 computer has four ignition outputs and can directly drive four of these coils .
Would be neat to lose those long std EA HT leads and forget about cross firing in them or distributor caps .

Why should only EJ's have direct ignition systems , some of them anyway .

A .

[discopotato03]

Christmas in Diet Coke Ville , computer and four Bosch LS1 coils should be in the bag on Monday .

Std system going into mad (lean mode) without warning is getting to be a joke . I'm not sure why it does that but suspect the TPS may have a dickey open throttle contact switch .

With the next computer I should be able to get away with a 4P ECU type TPS because they have the potentiometer as well as the idle switch . It bolts straight on too .

A .

[twilightprotege]

having tuned numerous microtech ems's, i wanted to comment about the common belief that people should always have 2 tunes, 1 for power and 1 for economy.

it's simple really - this is an utter waste of time and effort and doesnt achieve anything. all you need to do is ask yourself the following questions. 1, when do you want power, and 2, when do you want economy? the answers are simple. 1 = WOT and when on boost, 2 = everything but 1.

so just tune the maps for that. that way you're always getting the best economy when you're driving normally, but max power is always there when you floor it - anyway, i'd never ever ever want economy at WOT or high rpm - engine safety (ie not running too lean) is more important than that!

and disco, have you looked at the new haltech platinum sport 500? i'm more than likely going to be taking the microtech off my astina and using it on my suby when i go ej20, then getting the PS 500 for the astina. it's a pretty impressive ems, and certainly takes over the microtech LT series for the value for money crew.

i've been a big fan of microtechs for years and have had plenty of dealings with jon, maretta and dom @ microtech. they are great guys, always happy to chat and even help out at drag meets (well jon is anyway lol), but i really think that they are falling behind and may eventually fold. dom's the only guy there doing development, whereas haltech et al have many people developing the units so of course they are going to jump ahead of microtech in leaps and bounds. tis a shame, but that's business.

[discopotato03]

I musn't be making myself very clear , I dislike Haltech products and never had any time for Microteks . They may do it for some people but I'm not interested .

Tuning an engine , performance is all about volumetric efficiency and torque so most would do well to forget about horsepower/kilowatts or killerwasps etc .

There is a reasonably narrow window of AFR and ignition timing that makes for best torque at any load point so going significantly richer/leaner or more/less advance doesn't do anything positive for you torque wise .

Now , things may have changed at Haltech and Microteck but last time I looked they weren't using anythnig real sophisticated in the way of a processor for their systems . Good boxes nowdays use 32 bit microprocessors specifically designed for engine management ststems . They also contain modern injector drivers so using large and modern squirters is not a problem .

Now cruising load mixtures and timing has nothing to do with full load wide open throtle mixtures and timing .
To make a rapid transition is as difficult as getting in as much fuel as it takes to not get the lean out when the throttle/s open quickly - meaning rapid pressure rise in the inlet manifold to atmospheric and above it FI is there to push it . Timing is absolutely critical in these conditions because the sudden rise in dynamic compression means beating momentary detonation is challenging . So if your computer can react quickly enough and is tuned properly the AFR and best torque timing can be maintained .

To hose enough juice in quickly means reasonably big injectors and I challenge most Haltechs and Microteks to do this quickly and accurately .

I think you must have taken me too literally when I said 14.5-14.7 AFR for a lot of the overall load range . I mean a lot of the normal use load range .

Now to the two sets of maps bizzo .

Basic computers are not terribly quick or accurate and surprisingly that may not be so critical on a race engine where if its a little rich or doughy down low no one cares because they don't spent a lot of time at low revs .

The greatest challenge is getting everything running nicely on a road engine because they spent much of their life at lowish revs and constant transitions around the burbs .
A serious management system does all the nice around town stuff and turns into a ripping snarling white hot ball of terror when the hammer is down - if the engine is capable and the computer system tuned to allow it .

A budget sysytem has to have two sets of maps because it isn't sophisticated enough to cover the requirements of any use mild and wild thrown at it . It is made rich to cover the fact that it can't operate adequately sized injectors well enough to get the appropriate mixtures and make proper transitions .

To have it all the system has to be able to detect rapid movement of the throttle/s and rapid transients in the crankshafts speed - ie to the degree not to the quarter of a revolutuion . You may be surprised at just how many aftermarket boxes are not sophisticated enough to read many factory crank angle sensors - they have to have a dummed down disc fitted because they are not even up with the smarts of the factory management system .

I'm am not doing this to a price , it's a result I want and I'm not going to hamstring it with a budget system . Later it can be used to run something more serious than an EA82T - like a 4G63 turbo engine .

In time , A .

Gannon will try to e mail you pics of the bits as they arrive .

Steptoe yes still have the Skyline .

[fredsub]

I disagree with sweeping statements like
"Good boxes nowdays use 32 bit microprocessors...."
There's many reasons to choose a processor, 32bit is certainly easier to develop to by new engineers given the expensive tools that go with them hence it is becoming a leading choice in industry - developing the software is still far more the expensive exercise.
Heres an impressive 8-bit ECU http://www.vems-group.org/
clever software and a fast 8biter. I particularly remembered this one since your mention of volumetric efficiency, as this has some particular algorithm to
calculate VE for its injector timing.....

[discopotato03]

Be that as it may and I'll look into that one .

The fact is that for some years the highest forms of motorsport are demanding performance from their engine management systems that 8 bit processors can't deliver .
Yes software development is expensive but at the end of the day thats not the customers problem is it . Do you really know or care who's developing the code and if so would that make any difference .

The point I'm making is that older 8 and 16 bit systems available here in Australia don't cut it , the ones that opt for faster processors do it because they feel the need to have an engine management specific processor and ones thats fast enought to do all the processing and not slow down the hardware - least number of compromises inc component cost .

Have a sniff around and see if any current OEM engine management systems use 8 bit processing . I don't know but I wouldn't think so on anything even reasonably modern petrol engine wise .

A .

[Outback bloke]

Do you really know or care who's developing the code and if so would that make any difference .

It seems it does to some. I know of people on this board that judge a product by it's name or manufacturer not on its' ability.

The fine tuning of an engine is a precise mathematical equation. Now the brand name on an EMS is not going to make that equation any better or worse. If you have it set within the correct parameters then it will not matter who made the ECU.

As for the two tunes thing, just think about all major car manufacturers for a minute. Think about how economy has a huge role to play in the design process of the complete vehicle including the ECU.

Now think about a major GT production race team that is running the same vehicle around a race track. Economy doesn't play as big of a role now as it did when considered for general public use.

2 vehicles, same engine completely different tunes.

You can have a vehicle that you take circuit racing on the weekend with a cranky tune ready to tear those balls out that were talked about earlier. Then when the racing is done you can get in that car and drive it sensibly home.

Australias fastest 4wd at one time which just happened to be a WRX had this set up. There was a write up about it in a top ranking car magazine. It showed the rex on the strip putting in very quick times on methanol and also being a very sedate, relatively economical car to drive home in.

Then there is the "wank" factor for bragging rights. You don't need a sledge hammer to crack an egg. Same as you don't need an EMS that Ferrari's F1 team would use to run a street driven EA82 no matter how many turbos, injectors or coils you want to run on it.

[discopotato03]

Sigh , still not getting the message across .

So according to you any and all "boxes" can do the same thing based on the same settings , no oxygen there .

To OEM spec cars , OEM state of tune does not take into account someone driving the thing in a competitive manner because that's not what production cars are set up to do . The fact that someone may want to alter its mapping assumes that there is the ability to do so meaning flashing late systems or using aftermarket management .
If you take option 2 you're highly unlikely going to have the mapping exactly the same as the manufacturer did because the green police aren't there to see the tuning .

What I want to know is if you have the same car/engine/computer why does the tune have to make the engine a cranky pig ? Idle/low load/cruise load mapping doesn't have to give choppy running in order for the engine to lift its skirts at wide open throttle running .

Exotic fuels open up a completely new set of circumstances and while I have mentioned an interest in E85 the only reliable source at Rozelle is not just down the road from me so not going to be a permanent arrangement .
Blind Freddy can see that fuels with completely different heat values are not going to be able to be run with the same mapping , not without lunching something anyway .

The wank factor . Well Ray Hall and his re branding of the Link G4 Storm or Extreme to Vipec V44 and V88 is irrelevant to me . I don't get involved beyond learning of a products abilities and suitability and my aims .

You will never see a brand sticker on my car/s and its handy that the 4WD Turbo RX decals on the side have been polished virtually white by previous owners .

I have a policy when I buy expensive gadgets and that's usually to go a little better than I possibly need mainly because it saves doing it twice if the first effort falls short . Cheaper in the long run .

Now the tuners I've spoken to all reckon the computer I like does it all nicely and is a breeze to set up , they didn't have anything glowing to say about Haltechs and universally everyone relegates Microtec to the bottom feeder level .
Now you can pay a bit more for a system that goes in nicely and is straightforward to tune - or you can buy something a bit cheaper and cop larger tune bills because the alternate box takes more time to punch the figures and end up with an acceptable tune .

The feller who will tune this thing I have known from a distance for - must be about 10-11 years and also comes highly recommended by someone from the Skyline following who's used his services and likes his style . SK if you know his nickname .

Dam shame not coughing up retail and still will cost a tidy sum of money , no don't ask .

My opinions and findings are exactly that , maybe in future I shouldn't think out loud with a keyboard and keep the results to myself .
If I have a win great and if I don't tough .

Yes agreeing to differ , that is the civilised approach .

To each their own , cheers A .

A "link" from NZEFI just in case anyone wants to see the specs or is just curious .

http://www.nzefi.com/forum/viewtopic.php?t=28

[twilightprotege]

definately dont want to start an argument etc, but theres a few things to take in to account.

1) 32 bit does not (neccessarilly) = faster than 16bit. It refers to the width of the processors internal registers. Has a lot to do with addressable memory (and given that its unlikely that processores like this would have address memory extensions, or any form of memory windowing, this could be in fact the upper limit of addressable memory for the computer). Speed is determined by frequency. That is how many times the processor can cycle per second. All modern ECUs that are available are capable of operating fast enough to manage a motor in real time to inexcess of 10 000 rpm. Microtech's and haltechs can actually push this quite a bit further, microtechs to 16000rpm! Add to that that microtech's use a 32bit processor and have been for quite some time
2) most if not all engine management systems are "reactive" types...that is they are reading their information "after" the fact (determining RPM based on the last 2 revolutions of the crank, and so forth)
3) calculating timing to a fraction of a degree is fairly straight forward, even with only a home signal. Each trip passed hte home signal is 360 degrees of crank rotation. That will take X seconds. Divide X by 360 and you have Y, the time it takes per degree. All modern ecu's read crank speed in real time, and are more than capable of determining crank position to the degree (capable of, but hte software may not allow for it). Haltech computers can calculate crank position to the fraciton of a degree.
4) a computer's injector drivers have nothing to do with how a computer can "pump" fuel in. A fuel injector is simply a solenoid. a PWM solenoid driver will simply open the injector by an amount. the amount of fuel that flows is determined by the injectors jetting. if a computer can drive a 300cc injector, it can drive a 1600cc injector (injector impedance not withstanding).
5) if you're talking about the amount of time it takes for the ECU to "react" to changing conditions, its got less to do with the computer than you might think. The reaction time is usually down to the sample rate on the sensors. If you stamp on the throttle, the MAP sensor wont read changing conditions before the engine is demanding more fuel. So you get proactive in your tune, and measure throttle posisiton changes, and pump fuel in regardless of sensor readings to compensate. ALL ecu's (right up to the most expensive motec units) manage transient fuel requirements in this fashion - microtech call this setting the "pump" settings

[discopotato03]

A lots of truths (IMO) but not sold on point 4 , not all solenoid injectors and pulse width modulated drivers are created equally . The best around now are better than they've ever been , it's not that many years back that Haltechs had major issues controlling even moderate increases in injector size without running pig rich at low and idle speeds . I dread to think how an F3 or E6 would have gone with 1000cc injectors and 500cc cylinders .

Some of the manufacturers go to a lot of trouble to make sure they get crank position and speed updates a lot closer than 360 degree intervals .
I don't think they'd spend the dosh to achieve it for no reason .

I chased up a link to EFI University and at the bottom of their home page is an interesting book on high performance fuel injection . It is available at Pittstop in Perth and hopefully close to here , Sydney .

http://www.efi101.com/

A .

[Outback bloke]

It seams now that the correct message has been relayed thanks to twilightprotege. Well most of it. The tune of the car through ignition and fuel timing has a great deal to do with the performance of the engine as well as the economy. It also has a lot to do with the way the engine runs at idle, at cruise and at WOT.

You can take a stock standard engine and EMS from the factory as already has been stated and retune it. Ever wondered why people retune factory computers? It is to unleash more power which in turn most of the time creates a slightly rougher idle with the more aggressive cam/ignition timing and fuel delivery.

It is hardly ever, if at all, done to gain better economy.

I honestly don't care if you spend 10k on an ems, it is your money. What I do care about it mis-information to people that may not know, that you do not need to go to that stage to do things correctly. The write up on and mods on your RX are interesting to read most of the time. Maybe just some times you should take on some of the advice that is given from people that have already done this numerous times.

[discopotato03]

You people do it your way and I'll do it it mine .

A .

[Gannon]

I fail to see the issue here

He is choosing an ECU that is suiting his purpose of
1. It supports high impedance injectors, which have a lighter pintle (which means they can react quicker), which make a steady idle possible with large size injectors.
2. He can tune it with his wideband o2 sensor.
3. A local tuning shop has experience with the particular model
4. Its totally universal, when he blows up the EA82T, he can unplug it all and put it on a 4AGE or maybe even an EZ30TT

And so what if it is expensive, ive seen other car enthusiasts spend big money on pod filters behind radiators, 4 inch exhausts on sub 200hp cars and blow off valves and fancy boost controllers that do nothing more that keep Chinese/Taiwanese people in a job.

Where did the idea that you cant have heaps of power, and get less than 10L/100km on the same tune?
Maximum power is made at 100% open throttle, so if you have a TPS, the ECU will know that you have just stomped the floor and will control the engine accordingly.
If you are cruising, the throttle will be 10-20% open and the ECU will do its best to keep AFR's at 14.7:1 and keep ignition advanced as possible.

I think the miss conception of poor fuel consumption with tuned ECU's is that the tuning shop spends most its time tuning at WOT to get as much power as possible and spend very little time on cruising maps. I know some people dont even bother installing o2 sensors, which make closed loop fuelling impossible.

Oh and while im at it, factory tuned ECU's arent tuned for outright fuel economy, they are tuned for longevity, which means slighty retarded ignition and slighty rich fuel (to compensate for carbon deposits in the cylinder heads and the injectors wearing out/gumming up over their lifetime, respectively)
This is why a reflash tune can unleash some hidden power, just by advancing timing and leaning out WOT fuel mixtures. One shouldnt have to touch the idle part of the map.

Looks like essay writing is contagous

Here is a picture of a a typical fuel map, with load/manifold pressure against RPM against injector duty.
You can see there is quite a difference in fuel use between idle/cruise (dark blue) and high load (light blue, purple, red and pink)

[Outback bloke]

Where did the idea that you cant have heaps of power, and get less than 10L/100km on the same tune?

I don't know. I am happy for any one to point out a factory standard car, let alone a modified one that has lots of power and uses less than 10/100.

I'll keep my mouth shut on this one now, seems as though tried and tested is no match for theory these days. I will be looking forward to the outcome and will be pleasantly surprised if my information is incorrect.

[fredsub]

I concur with disco on point 4)
"injectors simply solenoids"
At the speeds these mechanical devices must be run, if you don't take into account inertia and hysteresis you won't get very accurate opening times,
especially near idle, this even more so applies to peak and hold injectors.

On the other hand, marketing has done its job on disco:)
but hey its his money.

just like to add one more thing about the 8/16/32bit marketing......

exaggerated example....
so lets say we have a calculation that must be done C.
input samples once a second. Need a calculation result 1/sec
so we have 3 computers
Fast 8bitter
32bit embedded
32/64bit gigahertz pentium MB

Time for calculation
8bitter 0.1sec
32bit embedded 0.01sec
32/64bit gigahertz 0.0001sec

so going by marketing, the 32/64 bitter is the most powerful and best computer to use, will beat the competition - well thats true, BUT its not honest.
The truth is all 3 do the job perfectly well.


I will also add that still yet there is far more automotive spec'd 8/16bit cpu's than 32biters.
Also the embedded controllers on nearly all hard disks are 8bit!!!!
And very usually a very fast derivation of a 8051, a cpu architecture which is positively ancient in terms of how long its been used.

I too am not convinced on the need for more than one map/configurations set to cover the entire power/economy range of an engine,
more likely this is simple from a habit by engine tuners.
it would be handy as say a reference set obtained on a dyno, and another set to play around with for on-road changing.

Nice discussion going here:)

[discopotato03]

I think it's time to wait and see what the result will be .
My car is is a bit unique in some of its bits which means its in uncharted waters . It has and continues to be developed with a view of what others have done in the past and a few different ideas based on different trains of thought .

The guts of it all is how much better the aftermarket system can potentially run my engine compared to the factory one and how successful my combination of bits ends up being .

What I sought to do was make this engine have a bit more part throttle torque with better breathing via cleaned up ports/slightly higher static CR/NA EFI camshafts/better than L manifolds/more modern turbocharger .
With a bit better dynamic compression from less gas flow restriction and a slightly reduced clearance volume the theory is it will make more torque and be able to use best torque timing (octane dependant as well) to make more efficient use of its air and fuel .

Driven flat out I'd expect this engine to make a bit more torque and horsepower but the proof of the pudding is in the eating meaning suck it and see . I never been near a truly powerful feeling EA82T so I have no yardstick to judge by . I don't know if all my theories will come together for me and the only way to know is to try it .

Time will tell .

Cheers A .