HI Todd,

Good question. One that has been buggin me from the start. At peak hp, the 185 wheel hp EVO is running 17psi or a pressure ratio of 2.16. At peak hp, the 160 wheel hp WRX is running 10psi or a pressure ratio of 1.68. We can get a better picture of general operating efficiency by running a few numbers.

In stock form, WRX making 48 wheel hp for every 1 bar absolute of pressure and 1 liter of displacement (160hp/1.68bar/2liters). The import EVO VII made an efficiency number of 50. The AWD Porsche 911/996 Twin Turbo makes 49. Digging up some old records, even our aftermarket turbocharged Impreza 2.5RS cars make 49. The lowest car I could find is the Audi S4 which only works out to 42-- which is not surprising considering that the ones we've tested run stoich a/f ratios for the sake of emissions.

The stock EVO VIII is at 43 wheel hp which is certainly on the lower end of the spectrum of all the AWD turbo cars we've tested here in CA. This is surprising considering it having only 1 cat, a reasonably well-designed exhaust system, a fantastic exhaust manifold design, good induction plumbing, etc,. Yet, it's still well below the WRX which has two more cats and a rather bad exhaust manifold system (pre turbo), relatively small turbo, etc,.

-Historical evidence of tuned WRXs

With the WRX, through simply remapping fuel, timing and boost curves, we're able to bump wheel hp to ~190. Boost, at peak hp, rises to 14psi. This translates to an efficiency number of 48. In other words, we're able to improve output significantly through ECU tuning alone without harming operating efficiency (which is good considering we are raising boost without opening up the exhaust side with cat-backs, downpipes, up-pipes, etc,.)

With an up-pipe and 3" turbo-back exhaust (so that only one cat remains), we're usually able to bring WRX output up to at least 210 wheel hp (and even more with an UD pulley and silicon IC hose). Now running 14psi, such modifed WRX are running efficiency numbers of 54 wheel hp per atmos per liter. As expected, we end up trading some exhaust noise level and margin of safety (ie. one can no longer run 87 octane as some stock WRX owners have gotten away with!) for a good increase in output and operating efficiency. No surprises there.


-And now the EVO VIII

To achieve a WRXs (one with all its cats, stock exhaust, etc,.) efficiency number of 49, a stock EVO VIII would have to make 211 wheel hp on our dyno. Interestly enough, the import EVO VII we tested several months back did just this. Yes, on 91 octane. Also, all evidence suggest that the pre-production EVO press car, which dyno'd 3 or 4 months ago at 240 wheel hp on HKS's AWD Dynojet (see current issue of SCC), would have accomplished something very close to this as well.

So why are the stock EVO VIIIs we've tested only making ~185 wheel hp tested under very similar conditions (91 octane, temps, humidity, etc,.)? Let's look in the ECU tuning. I've spent the last two weeks and nearly 600 dyno runs remapping our EVO's fuel and timing curves. Did I find the missing 15% power output (on 91 octane, that is)? Nope. Did I even find 10% more power? Yes, but not without random detonation and general nconsistency/unhappiness. More on that later.

To achieve the modded WRX's efficiency rating of 54, an EVO VIII would have to put down 233 wheel hp. Can it do this on 93 octane with some simple exhaust upgrades? Yes, it can. Can it do it on 91 octane. So far, no. But we're working on it. The hard part of the job right now is to see what can be blamed on octane (or lack thereof), what part can be blamed on conservative ECU mapping, and what part can be blamed on yet-to-be-determined hardware changes (compared to the EVO VII) that we don't know about yet. The latter is a distinct possibility given the data we accumulated. And we're investigating that possibility through a few avenues we're fortunate enough to have access to.

But here's some interesting info. Running just 13-12psi of boost, with nothing more than a cat-back and revised fuel and timing curves, our test EVO just barely exceeded the stock EVO's peak hp output (running 19-17psi of boost). Below 5000rpm, the stock EVO had the advantage (6-7psi of extra boost certainly helps). Above 5000rpm, the modded, lower-boosted EVO was well ahead. The difference at 7000rpm was, IIRC, 20-25 wheel hp. Not bad for much less boost, huh?

Just some info to chew on. I'm sure it will be rejected and projectile vomited by a few

Cheers
Shiv S. Pathak
Engine Calibrator and Owner
Vishnu Performance Systems

 

It smells like cam changes are in order here!

 

Shiv--

Wow. Great, informative article. Sounds like the low octane, high boost situation is wreaking havoc with power output. Interestingly, the low-boost Evo setup has a much higher efficiency rating than the high-boost one-- again, points to boost being wasted on poor gas. Getting pushed over the peak efficiency and into detonation land...

At least we know the potential is there for the 4g63-- just a matter of extracting it. Thanks for all your informative postings, musings, ramblings, etc. By sharing your thoughts and info as you get it you demonstrate that you are, first and foremost, a tuner with a passion and commitment to the cars. When I get my Evo, I'll be taking it up north for you to work on.

Regards,

Heff

p.s. keep it (the info) coming

 

Shiv,
First, thank you for your quick and in-depth response to my questions. If what you are saying it true and the higher boost is more or less "being wasted" perhaps the internals aren't effectively being used, that and the combination of lower octane. I'm not a huge numbers guy as long as the car is everything I am expecting and able to make more power. I would like to justify my purchase over the STi since I can get either at MSRP.

Perhaps, if its not just octance, maybe the differences are held within a/f ratios, etc. I'm not super experienced in internals and a/f ratios, but this whole scenario isn't making sense since the boost is there and supposedly the engine is the same as the rest of the world's. I'd like to extract all the power I can from this car w/o going overboard. I want this reliable and that's why I am asking you, b/c I want a complete solution and not a make-shift AFC, BC, etc in my car.

 

WTF is an engine calibrator? Vishnu, you try to hard to make yourself sound like a scientist. Do you know what calibration is? I can explain if you like. I am engineer, and I would love to hear this explanation from your point of view.


Please, keep it real.

PS,
I hope to meet you at a import event or something, I need to see up close what you are all about.

 

Vishnu, what's being used to tweak the fuel and timing curves in your EVO 8 test?

Thomas Dorris

 

Actually, the full term is "engine management calibrator" but that's a lot of typing. I could call it powetrain calibrator. Engine control system calibrator. ECU monkey. Exhaust inhalation expert. Carbon Monoxide taster. Etc., All describe what I do, all day long, on the dyno.

twdorris-- We have our own proprietary user-programamble engine computer being used to modify fuel, timing and knock control maps. It runs in parallel with the factory computer. It uses a windows-based tuning interface, allowing users to make real-time changes on-the-fly. It stores a number different maps internally allowing the user to upload maps for 91 octane, 93 octane, etc., User-definable rpm and load breakpoints with a maximum 20x20 map resolution making it have better granularity than the factory ECU. Different maps can use different breakpoints and load variables (either TPS, MAP or MAF) depending on the intentions of the user. User-definable adjustment ranges, datalogging etc,. We originally designed it for WRX applications but found it to work alarmingly well with the EVO which doesn't exhibit the same number of emission and knock control related hurdles (such as TPS-dependant open/closed loop fuel control behavior, active spark advance map, floating ignition advance multipliers, etc,.) It took only 2 days to adapt it for EVO use and only a week to develop the software and to generate some basic calibration maps. Without it, we would have to resort to AFCs, ITCs and other old-school low-resolution approaches which would leave quite a bit of output potential on the table.

Cheers,
Shiv S. Pathak
Chief Dyno Dude
Vishnu Performance Systems

 

I'm trying to understand why the engine has a weak top end compared to modified WRX and Evo VII and why additional boost at the upper rpm actually hurts output.

I assume Shiv was able to optimize timing and fuel with his test car. So its not a timing or fuel issue that is limiting power.
But even with those corrections the engine does not like boost at higher rpm and (in general) does not put out the HP one would expect in the upper band.

So what could be caursng this, assuming the fuel and timing are not the culprits? It would seem we are left with only a few possible things that could be the cause: (1) Poor cylinder heads and/or combustion chamber design. This is unlikely as we should be able to expect that our cylinder heads are better or at least equal to Evo VII. Heads/combustion chambers generally get better in time and not worse. (2) Restriction in exhaust. Shiv is already running a cat.back exhaust and still has these top end restrictions. In another post someone stated the cat. is very non-restrictive, so thats probably not it. The downpipe is probably thesame as Evo VII (?) and so that is probably not a big problem. (Also, it does not look very restrictive). This leaves us with the exhaust mainfold which appears to have a great design from the outside. (3) Restriction in intake. Evo VIII piping appears generally same as Evo VII. Perhaps intake manifold itself is a bit different but it does not look like it would be more restrictive.
(4) Cams. This would seem the best possibility for our "problem". Perhaps our cams are more "mid range" cams and less "top end" cams. Perhaps this was done to help pass USA emissions. Perhaps the engineers decided to give us a really fat midrange and sacrifice top end in order to make the car really fun in normal driving (which it really is!) while at the same time meeting/exceeding US emissions. If this is the problem it will be an easy cure. But I'm addicted to the mid-range power; hope we don't lose much of it - I doubt we will!
Summary: This is not too complicated actually. Once timing and fuel are correct its a matter of getting more air/fuel through the engine. The complicated part is the stuff guys like Shiv are doing with the ECU, etc. The more simple part is finding the restriction to additional power. It would appear it is either intake or exhaust restriction or the cams. Someone will no doubt soon take the complete intake and exhaust systems off the car and look for any restrictions and if nothing of consequence is found it would appear that only leaves the cams as the "problem", assuming the cylinder head/combustion chamber design is not screwed up.

Sorry to be so long winded. And I am new to this age of tuning so mabey I am missing something here. My question would be: What else can it be if the intake and exhaust systems do not have restrictions except the cams? Any other ideas would be great to hear.

 

It sounds a bit like an old-school EFI PMS.

 

XEDE for the EVO perhaps?...

Shiv...alittle far fetched, but any chance of stock ECU reprogramming in development ala ECUTEk?...


Thanks,

Ryan

 

Ah. Thanks. Would it be fair to compare this to a PMS, then? Is the system directly driving coils and injectors or is it attempting to use factory ECU hardware to do that?

Thomas Dorris

 

EVOwhat-- Eventually. But it will take a while. The ECU is a bit different from earlier generation EVOs.

twdorris-- I don't know much about the PMS but if I'm not mistaken, they used to have an FD3S application sold through PFS. If that's the case, I'd say that they are quite different. The XEDE introduces no additional drivers and relies completely on signal intercepting/replicating/modifying. This is the only way to retain factory knock learning systems which, historically, have been nearly impossible to improve upon from scratch without introducing some unwanted artifacts.

Cheers,
Shiv

 

The PMS intercepted the airflow signal. But it would only modify it in that it would clip it at a certain point to keep the ECU from fuel-cutting. Other than that, I think it left it alone.

The PMS also intercepted the fuel injector lines and modified the signal based on user input. Thus it did have fuel injector drivers.

Same goes for timing. So it retained factory knock control. It also offered a datalogger.

http://www.efisystems.com/

I can't see how your box wouldn't have fuel injector drivers. You can't just intercept and modify a signal and put it back on the same wire. You must cut the injector wires coming out of the ECU, read the pulse-width, and send out a modified ( lengthened/shortened) pulse-width from drivers on your circuit board to the injectors.

How well are you keeping track of the time that the pulse leaves your box? I hope it doesn't get delayed too much. That would sortof undermine the whole reason for having sequential fuel injection. Or at the very least just delay it til the next time that particular cylinder was supposed to have an injector event.

 

It doesn't. I promise.

shiv

 

Shiv....maybe you should just call yourself an "Head Engine-eer" huhu.

SC~ who thinks it's funny how some engineers and sciencey types get all caught up in the definition of things.....