I just spent the last hour going through the GTA thread and came up with a compilation of good info WRT 4G63/DSM stuff. Keeping in mind some of the info will be slightly out of context but I think its still valuable to have all that info condensed in one post. Thanks to the contributors!
Christopher Lambert:
Basically you have two flavours of traditional 4g63 engines, a 6bolt and a 7bolt. The bolt number refers to the amount of bolts located on the crankshaft where the flywheel connects. 7bolts had a notorious crankwalk issue that happens from time to time, but technically the block itself is more stout. In factory form these engines handle 400awhp with ease, with the highest I've seen being 610awhp from a complete stock long block minus head studs. The transmission should be more than up to the task at those power levels in such a lightweight 2wd application. You would want a clutch upgrade, but you would be dealing with a 300 dollar new ACT 2100/2600 style clutch. The 1995-1999 cylinder head is the ideal head for direct flow to the intake valves, and provides the best across the board power. In factory form the engines can rev to 8k fairly consistantly with a factory redline of 7k. The fuel injectors are 440cc from the factory, so they would need to be upgraded, but at 400whp you are looking at very cheap injectors, likely on the order of 150 bucks for set. DSMlink is a great option to keep a factory-like idle and startup conditions while still offering speed density tuning with datalogging. DSMlink is simply a chipped factory ECU that allows programming by the end user and runs all of 500 bucks.
Stock head
Stock block
ARP standard headstuds
Stock exhaust manifold
Stock intake manifold
650cc injectors
DSMlink
ACT 2600 clutch
Evo 3 16g turbo
That would be what I would consider all that is needed to run 400whp. This is assuming that you would already be running a custom intake, intercooler, and exhaust setup since this is a kit car. Most of those parts I listed are stock, and you can likely get away for a while without most of those and keep it even cheaper.
Im glad to hear he's planning to run an AWD trans as they are stronger. There are companies out there that offer the fwd trans adaptation, but basically it requires welding the center diff, and blocking off the output shaft to the transfer case. Transmissions for these cars are fairly cheap and should have zero problem holding this kind of power. For those of you who don't mind running an auto trans, the DSM auto trans holds legendary amounts of power and is capable of damn near 9 second time slips on a DSM in almost factory form. In a car like this I would imagine it would be unbreakable and the car would scream out of the hole with the smaller turbos it would be forced to run.
For what its worth the Evo 1-3 all ran a DSM style 6bolt engine. They came with a evo 3 16g turbo, 550cc injectors, and a few other small odds and ends such as a dual runner intake manifold (not the best manifold IMO but it sells for a good 100 bucks on the market all day long). 6bolt engines had 7.8:1 CR, 7bolts had 8.5:1. The iron block and crank take all the abuse you'd ever throw at them. 6bolt rods are slightly beefier than 7bolt rods, but to put 8.5:1 stock pistons on you'd need to rebush the rods due to wrist pins (22mm vs 21mm). Some cheapskates go that route but you'd be better off going with a set of eagle rods and wiseco pistons which can be had for all of 650 bucks as a set if you want a bulletproof bottom end on this type of car. With proper L19 headstuds and a Mitsu MLS headgasket you can boost the hell out of these engines. I personally was running 43psi by the end on with Crower rods and Ross pistons (Ideally would choose Wiseco HD pistons which cost only a fraction more but offer a MUCH improved design with ultra beefy wrist pins and very deep/large valve reliefs for folks who want to run .450-.500lifts on high duration cams.
As an edit I wanted to correct myself regarding the Evo 1-3 motors:
EVO Motors: The valve covers of the EVO 1-3 motors will be a kind of titanium color. They are the same as a US 2G motor with nothing fancy. They have the smaller 2G style intake ports, a small runner intake manifold with a larger plenum and a larger throttle body than a 2G motor. Still they are hard to find. Many are from other 4G63 powered cars and can have small 13G sized turbos. They will not bolt into a 1G because there are no provisions for the front roll stopper motor mounts that bolt to the block. Everything in Japan from 94 and up has the front and rear roll stopper motor mounts on the transmission 2G Eclipse style. Many people are running around yapping about how they have an EVO motor in their car. If it bolts into a 1G, it isn't an EVO motor.
Just to give some food for thought and vendors where you can find components for these cars:
ECMTuning, Inc., plug n play ecu conversion to allow for datalogging/tuning with great customer service from Dave/Tom. Owning the system also gets you access to their forums which is the best forum for the DSM community IMO hands down.
Kiggly Racing - kigglyracing.com, amazing guy who currently has the fastest fwd car in the world. Has a lot of useful parts
Welcome To Extreme PSI, cheap aftermarket components
JM Fabrications, cheap fabricated components
JNZ Tuning, for OEM mitsu parts and
www.magnusmotorsports.com, more pricey but Marcus has some pretty off the wall stuff to offer
Forced Performance Turbochargers, Parts, Kits | Ball-Bearing Turbo Chargers For Sale Online - FP Red, Black, Green, great turbos, the HTA 68 would be a torque monster in this car, and the HTA 82-86 variants would put the car in the low 9's while still spooling fully in the low 4's
www.sheptrans.com, builder of the most stout DSM transmissions available.
Personally I feel a 30R variant would be all that anyone would need in a car like this, probably placing you in the low 10's, possibly high 9's depending on how the car is setup, and still spooling at least 20psi by 3500ish. The head should be a 2g head, the block can be a simple 6bolt setup. For cams, FP makes some of the best and I would highly recommend the FP4R's, or possibly a FP4R/5R combo depending on how high you rev it (9-10k isn't uncommon for folks with kiggly's valvesprings). The intake manifold should be changed to a sheetmetal or magnus' cast version. If budget is tight a stock Evo 3 manifold works in a pinch for a slight upgrade. DSMlink allows for speed density, which will make designing intakes a breeze, and give your turbo the unrestricted flow it needs to be efficient.
Some other food for thought. Depending on how well this car pulls G's, there can be starvation issues in these engines at high sustained G's. Kiggly's oil restrictor plate restricts the amount of oil feeding the lifters and returns the unused oil back to the pan. During high G track events this helps keep oil in the pan where it is needed most, and might be of particular interest to folks here.
For budget folks, stick with an 1g intake manifold, MHI Evo 3 16g turbo, 6bolt longblock, and FP race exhaust manifold, and you have a combo that has the potential for 400whp with the right setup such as this and be about as dirt cheap as they come.
I personally ran a 6bolt on my car, but the 7bolt is in many ways a better block it just is flawed by the possibility of crankwalk, and has a weaker OEM reciprocating assembly. To highlight some of the good points, the block itself already has a built in cast girdle, the crank is around 5lbs lighter, and the block has increased rigidity. Some other highlights are the fact that it has a dedicated crank sensor, which keeps electrical timing in-sync with the longblock in the event that the timing belt stretches. Kiggly makes a girdle for the 6bolt as well as a crank trigger, and you don't have the concern of crankwalk. It's your call which one you choose.
400-450whp is virtually nothing on this setup, especially considering it is 2wd so the losses are significantly reduced.
Exhaust:
An evo 3 16g with the right setup costs 500 bucks (possibly more new since they are harder to find nowadays), and would put you at that level. FP sells an HTA 68 and has had them on sale for as little as 799 bucks new, and it should spool very similar to the 16g, but push you easily over the 500whp mark. I would defiinitely ditch the factory exhaust manifolds (1g's are known to crack, and 2g's still have a lot left on the table) in favor of an FP cast exhaust manifold.
Intake:
Ditch the factory MAS and run speed density using DSMlink or another standalone of your choice. The intake pipe will be custom on this car ideally, but a 3" pipe should be all that's needed for this kind of flow. I'm not sure how this car allows for an intercooler so I'm going to skip over this part in particular. I would run a solid aftermarket BOV, I always prefer Tial products as they are as bulletproof as they come. I would pick up a JMfabrications or Magnus sheetmetal intake manifold and ditch the stocker.
Head:
For a cylinder head I would run a 2g stock OEM head with a nice set of cams. There are a ton of cams out there, I found the FP4R's to be a great set that seemed to work all over the power range. The HTA 68/Evo 3 won't be high revving turbos, the E16g losing steam around 6-7, and the HTA probably going a fair bit further. I would say just about any springs would work here, but if it were me I would stick with Kiggly's behives as while they are very pricey they are the best you can possibly get and Kiggly is an amazing guy to deal with and I whole heartedly support him. I would pick up a kiggly oil restrictor while I was at it to help keep oil starvation to a minimum. Many folks run 3g lifters for lower revving cars, and 1g lifters for higher revving engines. Your call there, both can be had for about the same price. OEM rockers work just fine. For head studs OEM work plenty fine at that level, and 2g engines use torque to yield bolts if I remember correctly so you'd need to replace them if you pull the head from a 7bolt. I personally would go with ARP L19's combined with a Mitsubishi MLS head gasket and be done with it. That combo has proven to hold the head down under extreme pressures (43psi on my setup). Stay away from solid lifters, the only time they should ever be used are when you are on the bench and dialing in cam timing.
Block:
In all honesty the factory block will handle these numbers just fine as long as you tune the engine properly. If you truly must pull it apart the Eagle rod/wiseco piston combo is as cheap as it gets and would be bulletproof. For even more protection for future mods Wiseco HD's are about the most stout piston I've seen in the common aftermarket capable of 400hp per piston. OEM crank is plenty strong enough. For bearings run ACL race bearings, 100 bucks a set and they are all you'd ever need for this kind of level. Be absolutely certain to use a OEM mitsu front cover, the frontline and aftermarket units are notorious for the oil pump eating itself up. A kevlar timing belt is a bonus if you want to have a little more security.
Fuel:
A walbro 255lph HP is all you need for these levels. A bosch 044 is louder and will flow more. Walbro also produces many other higher flowing in-tank designs now that I would tend to prefer over the walbro. If you need more flow down the road just add in a 255lph HP or bosch 044 in series and use the in-tank pump as a lifter pump. Ditch the OEM fuel filter, the banjo fitting causes too much pressure drop and it won't be a bolt-in affair on this car anyway. Convert to an aftermarket filter of your choice. Injectors are whatever you want, I would tend to stick with something like 1000-1200cc's, as they will idle just fine and only cost a tad more. While you don't need this flow, it certainly opens up future capability for someone not on an absolute budget. I picked up RC 1200's new from ebay for around 300 a set. One of the major downsides to the OEM system is the factory saddle tank, which is eliminated in this car, awesome OEM pressure regulator will need to be swapped out as it will be overrun by the pump, swap to something of your choice such as an aeromotive unit.
I'm sure I missed a few smaller things, but all of that is SIGNIFICANTLY less than your budget. If you wanted to keep it dirt cheap stick with a 6bolt long block and replace the OEM studs with ARP's (although there is debate that the 1g studs were as strong as standard ARP's). No block resurfacing required as you could use a composite OEM gasket.
Just to add, both engines are fully capable of your goals in factory form. A friend of mine put down 610awhp on a stock 6bolt longblock with ARP studs and cams, which would be equivalent to 650+ on a 2wd setup. It is entirely up to your budget, but realize these are from the days where Iron blocks still existed and they came with 7.8 to 8.5:1 CR from the factory so they aren't as detonation happy as turbocharged NA engines.
As any good designer knows at elevated power levels the OEM components lose their factory of safety, and can become susceptible to failure over the long haul. That may be years, or that may be months as tuning is a tremendous factor. I've always come from the mantra of overbuild for a given setup and give yourself a bit of a FOS. That may not win on race day, but I always built the car to be a street car foremost and I never once had a catastrophic failure.
Leon R
It is true that it is easy to modify an AWD transmission to function as a FWD transmission, but you need to settle on which axles you will use.
90-99 AWD transmissions use the same size spline front axles and use an intermediate shaft on the driver’s side to give you “equal length axles”.
90-92 FWD transmissions use the same size splines as the 90-99 AWD transmission, but there is no intermediate shaft, so driver’s side axle is almost two times longer than the passenger’s side axle.
93-99 FWD transmissions also have longer driver’s axles and use larger (stronger) splines inside the front differential, so other axles do not fit this transmission.
Since you are targeting ether GST or GSX as your donor car, you might have to offer two different axle options. Or… just stick with one option and make people buy an appropriate transmission. They are pretty cheap, so it isn’t a big deal ether way.
If I had to pick one transmission for this project, I would select a FWD because they are cheaper, lighter (~25-30lbs), less desirable (by DSMers) and have stronger axles.
What engine/transmission mouning scheme are you looking to accommodate?
90-94 (1G, 3 motor mounts and 1 tranny mount) or
95-99 (2G, 1 motor mount and 3 tranny mounts).
7-bolt motors (93-99) are not desirable: almost most end-up ether crankwalking or throwing a rod. Rebuilding them (before they crank walk or throw a rod) is more challenging and expensive, because you have to replace the hardware (torque to yield type and should not be re-used) and match the bearings. It can be done, but ONLY by a good machine shop which has experience with rebuilding 7-BOLT MOTORS!!!
- 6-bolt motors (90-92): CHEAP! $200 for a good core shot block or $500 for good used short block. Much easier to rebuild (almost any good machine shop can do it) and it almost never crankwalks.
So when you are looking at $20k+ project, you would be much better off spending extra $200 for a core or $500 for a used short block. Then, you can focus on finding a good, clean shell, without worrying about the engine.
Crankwalk numerous theories:
- Shift/misalignment of the main caps/girdle
- Oil squirters that are stuck “open”
- Heavy clutch pressure plates
- Lack of crankshaft nitrate treatment
- Thrust bearing design (thus the 98/99 redesign)
- Low oil pressure
- Loose(er) tolerances in 7 bolt cranks.
- Clutch safety switch (requires load on thrust BEFORE there is oil pressure).
The only thing that has was proven is none of the above are THE solution! For whatever reason, some 7 bolt motors do not crank walk, and no one knows the reason why… Anyone telling you otherwise, is either a fool, a rip-off artist, or a genius! But we haven’t found such genius yet
If you happen to have a good 7 bolt motor, use it until you run into a problem! If you really want to refresh it, put in new rods/pistons/rings and get rid of the balance shafts. But I wouldn’t mess with the main bearings (unless it has over 150k miles or so).
The only thing on which people agree is that the most reliable solution for crankwalk has been 6 bolt short block swap! Considering that a good used 6 bolt block/crank combination can be found for $150-$300, this really is a no-brainer!
2.4 make a lot more torque at lower RPMs, but... No more power than the 2.0s. So you end up with with more drivetrain braking torque, but you are no faster than the smaller engine. That torque is better for daily driving heavy DSMs, but I do not see it being very useful on a light Apex.
4G63 with stock turbo, cast exhaust manifold and a sheet metal intake is 250lbs! (Not counting alternator or PS)
Flywheel/Clutch 12-20lbs
Tranny: AWD 130lbs / FWD 110lbs
This project would ether use a FWD tranny or AWD tranny modified to remove AWD parts (thus weight closer to FWD).
So we are talking about total weight of around 380lbs
7 bolt block 95lbs + gurgle 11lbs = 106lbs
Crank, rods and pistons 42lbs
Head (complete with valvetrain, cams and sprockets) 52lbs.
So here you have a long block coming in at 200lbs even! These are stock parts that can easily take 500-600 whp!!! That is one of the reasons why it is such a special engine!
The only time you see balance shafts in a rebuilt 4G63, is when it is required by a sunctioning body (like SCCA in some classes).
Stock FW is 20lbs. But you wouldn't use it on a light RWD car (plus, they are known to come apart like a handgranade and should be avoided on high revving applications). Aftermarket FWs are 7-11lbs. There are numerous dual disc clutches ~15lbs total weight. In ether case, I did not count it in my 250lbs, but I already listed it separately.
Alternator and other accessories weigh the same between different motor brands, so I do not see the need to get into those.
I saved some weight by using sheetmetal intake. More weight can be reduced by using a short tubular header. They are known to crack on daily drivers, but should be OK on a race car.
On the other hand, you could get carried away with a huge T4 turbo and a long runner header and add 40 extra pounds in the turbo weight! But that isn't necessary until 800+ hp levels..
So the summary of my measurements (lbs):
Block (7 bolt)……………………….106
Crank/rods/pistons…………………41.6
Head…………………………………51.2
Valve Cover…………………………5.3
FP Red……………………………..14.8
Tubular Header……………………..9.1 (Cast header 11.5lbs)
Sheet metal intake w/TB…………. 5.2 (cast Intake manifold 19.2)
Pulleys, Pumps, front case……….17.7
Oil pan………………………………..4.9
Total:…………………………………255.8lbs
I would like to know if I were to pick up a good condition 2g DSM for cheap, does it matter which chassis (awd vs fwd)
Speaking strictly from the 2G Eclipse body perspective, there is no difference between FWD and AWD (beyond some driveshaft mounting bosses/studs). Engine harness will be radically different on non-turbos (since they use ChryCo engine), so expect to source it elsewhere (typical cost for used engine harness is around $100).
I hear some comments about picking such an old car, well, there is advantage to it! I do not know how much people are aware of this, but OBD2 emission inspections started flagging non-native and/or modified ECUs, it is becoming a serious problem for newer cars! One of the advantages of using 2G Eclipse is that 95 model year has an excellent OBD2 ECU, but it is pre-1996, so it is exempt from OBD2 testing! That should make annual inspections much easier in many states!
As for people who just can’t stand a 4 banger and “must have a V8”, some will never change their mind, but some will become 4G63 converts once they see what this tiny engine is capable of!
Price wise, it will not compete with the $10k 818 (natural comparison) and once you are at $20k, going to $30k is not huge leap, especially if you justify it with increased quality.
end of quotes
Cheers!