View Single Post
  #3  
Old 08-18-2007, 01:08 AM
XT6Wagon XT6Wagon is offline
Registered User
 
Join Date: May 2007
Location: Vancouver, WA
Posts: 417
Standard answer I give, doing a turbo setup right costs the same as doing it cheap once you get down the road a bit.

Doing a low boost kit is in my opinion risky as very few turbos work at all at low boost, and even worse... none of them in a size that the SVX will like. If you have to do it, find a BIG turbo with as good of a low pressure ratio map as possible. You will note that turbos are very inefficent at high flow and low pressure ratios, this is what happens when you try to run a small turbo at low boost. A big turbo at low boost might be "soggy" on up to out and out laggy, but it will atleast make a performance difference on boost and not just burn your motor up like a small turbo. Even worse a small turbo will have a small turbine or "hot side". Small hotsides equal large backpressure which equals lower performance with higher stress. Going with a larger hotside can actualy decrease lag despite the common ricer wisdom. This is due to the fact that the compressor side is powered by the turbine, and too small of a turbine will not provide sufficent power to the compressor. Imagine if you ran a second engine to power the supercharger on a top fuel dragster... A 3cyl Geo Metro engine just won't get the job done, yet thats what many of the very popular aftermarket turbos DO, is put a tiny turbine to power a huge compressor.

Honestly If I was to do mine today, I would start with a built Forced induction motor from ECUTune, a built transmission (or 6spd swap), and then a properly built Turbocharger setup from someone who knows what they are doing. Proper external wastegating, proper bypass setup, proper piping, proper intercooler design, etc. Its comical the number of times I've seen a Stock turbo STi drop 10-15whp with a FMIC that wasn't exactly thought out. Thats compared to the stock topmount which people wrongly think of as too small and incorrectly placed. Sadly the SVX doesn't even have room for a proper elbow into the throttlebody, much less a TMIC. In this application I'd honestly say that doing a Air/Water IC replacing the plenum is the only way to fly on the SVX if your budget can take it. Super low pressure drop and minimal lag. Running the pipes to the bumper for a small FMIC just adds alot of volume to the system and alot of pressure drop. Again I'd assume ECUTune could make you a very nice Air/water intercooler intake manifold combo with his experience at doing the fabrication for the supercharger kits.

Last, don't be shy with exhaust pipe diamiter. I've seen cars that I'd be running 3.5" or 4" if I could actualy get the parts to make an exhaust in that size properly. One car shoved a 20ft long aluminum chimney "pipe" a good 8-10ft off the tailpipe once it hit boost as the internal 4" diamiter just ah... wasn't enough. Oh and DON'T taper the exhaust like some companies do claiming that it makes more power. LoL just because the exhaust is cooler and more dense, doesn't matter, you still have the same mass flow. The reduction in energy does make it more dense, but it moves slower too. So tapering down to 2.5" say from a 3" is still bad as the slower cooler exhaust gas has even less room to wander out, and thus slows down more, which of course provdies more restriction.



Last thought, it would be a ***** to tune perfect without running the engine as a pair of 3cyl engines, but you could use an asymetrical setup on the turbocharger. This means only running one bank of cylinders into the turbocharger's turbine. This is normaly done to keep a large engine from overspeeding a smaller turbo and er.. breaking turbo and motor. So if you say got a VF-39 from a STi, this is how I would run it. More lag, but likely more peak power as you have far lower exhaust restriction. This lack of restriction on the one bank is what causes the tuning issues. I'll finish with this style of setup is honestly for torque and MPG boosting, not pure power. The difficuties doing this right prevent me from recomending this as a good idea unless you want to spend the time and money for real heavy R&D to make it right.
Reply With Quote