Thread: New type of BOV
View Single Post
BoxxerSix BoxxerSix is offline
Registered User
BoxxerSix's Avatar
Join Date: Jun 2006
Location: Ballston Spa, NY
Posts: 528
Alright this is good, we're back on track with a good Pelican thread here I'll first go an apologize for jumping down Gorans' throat. I was tired this morning and one thing that bugs me bad is discrediting someones work w/o first asking questions about it to better understand the general idea and knowledge behind it. No hard feelings I hope?

Ok I'm going to try to explain the issue here with these Evo's and my reasoning for doing what I've done the best way I can here in a post. I have alot of information jumbled in my head right now and this may not be that easy for me to translate it all into text without it getting confusing or off track so please please please bear with me here. This may be quite long and may jump into multiple posts pending if there's a character limit. This is most definitely OT and not Porsche related but here it goes:

The Mitsubishi Evolution is the car in question, both the older Evo 8 models and the newer Evo 9 GSR/MR models with the MIVEC 4g63 engine. The car is equipped with a 2.0L 4 cylinder forged internal engine. Boost is provided by a big 16g mitsu compressor mated to a 9.8(8) or 10.9(9 GSR or MR) twin scroll turbine and manifold setup. These cars are well know for making crazy power at rediculous boost pressures(1.8-2.6bar) and doing so with the ability to open source tune any part/map in the OEM PCM. As such, you end up with a "import tuner" type of car with kids throwing a mass of aftermarket parts at the chassis to tap the cars potential.

The issue that affects 95% of all the Evo 8 + 9 cars revolves around the diverter valve and track use or any sort of low load/throttle modulation. The DV setup on these cars is not unlike any other car, however the car suffers from what appears to be an EXTREME level of DV sensitivity to throttle input against engine RPM and load. If there is a slight load on the engine while driving and you attempt to modulate the throttle in any attempt to control the car(say like throttle steering through a corner) the DV dumps under throttle, causing a massive unsettling of the car. In some instances, this can also be noted on long gradual uphill jolts, when a very light but continuous hold of throttle will eventually lead to a DV dump at mid to upper rpms.
This one Evo here in particular has had numerous close calls at Watkins Glen while running through the esses. Car bucks and dumps mid corner causing unsettling no different that if you lifted off the gas completely or hit the brakes mid corner. This is very scary stuff as it's totally unpredictable. Most of the Evo's I've tuned and dealt with run anywhere from ~315-430AWHP, so an unsettling in a car with that kind of power usually comes at elevated speeds as well. Stuff that I know 99% of the members here are well aware of or have experienced themselves out on the track.

At first the cause was thought to have been issues with the DV itself. Then the 272 degree camshafts were thought to be the root cause, as well as the various intakes, charge piping, FMICs, etc etc. This however was soon squashed when Evo owners with nothing more than a little boost and a reflash caught wind of my research and started advising me that their cars suffered the same issue. More and more people started popping up, every car different and all had the problem. After a few months of probable causes, prototype isolators and parts, and countless tests on damn near every modified Evolution in the area nothing stood out as a root cause. All the car were different in their own right, but all had this bucking/dumping issue. Some thought that intake disruptions were causing a backfeed to the MAF, upsetting the PCM fuel/timing signals but again squashed when a few MAP based Evo's showed this signs. The last ditch effort and possible cause came about from talks with some aviation engineers who dealt with high altitude aircraft turbo charging. The one part that damn near all the afflicted car had in common was the stock turbocharger, with say but a few that had custom kits with spools like 18G's or 28R series Garretts.

The stock turbocharger spools incredibly fast. We regularly run boost pressure in excess of 25+psi all day long on stock blocks and don't break a sweat. The twin scroll turbine and low mass 16G compressor lead to some seriously fast spool rates. This one car here will see 25psi around 2900rpm from 2nd gear on out if it means anything. Now over the course of the trails with this problem we had discussed the compressor map in general and all the math you could compute kept putting us well in the efficiency range of the compressor but we never took into account the WAY the compressor got there nor how it was affected during low/partial throttle modulation.

Knowing a bit about turbochargers myself what I ended up doing was making a special intake tube to house an optical RPM reader from my VSR balancing machine so I could observe and monitor the CHRA speeds of a 16g while operating the engine at load. I also plumbed in pre/post IC boost gauges and pre IC charge temp gauge to log and diagnose exactly what the turbocharger was doing at these boost levels. Over the course of a couple weeks, I logged the information and discused it again with various people in this field and we determined(guessed more like it) that throttle modulation was causing the compressor to surge at mid level pressure ratios, right around where these turbos run in their sweet spot. Because the turbocharger spooled up so fast in reference to engine RPM and flow requirements the compressor would function fine at mid-to-WO throttle levels, but any sort of low level throttle modulation with a light engine load would hypothetically cause the compressor to shift into surge as the CHRA would still spool up rather quick at those throttle input levels, even though the airflow requirement of the engine was low at that point. It is thought that this compressor surge was causing flow disruptions to the air going into the motor itself, which subsequently tripped the DV into dumping while ON the throttle. Sound weird? Tell me about it. This is the best way I can explain it right now. Alot of it hypothetical yes, but it's the best I've got so far.

Adam Hennessy
Old 01-28-2008, 08:52 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #43 (permalink)