Borg Warner S200 (Bullseye S256) Review

First, I would like to say that I have no association with any of these companies (Borg Warner, Bullseye Power, FP, Waffle House, etc.), and that I am doing this solely to provide information to the community. To do that, I will have a post describing the background of why I chose this turbo and a brief section with installation notes, a post describing performance (including logs with wideband and MAP sensor data), and a post with followup on integrity of the turbo.



Background
There seems to be some interest in this turbo, however there is very little information on it floating around. With that said, I'll be attempting to document my experiences with the Borg Warner S200 (Bullseye S256).

The compressor wheel for the S200 series comes as a 55mm/80mm 50 trim and a 58mm/80mm 44 trim. The turbine wheel is available as a 61.5mm/69.5mm 16 trim and a 64mm/74mm 20 trim. My turbo is made up of the 55/80mm compressor wheel and the 64mm/74mm turbine wheel. The housing used will be a Mitsubishi-flanged .55 A/R Bullseye stainless steel housing.

In trying to pick out my turbo, I was looking at everything from a standard PTE SCM5031E to a divided housing full T4 3582R. This turbo first caught my eye because someone was offering a deal on a brand new one which they didn't want, so, my deciding factor was actually price. Putting that aside, I also looked at the performance characteristics of the 3582R and the S256. In sheer terms of wheel size, the 3582R would apparently win given that it has a 61.4mm/82mm 56 trim compressor wheel and a 68mm turbine exducer. Pulling up and proceeding to overlay their maps as best as I could brought up some interesting characteristics, however.

You must be logged in to view this image or video.

I realize that this may be hard to read, but I'm lazy and you can feel free to try for yourself if you think you can do better. The S256 map is found at http://www.bullseyepower.com/S200.asp and the GT3582R map is found at http://www.turbobygarrett.com/turbobygarrett/catelog/Turbochargers/GT35/GT3582R_714568_1.htm

An analysis of this overlay shows that they both actually have very similar airflow characteristics, but there are a few very key differences.

1. The choke flow of the 3582R (62? lbs/min) is greater than the S256 (58? 59? lbs/min).
2. The S256 achieves choke flow at a higher PR than the 3582R (roughly 35 psi versus around 22-28 psi) based on the maps alone.
3. The surge line of the S256 is significantly further to the left and does not shift to the right like the 3582R.

So, here is a complementary picture of parts--

You must be logged in to view this image or video.

Ignore the assorted crap on my desk.



Next I want to go through the housings to point out some observations. Here's what the two turbine housings look like:

You must be logged in to view this image or video.

Unfortunately, I managed to forget to take a picture of the weird 6 bolt output pattern of the Borg Warner housing. It's basically a hexagon-pattern of 6 bolts situated on the other side the housing (which is also a circle, for anyone who wasn't completely sure). You can use your imagination to figure out how the O2 housing/downpipe would attach. IMO this would be pretty handy, but it's not very common.



The Bullseye housing is .55 A/R and stainless steel with a Mitsubishi flange. This particular Borg Warner housing is .85 A/R, steel, and has a divided T3 housing.

You must be logged in to view this image or video.

The prospect of a divided T3 setup is interesting, except that no one really makes a divided T3 exhaust manifold (nor do I forsee people starting to make them).



The Bullseye housing seemed to have some casting defects, mostly rough patches within the turbine wall, including this bugger situated at the very end of the exhaust path within the housing:

You must be logged in to view this image or video.

Luckily, the pit has no chance of hitting a turbine blade as it's situated behind the smooth wall to its right. This housing does seem rather small, especially given that the turbine wheel is almost the size of a P-trim. It would be nice to have a little better QC in terms of the casting, as well as the option for a bigger housing (.63, or a .82).



The Borg Warner housing is huge in comparison--about the size of the compressor housing (which is roughly the same size as a T04S cover). It also seems to be much better made, with smoother lines/transitions (as expected of manufacturer of their size).

You must be logged in to view this image or video.

Many people have seen the compressor cover from various websites selling the turbo, and it shows the "map enhancement groove" as a sort of inner/outer set of intake areas. I for one never understood what was going on, especially because everyone shoots the same picture from directly overhead. Here's a shot where you can actually see what happens:

You must be logged in to view this image or video.

The outer circle leads to a thin passageway through the inner circle which feeds a second set of inducer blades (SO THAT'S WHAT THEY MEANT!!). You can see one of the inner blades at the very top of the picture, and there is one underneath every exterior turbine blade.



Install Notes
It should be a pretty straightforward install for most people (not for me because I have the worst luck ever). A few key notes that people need to know:

1. I ordered a flange from Bullseye to take care of the oil feed. It takes care of the fact that the Borg Warner turbos use a BPT (British Pipe Thread), as well as any possible sealing issues. The flange also has a built in .080" restrictor, which will take care of any oil pressure issues.
2. The cover SOMEHOW manages to clear the front water pipe without needing to dent it. This may also be due to the fact that I'm running an aftermarket cast manifold which has a slightly thicker flange to move the turbo closer towards the radiator (I think). Basically, your mileage may vary, but if you do have to dent it, then it's just like any other install.
3. Contrary to what I was told, not everything from a Garrett oil drain kit will fit. I'm using a flange/push-lock setup, and while the flange fits perfectly, the bolts do not. The bolts originally on the Bullseye T04B 50 trim were M10x1.5 25, while the bolts that go in the S256 were M8x1.25 25. Just be aware that you will need two new bolts!
4. It is recommended to run an inline oil filter. David from Bullseye verified that the FP 10 micron .100" filter (the red one) would work perfectly.
5. Just like any other install, take your time and triple check everything!
6. Another reminder--USE PROPER SUPPORT. I know some people don't ever use jackstands and just assume that their jack will do the job, even though they've heard that the jacks will sometimes lose pressure and drop. Well guess what, this actually happened to me. I didn't even notice it, and you know why? I had the jackstands positioned correctly to do their job--save my dad or myself from having a car dropped on top of us.

 

[2gGSX]

First Impression
After waiting for the inordinate amount of noxious fumes to clear out of my garage due to the PB Blaster boiling off as well as cleaning up, I did what I do after any hard days work--go to Steak 'n Shake. The first thing I noticed was that I thought I accidentally took off a vacuum hose or something. There was a weird whirring noise, like a BB turbo or a jet engine. Even under low load, it was fairly noticeable with the windows up and the radio off.

After warming up for 10 minutes or so of cruising on the local roads and getting up to operating temp, I tried making a pull while getting onto the highway. I quickly discovered that I forgot to reconnect something to my wastegate's input port, and, well, ended up with a surprise. I ended up spiking to ~30 psi or a bit over it before I even knew it. The turbo is spool crazy. On my 2.3L (which happens to spool closer to a 2.0 for some reason, cams I suspect) I saw 15 psi according to the MAP sensor at a hair over 4100 RPM. Within half a second, I was at 4600 RPM and 30 psi, and saw my CEL go off signifying I was knocking above my signal point set in DSMLink (peak of 6.4 degrees knock!). After that I just picked up my friend and went to eat a Chili Mac Supreme.

You can see the log for yourself (run11032007-2 pull.dat) using the DSMLink software, which is free for download off their website. If you do look at it, you can see that I have timing and fuel zeroed out (new turbo = time for a new tune is my take on it). A few things to note--

• The max airflow is ~44.3 lbs/min, at a whopping 4777 RPM and a maxed out MAP sensor (so, at least 30.7 psi)
• The ECU seems to freak out in a big way and dumps as much fuel as is possible. I don't have a wideband up, but 100% on ~700cc injectors (rough estimate based on how much fuel I have to pull globally to get decent fuel trims), while only flowing 43.8 lbs/min (see data at 6.80s) puts me at ~9.3:1. So, running the stock fuel maps was a big big mistake. Not taking the time to set base fuel pressure was also a big mistake.
• The spoolup is moderately fast, but once it hits it hits like a freight train.

First Pulls
Reference run11032007-2 pull.dat for the 2nd gear hoo-ha, and run11052007-2 3rd gear pull.dat for the stock tune data.
After getting everything hooked up I managed to get a (still untuned) pull through most of third gear. The log is listed as attachment "run11052007-2 3rd gear pull". The important parts first.

• 15.6 psi comes on at 3800 RPM in third gear. Please note that this is a 2.3L with spoolup characters similar to a 2.0. In comparison, the old T04B 50 trim spooled up about the same if not a tad bit slower. Unfortunately, I don't have any logs with MAP sensor data.
• Peak airflow at 23 psi was 43 lbs/min by roughly 6200 RPM.
• I really need to tune.

It's important to note that all this is on a completely stock tune, which come to think of it may have been a good idea since it removes an extra variable. With some tuning tricks, I'm confident that the 15 psi spool-up can be attained at least 300 RPM earlier.

Another interesting point is the massive airflow difference between 30 psi and 23 psi. This turbo is definitely not the same as your average 50 trim which will shine at around 26 to 30 psi--this thing really really likes boost. This is especially apparent when you look at the map, which shows that the turbo should really be happy around 30-36 psi. At this point, I'm considering if I need to pull the head off again and get it O-ringed or something. If you compare the log of the 30 psi spike and the log of the third gear pull at 23 psi, you'll notice that at 30 psi and much less load, the turbo performs amazingly by flowing more air at ~4600 RPM than it did at 23 psi and up through 7000 RPM. Bottom line, if you're interested in this turbo, be prepared to run lots and lots of boost to get in the sweet spot of the wheel.

After Tuning (sort of)
Reference the file run11172007-5 3rd gear pull.dat for the pseudo-tuned data.
With boost set at 23 psi, the car made full boost at 4500-4700 RPM in 2nd gear, and 3900-4000 RPM in 3rd gear. Given that I haven't increased the boost at all, airflow numbers are still the same as before. The 70-90 mph time is approximately 2.1 seconds throughout these pulls.

All in all, the car pulls very very hard and meets/exceeds my expectations. If you're in the market for a 50 trim/3052/60-1/3065/GT35(61mm wheel) sized turbo, then this or the S258 should be considered. The turbo has the spool characteristics of a 50 trim, but *should* be able to flow as much as the larger turbos.

E85
So, I realized that the Dejon SMIC isn't really cutting it anymore but I don't have the money to drop on a FMIC or injection kit right now. I did however have 20 bucks to upgrade to FIC950's after some hawking on the classifieds, so I went with that and some E85 instead. The stuff is definitely strong enough to make up for my lack of intercooler, and between the latent heat and extra octane all the knock is gone. So far all I've done is adjust my global fuel and adjust for LTFTs, lean the WOT map out to ~10.5:1, and zero out my timing map. The car feels much better now that it actually has some timing and the 70-90 is down to 1.9 seconds. Nothing else has changed, including boost and peak airflow. With any luck the tranny will hold up to a little more abuse and I'll be able to up the boost some more before dynoing or trying to make a pass down the track. Log for this data is run02242008-4 two third gear pulls.

 

[2gGSX]

Two weeks have gone by and everything is fine as expected. The turbo has put up with plenty of use, and there is nothing wrong at all. Visually, everything still looks the same, and shaftplay is in check. On another note, the oil feed flange is sealing perfectly, and I would recommend it to anyone ordering a Borg Warner turbo.

About 4 months have gone by now, although two of those consisted of limping around on a dying clutch and breaking the new one in. The turbo is still fine in terms of shaft-play and blow by (zero). I haven't had any troubles with this thing, knock on wood.