Dyno day: the story (numbers inside graphs to come Options

[KOU In3]

All the tests were done on a ’90 Eclipse GS-Turbo. The car was in the process of being returned to stock for resale. We did learn a few things in the process of this dyno comparison. Largely that there is good information to be had but most of what we garnished needs interpretation and explanation. I’ll furnish the step by step info, then give my own interpretation, but also provide the raw data so you can draw your own conclusions. One of the difficulties we ran into was one of the main benefits of DSM ownership. Tunabilty was definitely a double edged sword here. With so many factors that can be played with on a given set-up. We tried to keep everything consistent making few or no adjustments in order to produce as true a comparison as we could. Here’s what we found:

Performance mods:
3” down pipe and catback
larger than stock fuel pump
mbc @ 15psi,
thacked air can (paper filter)
stock 14B and largely stock car
please note the stock ’90 cams with their lesser top-end
93 octane pump gas used

We ran two back to back runs with each manifold in order to have a verified number for each baseline run.

First two runs were made with a professionally ported 2G ex-man. This had been bolted on the night before when all the exhaust studs were changed in preparation of the upcoming dyno. This manifold had come from a car successfully running 12.6 on a 14B set-up with full weight, full interior, and AC so we knew it would be a good candidate for comparison. Unfortuanately it was bolted to a stock 14B so there was not the degree of port matching that would make it as solid a performer as on a port matched 7cm turbine housing that it was intended for.

1st run: 224.48hp @ 5800rpm 2g ported
2nd run: 222.99hp @ 5800rpm 2g ported

Second two runs were made with a stock 1G ex-man. The swap was straightforward and took only a few minutes. I consider this one to have been an accurate baseline run for the car as no issued presented themselves for explanation. Smooth runs and repeatedable results.

1st run: 232.63hp @ 5800 rpm 1G stock
2nd run: 228.92hp @ 5800 rpm 1G stock

I’m sure you’ve noticed the initially surprising gain over the ported 2G ex-man we’ve all grown to love. Take a 7cm opening on the ex-man at the turbo side and slam all those gasses into a 6cm turbine housing. Although the video doesn’t so an obvious exhaust leak I’d assume one was still present from the overlap. Secondly there is obviously some flow issue as there is no longer a smooth transition.

Now we went for the SSautochrome stainless steel tubular manifold. This turned out not to be as easy a swap as the 1G to 2G. Here’s what we ran into. I’ll try to list it chronologically. Ex-man studs. They had been replaced with slightly longer studs the night before. There is enough clearance for stuck studs but extended ones will hit the tubes on the 3rd and 4th runners. We trimmed those down. Turbo location. The 14b will shift to the passenger side and down (towards the ground) by about 2 inches overall (imagine the 2inch diagonal line). With the 20G install I’d taken pictures of this was not a problem. Especially given the lightly cunstomized install of most big turbo set-ups (bolt on but still custom to most).

On a 14B set-up this will pose a problem with your water and oil lines. We used a brake line cutter to sever the water line and installed a length of hose to gain the couple inches of necessary length. The oil line was ‘massaged’ a bit and made the grade. The downpipe was the next issue. It took a lot of leverage to move everything in place with one of us under the car and another above. The FWD crossmember is located differently than the AWD version. Although we were able to bolt everything back up and the flex section did let everything realign, I’m not happy with the downpipe’s placement directly against the crossmember.

Next hurdle, the radiator fan. Ultimately we left the fans out for the runs so as not to hack the owner’s fan. The turbo outlet elbow on a 14B set-up sticks out just enough to hit the fan support arm. It would need to be hacked on a fully stock set-up. Again something I hadn’t run into on the 20G install given the different routing of the piping.

It took us much longer trying to miminize what we removed in the interest of speed (we were strapped to the dyno, on the lift, and on the clock. My reccomendation to those of you planning the install is this:

Expect it to take an afternoon not an hour:

Remove the following:
Manifold
Turbo
Lines
Radiator (with fans attached)
Downpipe

Install:
Manifold
Bolt in turbo
Reattach lines to turbo
Slide radiator complete as a unit
Bolt downpipe back in
Re-attach turbo outlet elbow
Re-attach radiator hoses
Add coolant and check oil levels

I was able to straighten and rebend my turbo lines by hand on the 20G install and it is most likely the route I would take again. Those of you running stainless braided lines likely won’t run into an issue here.

Personally the install, although in depth, I don’t consider to be highly technical or difficult. The FWD downpipe placement was not ideal though. I do think we ran into an issue of aftermarket parts mating to aftermarket parts that might have slightly exaggerated the issue though. Basicly, expect run into similair problems as you would when installing a 20g and assume it won’t be an hour install.

Now the important snafu: the turbo mounting flange. The bolt pattern was correct. But the outlet shape of the flange on the test sample here was not. The 14B has a round turbo inlet. Our tubular manifold had a square outlet. Round peg in a square hole doesn’t flow too well. We bolted it up knowing the leak potential. That led to our next problem. The included bolts did not match the turbine housing threads. The stock bolts are too long to be of use. That left us scrounging in R&T Unlimited’s (the dyno shop) box of misc. spare bolts for some suitable candidates. Although a couple of the bolts were iffy, we did get it bolted up. Note that there were gaskets included with the sample we dynoed. The manifold to turbine housing gasket also had the square hole.

On to the runs:
We noted a visible and audible exhaust leak at idle so we went back to re-torque. We’d left a couple loose somehow. Tightened down we went for a couple pulls.

1st run: 235.09 @ 6000rpm
2nd run: 230.22 @ 5900 rpm

Conclusion: You can evaluate the data in a nearly infinite number of ways. There’s even the marketing claim of an 11hp peak gain over the ported 2G ex-man. My own particular assessment of it is this. The flange doesn’t match. There is a both poor flow from this as well as a visible exhaust leak. You can see the exhaust pooling out around the radiator and escaping from the mating surface of the header and the turbine housing. It’s most likely what blew out the gasket in the run. If you go frame by frame through the video you can actually see the gasket blow out. Additionally, most headers that I’ve seen tend to start to glow at the thinnest portion of the header, the tubes. The way in which it consistently glows at the juncture with the turbine housing shows the restriction there. Despite the flange, despite the leak, taking it all in, the best power numbers were still made with the header. At its worst, it still makes more power than the other set-ups did.

There are two things I’d like to see. First and foremost, the flange to be changed to the correct Mitsu flange. I’m hoping that SSautochrome will alter the ones they have already made and make them available for the group buy. Issues of 6cm vs. 7cm opening should be addressed as well to avoid a similair issues to the ported 2G ex-man runs on a 6cm housing.

Second, I’d really like to see a big turbo run. Everyone present seemed to be in consensus that a less restrictive turbo with properly matched flanges had the potential to see HUGE gains here. But I’d like to see it rather than talk about it.

I’ll be posting up the actual dyno charts themselves as soon as I have a free moment. There’s a lot of data to analyze here beyond the peak numbers. We’ve got some a/f numbers to show if there is more power to be made simply by turning up the boost (we’ve got one unpublished run that shows yes). As well as the power band to analyze. We’ve also got datalogs of most of the runs to go through as well.

Hope this helps for starters. We haven’t been idle on this end. Just trying to send through what I can whenever I have a spare moment (90 hour work weeks).

Apologizing in advance for typos and poor wording. Too many 3 hours sleep nights are getting to me here.

Have to mention all the incredible help and efforts by Slopony Automotive, Crebotech, R&T Unlimited, and DSMcentral. All of the aforementioned pitched in to help me out with this. There is no way the dyno comparison (even with all its issues) would have happened without all these guys. Many of whom are on this site. So I have to say a quick thanks there.

[90GSX]

Actually to save you time with the digital camera, you can just push print screen button on your keyboard and then open your paint program and go to edit-paste and it will give you a screenshot of whats on your screen.