The Garrett range of turbochargers are well known, tried and tested on Coupes. It's pretty much de rigeur: T28 and GT28-based turbos for the 20VTs and T3 and T04-based turbos for the 16VTs, with a very few exceptions to the rules. But time move on and new products come to the market. Understandably most people don't want to gamble their hard-earned trying out new products. So that's where good tuners like me come in we risk our money trying out new stuff so you don't have to!

If you followed my latest dyno results thread in the Rolling Road section, HERE, you'll have read that the results were actually a consequential part of testing a new make of turbocharger. The purpose of the test was assess the performance of this new turbocharger and, if it performed well, develop a kit including the turbo and all the oil/water fittings to make it as easy as possible for users to bolt on and go (after mapping of course).

So the first test subject is my 16VT.

Brief spec
>fully rebuilt engine
>Wossner forged pistons (8:1 comp ratio)
>ZRP forged conrods
>Custom uprated cams on Voodoo Tuning (VT) adjustable cam pulleys
>stock cylinder head, no major headwork
> Tubular exhaust manifold
>Precision external wastegate @ 1.55 bar
>VT 3" custom downpipe with V-band turbo flange
>2.5" exhaust system and decat

The turbo
Kinugawa turbocharger. Fairy new on the scene, their relativel well known in the Scooby and Evo world but pretty much unheard of outside.
My previous turbo was a Turbonetics T3/T04E 50-trim, known to be good for approx 450 bhp. I wanted a slight upgrade if possible so I chose the TD06SL2 25G, 8cm hot side, anti-surge cold-side, journal bearing. Apparantly good for over 500 bhp.



A quick word on Mitsubishi-based turbos
It's probably fair to say most of us on this forum are familiar with Garrett-based turbos and their designations; T3, T28, a/r etc. MHI (Mitsubishi Heavy Industries)-based turbos, like Kinugawa, use different desgnations. Take this turbo the TD06SL2 25G. "TD06SL2" refers to the turbine. 25G refers to the compressor wheel. TD06 25G is the biggest turbo that Kinugawa do. Above I mentioned "8cm housing". MHI don't designate turbine housings by area/radius ratio, like Garrett (Google if you want to know more about a/r ratios!). Instead they are designated by the area part-way down the scroll instead. You can't directly compare Garrett a/r's with MHI sizes, but I can say that the TD06 8cm housing spools slightly quicker than the a comparable Garrett with a .63 a/r housing. The 10cm spools about 500 rpm slower than the 8cm and is probably comparable to a Garrett .89 a/r housing. Having tried the 10cm housing can say it's a not a great choice for a road-going 16VT.

8cm housing vs 10cm housing



Fitting the turbo
First thing I noticed when picking up the turbo is it's significantly smaller and lighter than the Turbonetics T3/To4e, which is nice, because the T.netics is a fairly hefty ol' lump.

Second thing I noticed is [unsurprisingly] almost nothing fits! This turbo comes from the factory with a really nice fitting kit, nonetheless all the fittings are meant for Japanese cars, hence the only thing that's bolts straight on is the T3 turbine flange. I had to make several custom parts. The most difficult of which to make was the ol return fittings. The measurements were critical; few mm to long or too short, or 2 deg angle either way and the oil and the oil return fitting would end up going uphill, which would mean that oil would pool inside the turbo and eventually work past the seals.

Custom oil hose. Final design will be fluoro-lined silicone


Custom oil return turbo flange:


checking oil return angle


Oil return fitting in place




Braided oil feed hose with anodised aly banjo union



Braided coolant hoses fit onto original Coupe coolant hoses.


Testing
Next on to the dyno to compare with the previous turbo. In order to make a truly useful test you need to control as many variables as possible. All dyno runs were done on the same car with the same tyres, at the same tyre pressurs and on the same dyno, with the same settings. Both remapped to suit. Same boost level too. The only difference is the turbo.

Results
Turbonetics T3/T04e 50-trim: 396 bhp ATF / 337 bhp ATW, 379 lbsft

Kinugawa TD06SL2 25G, 8cm: 413 bhp ATF / 352 bhp ATW, 395 lbsft


Boost and AFR. AFR was a little sub-optimal on the old T.netics pull but only worth approx. 4 peak bhp and I don't think it had an effect on peak torque.


The advantages of the TD06 25G over the T3/T04E are plain to see from the dyno graphs: a mighty +17 hp ATF (+15 bhp ATF) and +15 lbsft - not forgetting this is with the same boost. Not only that, but the boost plot shows that the TD06 25G spools a good 400 rpm sooner than the T3/T04E. This is a journal bearing turbo, yet it's spooling with the same sort of speed you get with the ball bearing version of the T3/To4E!

On the road the spool is even better than on the dyno (Dyno Dynamics rollers don't seem to load up 16VT engines enough for some reason). On the road the turbo hits 1 bar (which, in my car, throws you back against the seat!) around 3700 rom and hits full boost around 4500. Compared to around 4000 rpm and 4800 rpm for the T3/T04E. Power delivery is somewhat less "all or nothing". Lag between gear changes is reduced. Don't get me wrong, it's a big turbo so it still lags, bet less than the T3/To4E

Another important thing worth mentioning is on-boost compressor surge. The T3/T04E could surge a bit at part-throttle. But with the TD06 25G there's no hint of surge at all.

Conclusion
So in summary, testing shows the Kinugawa TD06SL2 25G 8cm, compared to the Turbonetics T3/T04E 50-trim:

>Makes more power and torque
>Has a smaller profile and lower weight
>lower boost threshold
>less lag
>Less prone to surge.

Also, I forgot to mention it's about £200 cheaper!

So no surprise to find I'm impressed with this turbo so far. I'll definitely be adding Kinugawa turbos to my range of performance items.

What's next?
Select another Kinugawa turbo for my 20VT and repeat the process (build, develop fitting kit, test, write up results). I noticed recently that a lot of 20VT owners seem to be interested in more "down to earth" performance. So this time I will be using an internal-wastegate turbo, aiming for around 320-350 bhp, with low boost threshold and low lag.

Nice one, good to see something different on the scene, and an interesting post to read even for technical numptys like me!

I came across this turbo not long ago but didn't know anyone trying it on an FC and then you posted your thoughts on it just after I fitted a brand new RS... I wish that you have posted earlier I would have tried the Kinugawa on my 20VT.

Why do you use so much fuel? AFR below 12 is not good if you ask me. You are losing a lot of power.

Back in 2009 when I took my car to the RR I could see a big difference when you change the AFR.

First run I had 0.77 (lambda) @ 6100 and 291 whp.

I turned down the fuel and at the next run I had 0.86 (lambda) @ 6100 and 312 whp.

So +21 hp just by reducing the fuel.

Because that's what my Coupe likes! Tuning the AFR isn't just about making power. On 16VTs the EGTs tend, depending on the engine spec, to rise sharply with AFRs above 11.5, that includes mine too. I did run at 12:1 at points during tuning and it certainly doesn't lose anything like 21 bhp. But this is my engine and that was yours. When you tune an engine you tune to what's best for that individual engine, not to a general rule of thumb - because there isn't one! Some turbocharged engines can tolerate AFRs as high as 13:1, others will nuke themselves above 11.

Ask Porsche why their 90s cars ran so rich.

Great informative post Kj.
An inspiring read.

Cheers

Thanks for sharing experience. I looked this kinugawa turbos some time ago and they also made a (not actually in offer) 25g in twin scroll bolt on variant for lancer evo. Ultimate compact specs for an 450ps 4cyl roadcar i think.

To comparing T3T04e, you did not much mean, that your has .63 housing, so as you explaned against 8cm, its bit bigger=more laggy. Also turbine wheel is not always the same on other T3T04e hybrids.
I only want to say for others, its not overal review/comparation to "T3T4" at all. Its comparation to particular turbonetics T3T4 (which has particular part number).

This i meaning only for defence old garrett unit :))

Personaly runned T3T04e with .48 and smaller turbine wheel, and looks similar characteristic like your new 25g.


Do you planning more boost?

Hi, PhoB. Good point, I didn't mention wheel sizes.

Kinugawa 25G comp: 61.5mm in / 78.6mm ex, 60 trim
Kinugawa TD06SL2 turb: 61mm in / 54mm ex, 78 trim

Turbonetics T04e 50-trim comp: 53.9mm in / 76.2mm ex
Turbonetics st. 3 F1 turb: 64.9mm in / 56.6mm ex, 76 trim

Size-wise the TD06SL2 wheel is much closer to a Garrett (not Turbonetics) st. 3 wheel at 65.5mm / 54mm. It also looks similar to a Garrett st.3 wheel, with big gaps between the exducer blades.

Hmm, I could be wrong here but I'd have thought a 48 a/r housing and smaller wheel would struggle to break 400 bhp? St.1 wheel absolutely no chance - more like 300 bhp. St.2 wheel, maybe - just!

Yep I do plan more boost in the future. Want to do something about handling and paintwork first though!

Nice work KJ. 25G, wow that is massive! Even for my own amusement, I like that you have gone out on a bit on a limb there with such a large comp. It would be nice to run a 20G back to back as that would be the more common choice for that sort of power, although the same setup on an evo would probably make more power.

I like the Kinugawa turbos but I would only bother if you were keeping the standard manifold and internal wastegate. Although the standard Kinugawa WG does not look like it will fit. If you want to build a manifold the 10.5cm twinscroll Evo fitment would be awesome to try. I think there is a guy in France with a stock Evo 16G setup. I think a 6cm TDO5 16G would make a pretty fantastic replacement for the stock turbo. Likewise, maybe a TDO5 18G 8cm for a healthy upgrade.

btw MHI turbine sizes are still a representation of A/R just like Garrett only they measure it in centimetres where Garrett of course use inches. A Garrett .63 turbine is actually .63 inches but they drop the inches. A/R has dimensions R^2/R which cancels to give just R in whatever scale you measure it in.

Is the turbine really that big? I've had these crazy thoughts about a GT30 turbine in a stock .48 housing, Its the same size as a T3 only a much larger trim. With the 71 or (GTX)67 comp it would perform well but top end power would always be limited.

With stage2 turbine wheel(only a little smaller than stage3, but much bigger than stock stage1) and ar .48(its aftemarket 5bolt pattern, not stock fiat) i have proven 400hp+ on two different car(one was 16vt) at two different dynos.
But of course, not always smaller AR means faster spool with solid power.

Im considering my new turbo, the sizes of 25g are area where im looking for, but i didnt saw/find any 500-550hp cars with it. Also kinugawa didnt spec it over 500hp. Therefore i asked if you are looking for rising the boost

Cheers. It's funny, I've been so used to great big E covers that the 25G looks almost fun-size to me! The turbo as a whole is certainly a lot smaller than the T3/T4e frames. It leaves a lot more space around it, particularly by the gearbox bell housing - this a bit of a problem area (on manifolds that place the turbo in the stock position).

Yeah there are several MHI turbos that would make great Coupe upgrades, from stock to big power. I'll be looking into one or two of them as "intermediate" upgrades. I'll probably still be keeping my T3-to-T34 upgrade as the low-cost 300+ bhp option. Its uses all the stock fittings, including downpipe flange, so it's an easy "entry level" upgrade.

This TD06SL2 25G will definitely be my "big daddy" turbo upgrade kit though. BTW, I can also get an internal wastegate version (5-bolt downpipe flange only) and there's even a billet 25G comp version too.

When I get a Kinugawa turbo to test on the 20VT it will definitely be internal wastegate. Not sure about the rest yet - maybe TD05 18G.



Actually a/r is purely a ratio, area / radius, it has no units. It doesn't matter what units you measure the area and radius in, the ratio will always be the same. If you were to measure the CSA of the scroll and measure its radius from the centre of the turbine at any point the ratio would be the same. This picture is a lot easier to understand than what I just wrote!


MHI's "8cm", on the other hand, is a measurement of the CSA of the scroll at one specific distance from the mouth of the turb. housing.

That's good to know. Do they ever suffer from surge? That was a big problem when I was running a To4e on a Garrett 48 a/r, st.1 wheel.



When researching MHI turbos I spoke to a Evo specialist and he said that it can make over 500 bhp on Evos. Looking around on the net it seems that Skylines make that sort of power too. Guess we'll have to see what it can do on a Coupe!

Semantics really, it's the same thing. Sure I agree A/R is a dimensionless ratio, that's why when you actually measure the turbine A/R it will be the same where ever you measure it. It can also be expressed as a physical measurement of Area divided by Radius which will have units of length. FWIW Garrett use this definition. And, as A/R is usually written as A/R (i.e. Area divided by Radius) not A:R so I prefer to use the Garrett definition.

Terminology



Sure it is, but it's the same thing as A/R. MHI just like to complicate the whole thing solving the flow equation for A not Q

Good work and result KJ, couple of observations

The 25G is a fair bit of compressor when compared to the T04e (ind 61.5mm vs 53.9mm) with a slightly larger back plate, so there would be an expectation that the compressor would be good for more bhp, and I note you have picked up spool, with the TD06SL2 being a smaller turbine and newer turbine design

Have you measured exhaust back pressure? - I suspect it would be higher in the Kinugawa, how was the EGT's comparability which of course has flow on effects

BTW - I never had any luck luck breaking 400bhp on a STG 2 garrett turbine

Very good thread KJ !

I appreciated the time you spend to describe the details of your comparison.

I have been interested in MHI turbos for a few years..

The best dyno results on dyno for an evo with TD06h-20G was around 600whp (don't know the pressure and it was on e85...these turbos are able to run above 2.0 bar and 8000rpm without any problem)

There are 3 16VT coupe in France which runs EVO twinscroll manifold + td05H-big 16G on e85...it's just awesome : full boost before 3000rpm and 350hp+ @ 1.5bar

For the comparison with the TD05H-18G :
I'm running one on my 16VT coupe since 2 years...
- at the beginning whith internal WG and 8cm housing : it's not so laggy (full boost @ 4000rpm with stock header) and it can easily breaks 330hp+ @ 1.4bar
- now on external wg & e85 : should output something like 360hp+ @ 1.5-1.6 bar

As told KJ, these turbos are quite ON/OFF and give you a very good compromise for the price.

Julian, has anyone tried the Kinugawa evo fitment turbo or just OEM MHI turbos? There is some talk about Kinugawa not replicating MHI but making their own "improvements".

Are you using the T3 fitment TDO5 18G Kinugawa turbo?

Yes sorry, they are ALL runing kunigawa turbos...

We have also a 20vt with TD06H-20G and another with TD06SL2-20G they both run more than 400hp.

I noticed kunigawa compressor wheel on my 18G is lighter than MHI one (like 80gr vs 70gr) and the center of the wheel is smaller (it will flow more air).

I'm using this turbo with T3 flange yes

KJ, a very good test considering a rolling road was used and on different days. I however would still hold judgment of your conclusion by only plotting basic values such as BHP/TRQ/Lambda/Boost. I'm not trying to discourage your testing/research as it's good to see science projects in action and well executed, this is very rare for somebody to openly publicise results that all could benefit from so well done.

This issue with turbocharger selection is that it is genuinely a very complicated task and in order to conclude your results I would have liked to see at least the following items instrumented and logged in the same data set. I fully appreciate that this is a large amount of time and money but that's the cost of development.

Air Charge Temperature
Ambient Air Temperature
Barometric Pressure
Dyno Correction Factor
Engine Speed
Engine Torque
Exhaust Gas Temperature
Fuel Pressure
Ignition Advance
Knock Activity
Lambda
Manifold Pressure
Mass Air Flow
Pre Turbo Pressure
Throttle Position
Turbo Speed

Please don't feel like I'm having a dig because I'm not.

TJ

I noticed when I used the precision turbo that the coolant temp ran at 95 degrees which was 10degrees hotter than on garrett.
Also it ran nearly 200degrees hotter on the precision than garret regarding egt but did make more power at 1.6 bar all be it it came on 1500rpm later.
Bizarrely it ran slightly cooler on intake temps with comparable ambient.

Overall the garrett was a much better turbo and more linear progression of boost hence why I reverted back to it.
Surge on the gtx is a big problem which is why i still fit the gt series.

No offence taken

Believe me, I'd love to test ALL of those parameters and more - on a whole selection of turbos back-to-back. But, just as you mentioned, time and money (definitely money!) prevents this. Something I'd love to do in the future. Or if you do it first let us know!

For sure, this isn't a totally scientifically rigorous test. But it is definitely an objective test, the purpose of which to give people some genuinely useful information, show what I'm up to at the moment with regards to Coupe tuning, and to spark some interesting conversation. And it's doing that perfectly!

Hey guys, I have just fitted one of these turbos in to my 20VT Fiat Coupe and I would like to use this topic to share my first impression.
I bought almost the same turbo as the one above
4" TD06 SL2 25G billet wheel with 8cm T25 turbine housing with internal WG.




The turbo looks pretty good and the quality seems to be as good as it looks.

My engine is pretty standard:
- stock cams
- stock head
- stock pistons
- ZRP Conrods
- stock exhaust manifold

I wanted to fit the turbo to the standard manifold and surprisingly it did









The spool of the turbo is pretty good for its dimension - on 3500rpm 0.5bar , 4200rpm 1.4bar of boost.
So far I don't have any graph or I can't tel how fast the car is, because I need to finish my 2.75" exhaust system and then to be fully remapped.

I bought almost the same turbo,i have the Td06SL2 with the garrett 60-1 wheel,i have buold it on a 2.4 block....so far as i drove it i have positive boost from approx 1500rpm.
It only seems that i have to restrict the oil pressure to the turbo.

You should be able to see around 450hp with this kind of turbo. The turbine will be the limiting point of your setup, as the 20G compressor wheel manage to max out the SL2 turbine wheel flow, the only gain with the 25G compared to 20G will be the lower intake temp.
If you want bigger power, you will need external wastegate setup or/and TD06SL2-9blades/TD06H turbine wheel.

The 25G wheel is a 600hp+ capable wheel.

Yes, I know. For now I am looking for 400-420Hp , but wanted to have some space for future if I decide to go fully forged.
Kinugawa offers very cheap turbine housings, and I can change the setup from 450 to 600 very easy if I decide to. Or I can get the 9 blade wheel and hit the 500

You have stock internals??

Good luck

Just std pistons. I have ZRP conrods. I don't thing that i need luck for 400hp

Think you'd have been better off with forged pistons/standard rods
I'm sure it will be fine though with correct mapping!