Looks Like Garrett Slipped Another Gtx30 Variant In .

[zebra]

Yeah, damn confusing garrett numbers

[M@&k]

Just curious - why only 330rwhp with a GTX3076 at 18psi?? You should be much closer to 400rwhp.

Get's em every time

Tuner didn't mention what was causing the drop off specifically. He's convinced the turn-flow intercooler is holding it back, and was looking into setting up a pressure reading at turbo outlet to show me pressure drop across the cooler but ran out of time.

Wish I had more experience with other turbos, because I don't really have much to compare to. I will say that it seems to build boost in lower gears well too, in 2nd I feel like full boost is not far behind 3,500 rpm as well. In 4th and 5th, positive pressure is being made pretty much as soon as you start on the throttle.

The drop off up top is noticeable now, but I rarely get the chance to wind out above 5,500 so suits me fine.

In terms of peak power I do still feel it's maybe a bit low. At this level I don't feel it offers much over GT-RS or for example the new GTX2867 kits being offered with bolt on rear housings. Seems like something with my setup is possibly not quite right, but I spent a lot of time making sure I had all the right supporting mods in place...

But chasing numbers for kudos on SAU aside, it's definitely heaps of fun, and positive side effect is that I already am setup now for GT30 turbos if I want to go bigger sometime in the future.

[M@&k]

Below is rough plot of boost on compressor map. The map with speed lines was taken from VW forum where disco was asking about this turbo. I assumed these are correct and have just plotted boost plot from last dyno run on there. I guess the speed lines would shift a bit between the two housings as the vol efficiency changes, but I guess close enough for an idea.

Anyway, interested if anyone has any comments.

[Skepticism]

Below is rough plot of boost on compressor map. The map with speed lines was taken from VW forum where disco was asking about this turbo. I assumed these are correct and have just plotted boost plot from last dyno run on there. I guess the speed lines would shift a bit between the two housings as the vol efficiency changes, but I guess close enough for an idea.

Anyway, interested if anyone has any comments.

GTX3067 comp map with boost trace.JPG

Speed lines should be the same - boost is a function of compressor speed and airflow. so whether the power source driving the comp wheel is a turbine, or the crankshaft directly as in a supercharger, etc shouldn't change anything. it's defined by the aerodynamics of the comp wheel and I guess comp housing

The one thing that will change is engine VE.. the smaller housing should take a bit less timing due to being restrictive, so (unless I'm wrong, which wouldn't surprise me) I reckon this would have to be guesstimated when converting the HP to obtain the airflow in lbs/min

with the smaller ex housing, a given HP would need slightly more airflow as the VE is lower due to less timing. And the converse for the larger housing.

Maybe your calculations already took that into account tho, I dunno! It's pretty easy to work stuff out with tools like Matchbot around otherwise I'd have no idea.

Edited December 12, 2014 by Skepticism

[XKLABA]

Is it possible that that is just where the 67mm comp wheel runs out of puff trying to feed a 2.5ltr six ?

Mark, what boost controller have you got ?, can you turn the gain down a bit to smooth out that peak which should reduce how much it dips after full boost is reached

[M@&k]

Maybe your calculations already took that into account tho, I dunno! It's pretty easy to work stuff out with tools like Matchbot around otherwise I'd have no idea.

Didn't do the calcs for the speed lines myself. Was easier just to grab the pic off the internet. Everyone knows info on the net is always trustworthy.

Is it possible that that is just where the 67mm comp wheel runs out of puff trying to feed a 2.5ltr six ?

Mark, what boost controller have you got ?, can you turn the gain down a bit to smooth out that peak which should reduce how much it dips after full boost is reached

I also suspect this may be the case, and possibly supported by the plot on the comp map. Why I put it up actually.

Have a Profec boost controller. I'm sure the gain can be adjusted, but not too confident doing that myself to be honest.

[discopotato03]

Yes an interesting result .

What I think upsets perceptions is todays use of E70 / E85 and what that allows you to do timing wise . People do get higher numbers with ethanol like my 271 wheel wasps when I ran the GTRS . To put that into perspective I think Micko had 270 odd with his GTX3071R when running 98 PULP , open to correction if that number is wrong .

I noticed that Checkbuzz (Sp ?) got a little over 300 wheel wasps out of his 3076R 52T 0.82AR so maybe in some niches the old GT comp jobs do alright .

Anyway it could be possible that like the older GT3071R the GTX3067R isn't a perfect animal , both are GT28 turbo upgrades and its possible that the cropped GT30 turbine is a better compromise in this area . Many felt that HKS had it right with the GT2835 Pro S and Garrett seemingly hasn't really been able to get equal results . All that really stopped them was the lack of a T3 flanged IW turbine housing for the cropped GT30 turbine . I think they are doing the larger (90) trim size version of this cropped turbine for the new GTX2971R but not sure about a T3 IW housing to suit .

Bottom line is that you picked up 500 revs which is significant and another getting rare T3 IW GT30 turbine housing .

Cheers A .

[M@&k]

Mick got over 280 on 98 I believe. 288 or something?

Yep much better now with this housing, and worthwhile change. Should have had it from the beginning.

The way it tapers off above 5,500 makes me wonder if I haven't still got a restriction in the exhaust somewhere ( maybe the mufflers). Next year I might consider changing intercooler to Plazmaman setup piping over the rocker covers with stock manifold. Results from Hypergear thread (and Mick_o's reminder) pointed out this restriction is bigger than I realised.

Edited December 20, 2014 by M@&k

[skylinecouple]

I can agree with some that this turbo has definitely run out of puff. A boost leak check would go a long way to seeing if this was the case. after that we can assume the compressor has done all it can. We could guesstimate for hours but really, it is what it is. If you need any help with the Profec when you install it let me know. Mine has done well for around 10 years now. Will bring boost on earlier and help to hold it. I haven't read the entire post but have you got an 18lb wastegate actuator ? This will be a must with the smaller exhaust housing. From here there is only E85. A must for smaller exhaust housings where top end timing is limited.

I've never been a fan of smaller turbine/compressor setups as boost comes on too quickly and results in instant wheel spin. Good for slower drifting courses though. Not so good for real horse power/ torque that gets the car moving in the higher gears. should be a hoot at the traffic lights though as my 3071 is on boost instantly so yours should be even faster.

[dionkouki]

Thinking out loud here so if anyone would like to correct me or add feel free. I was considering a GTX3067 for my RB25 however the results so far are confusing. I currently have a 56 trim gt2871 and making 320whp on a mustang dyno. I would expect the GTX3067 to make more power but perhaps it has to do with different mods or possibly the power difference I've read about (AUS vs USA). The thing that confuses me is the GTX2867 has been shown to hit 400whp and more on 2.0 liter motors as Lithium commented on. A tuner I talked to mentioned he's tuned several 2.0 liters with the GTX2867 and was quite impressed that they were hitting these power levels. Keep in mind most of these are running 92 ocatane. Is the GTX2867 just a better turbine/compressor match? I would think the GTX3067 turbine wheel would be better suited to a rb25 plus I could actually get a Garrett T3 GT (3" outlet) housing instead of the T31 (2.5" outlet) I'm limited to with a T28 frame turbo.

Another point, RB25s over here are reaching 400whp or close to it with GT3071 so why is the GTX3067 having so much trouble. The only other detailed results I could find were for someone that is building a kit with a gtx3067 TIAL .63 A/R for subarus hitting 350awhp. Similar to this thread it had great response, just lower power than expected. Now that I think about it that subaru was running an oem TMIC and getting heat soaked, so perhaps there's a little more left to be gained.

[AngryRB]

It looks like the compressor map says it does have more left, 16/17psi you'd think is far from game over with the GTX compressors, with the right setup eg. E85, boost, timing, external gate's , a good tuner ,

then it could become a different beast again.

[M@&k]

We tried running more boost, but that was as high as it'd go. Prime suspect at the moment is the turn-flow intercooler. Graph from Stao in Hypergear thread showed as much as a 5psi loss over normal cooler which would make quite a difference.

Changing cooler is on the drawing board for this year.

[Mick_o]

With all due respect to Marks result & some possible bugs within his setup i think you're better off with a GTX3071R. My car had better response made more power everywhere than the 3067. I believe its simply too small of a turbo for a 25. Unless you plan on running WMI or E85 the .63 chokes 25's. Its a different story with 2L 4 bangers

[M@&k]

From my experience thus far I'd tend to agree. With standard sort of mods it's not doing much more than a gt28 turbo, and response even with 0.63 housing isn't mind blowing.

[lcxu105]

Dose anyone know how to tell what the turbine housing size is from the number stamped on the flange of the housing. Part Number is 1679-19 from what I was told it supposed to be a .82 housing. But wount build boost until around 5500 to 6000rpm.

[lcxu105]

Forgot to mention its a gtx35

[lcxu105]

Found it thanks.