Looks Like Garrett Slipped Another Gtx30 Variant In .

[M@&k]

Not necessarily. While I thought it would make more power, it's always been the response that I've been mainly disappointed in.

From the tweaking done so far, and in particular the last work (post #326) which didn't give any gain, I get the impression the compressor is basically maxed out for flow. If that's the case then I would think (or hope) the turbine housing isn't the limiting factor, and might be able to lower the boost threshold with the 0.63 housing without giving up much top end.

[discopotato03]

If you're getting 250 at the wheels then I wouldn't think you are getting all that the GTX67mm compressor can give . Assuming there are no other issues such as tuning or tuning around detonation for some obscure reason I think you can do better .

I'd have to search around again but I think you are the only one that ever used the 0.82AR housing on that turbo , most went with the 0.63AR version or some T3 nock off type housing .

It really sounds to me like your turbine isn't spinning its compressor fast enough to make it work properly . Maybe a problem is that the 67s unique 10 blade count needs more revs to make it pump enough volume compared to most GTX wheels with 11 blades . I have to wonder if Garrett was trying hard to get the best balance of compressor vs turbine performance in the GTX2867R and you'd have to assume they were struggling with hot side flow in a 0.64AR GT28 based turbocharger . Generally less compressor blades means more revs to pump the same volume and higher turbine speed means less resistance to exhaust flow up to a point .

It may not sound like a big change shoving a GT30 turbine up GTX67s ring but is a major change and one that I reckon needs consideration to make sure it spins 67 fast enough to do the job . The only GT30 turbine housing that was ever remotely going to make a 30 turbine perform like a 28 turbine in a 0.64 GT 28 housing is the 0.63 one .

Really turbochargers have to be looked at as a system not just a folk law sized turbine housing with a hair dryer clamped on one side .

I also doubt your intercooler is a major restriction but if you want to see then try measuring boost pressure before and after the core itself .

A .

[M@&k]

I'm not sure any more. The compressor map is pretty close to the GT-RS, and most people get around 240-250rwkW, a bit more with head work or cams. I'm not clear on what benefit you actually get from having a bigger turbine...

On the turbo speed topic, I don't know enough about how it works really. Are you suggesting that at say full power, it can't get enough energy out of the exhaust gases with the 0.82 housing to spin fast enough to generate the required boost? If this were the case it would mean that the wastegate would actually be closing as revs increase right?

[scotty nm35]

Good chance the wastegate can't close, even though it is trying.

Shame you didn't shove an external gate off the housing, I think this turbo had some potential.

[discopotato03]

What I am saying is that the compressor has to spin fast enough early enough in the engines rev range to get a pressure head (boost) on the inlet side .

I look at an accelerating engine as an expanding volume from the compressors perspective and you may have a situation where the turbo is playing catch up and not getting any worthwhile pressure up until mid range revs . Then up high it falls off capacity wise .

The way to increase the compressors speed in relation to engine revs , and load , is to put more energy into the turbine and that usually means a smaller turbine housing .

I think you really need to search for other peoples experiences with that GTX3067R and don't be too concerned if most of them involve 2ish litre four cylinder engines . These engines usually make less torque at the same revs than an RB25 and with your turbo in the 0.63 housing they are making good all round power .

I think its absolutely wrong to make the global assumption that the GT30 turbine 0.63AR turbine housing combination is an epic fail in any and every instance . Granted some GT30s have higher capacity compressors than others at high flow rates small turbine housings will become the limiting factor turbocharger wise .

This time round you have to look at turbine speed and exhaust flow rates with the smallest lowest maxing compressor option in the GT30R family to keep things in perspective .

Really this turbo should be viewed as a modified GTX2867R with a bit more hot side potential (flow) than the GT28 turbine/0.64AR GT 28 turbine housing can give .

If you research GT28 turbos with the larger (for GT28) turbine housing ie 0.86AR the result wasn't always good . I'd say Garrett went looking for the extra flow with a different larger turbine and it worked with what is considered to be the smallest of the GT 30 housings . I think the best way to look at it is that the 30/0.63 is larger and higher flowing than the 28/0.64 and that's why the 3067 version works better in your instance than a 2867 would .

Out of time but I think 0.63IW is the answer .

A .

[discopotato03]

My last post must've got lost .

The guts of it was that this turbo is a hybrid of the GT2867R and I reckon most of its dialling in would have been based on GT28 hot side development . That means a 53.8mm GT28 turbine in a 0.64AR GT28 T2 flanged turbine housing .

We already know that older GT series compressors ie the 71mm ones easily take that hot side combination to the limit and beyond . It was known decades ago that the cropped GT30 turbines in HKS turbine housings out performed the GT28 based versions - search HKS 2835 Pro S results to see the performance difference .

I think what a lot of people are missing here is that the GT30/0.63AR hot side combination is clearly better than the GT28/0.64AR one which is why you'd expect a GTX2867R to perform better in GT30/0.63AR format . I also think it doesn't matter a rats arse what happens with GT3076Rs/GTX3071Rs/GTX3076Rs because they are significantly different to a GTX3067R and the sort of potential you can screw out of post 300 RWKW turbochargers . I wouldn't expect to see 300 plus wheel wasps with a GTX3067R on an RB25DET even with E85 and for that reason I don't think its mandatory to have a 0.82 housing on it - because its unlikely to ever see the exhaust throughput that a 300 plus RWKW Skyline would have .

I think of all the things you could potentially do the 0.63IW turbine housing stands the best chance of getting you what you want . All else being equal it will bolt in and increase the gas speed through the turbine blades for any given engine speed and load . Assuming you have a gas tight system beyond the turbocharger and no glaring exhaust restrictions down stream of it there is no reason to expect to see less than about 270-280 RWKWs IMO on a healthy factory RB25DET .

I got to 271 wheel wasps with a HKS GT2871 52T in their unique T3 flanged 0.64AR GT28 turbine housing and obviously the only reason it didn't detonate to death was the E70 it was burning . Your turbo in 0.63 form is an upgrade on a HKS GTRS in every way meaning a more efficient compressor end and a higher flowing hot side , these alone should get you at least as much as I had IMO without needing 70% ethanol content fuel . I'm sure I remember people with HKS2835 Pro S's getting into this area without ethanol .

A .

[M@&k]

Thanks for your thoughts Disco. Hopefully I can get my hands on a 0.63 housing. It's what I'd wanted in the first place.

[M@&k]

Ok, bit of an FYI. I've recently got my hands on a 0.63 A/R IWG housing. Car is booked in for install and tune on 11th December.

Note this will be on the same dyno as before, just bolt in-out and retune, so should provide an interesting comparison if nothing else. Will post dyno sheet in a few weeks.

[Mick_o]

Good luck Mark hopefully this sorts your setup out and you can start enjoying it!

[M@&k]

Thanks Mick. Don't worry, I'm already enjoying it. this is more about optimisation.

[Dale FZ1]

Fingers crossed, this will produce the results you've wanted from the start.

3 weeks to workshop time and tune won't take long to pass. Will it??

[M@&k]

I dont expect so. They did the dump pipe, actuator and tune pretty quick last time. One day I think from memory.

[Lithium]

Good luck, man - hope it goes a bit more to target this time. You have a decent stiff actuator etc eh?

[M@&k]

Yep, actuator was changed to 18psi. Didn't change much at the time though.

[M@&k]

So, here we go. 0.63 housing in and retune. Boost on 500 rpm earlier, and peak power only dropped something like 8 hp from previous but is reached 1,000 rpm earlier.

Blue line 0.82 housing, Red 0.63.

Haven't had a good drive yet, but noticeably different, and much more like what I was after

Edit: haven't worked out how to embed images properly...

Edited December 11, 2014 by M@&k

[Dale FZ1]

Good news, hopefully it's putting out enough to make this spec turbo worthwhile as an upgrade option. Does it feel good to drive in terms of transient response and delivery? Be good to see the torque curve but probably fair to say it's a mid-rangey thing doing its best between 3000 and 5500?

[Lithium]

So, here we go. 0.63 housing in and retune. Boost on 500 rpm earlier, and peak power only dropped something like 8 hp from previous but is reached 1,000 rpm earlier.

Haven't had a good drive yet, but noticeably different, and much more like what I was after

Edit: haven't worked out how to embed images properly...

Nice work - that is a LOT more like what I was looking for, nearly 50kw up at 3500rpm and gets to closer to 100kw over before the .82 caught up... it should be quite a lot better to drive, especially in terms of "coming alive". The power falls over in a strange kind of way, though... I would expect it to struggle to hold like the .82 but that seems a bit worse than I expected... did the tuner mention anything? It seems almost like a blanket pile of timing has been pulled out above 5500rpm, I'd not be surprised if it was knock limited but not sure about that much everywhere. Nonetheless not

Let us know how it goes when you take it for a proper blat

[The Mafia]

So, here we go. 0.63 housing in and retune. Boost on 500 rpm earlier, and peak power only dropped something like 8 hp from previous but is reached 1,000 rpm earlier.

Blue line 0.82 housing, Red 0.63.

Hyperdrive_2.jpg

Haven't had a good drive yet, but noticeably different, and much more like what I was after

Edit: haven't worked out how to embed images properly...

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

[zebra]

Its a 3067

not 3076

[The Mafia]

Its a 3067

not 3076

okay thanks