Looks Like Garrett Slipped Another Gtx30 Variant In .

[Lithium]

I'd suggested early on from the advertised performance map that this 67mm compressor needed higher rotational speeds to do its thing

Leads me to conclude that advice I received a few years back was correct. Turbine choice/spec has the biggest bearing on turbocharger response and engine response. The compressor is the dumb device being driven by the turbine, and if it doesn't spin fast enough, early enough then it won't do the business.

I'd have some concerns whether the 0.63 IW housing is the best choice from a boost control standpoint, but if the targeted boost is going to be 18+psi that might be academic only.

You are definitely not the only one who mentioned that the thing is going to need higher rotational speeds to do it's thing, and there were a few of us who mentioned the .63 as being a good idea (or more specifically the .82 possibly being a bad thing) for this combination.

I don't think it has ever been a debated thing that turbine choice/spec has the biggest overall influence on the spool/response - however on the same note that doesn't mean it is 100% responsible, it's not. The whole thing relies on a combination of things, not just the turbo, and not just the turbine wheel/housing. There are all kinds of GT30 based combos out there, and they do not all follow the same boost curves - we have people with GT3071Rs, GT3076Rs (52 and 56 trim), GT3082Rs and GTX and HTA equivalents all with .82a/r GT30 turbines providing quite different response and spool figures.

You guys (Dale/Wolverine) seem to be implying that the GTX3067R is redundant as it has all the disadvantage and none of the advantage of a GTX3076R. Am I correct? And if so, any theory on why Garrett would have done something like that? And if not, what are your thoughts on how to improve this situation?

[wolverine]

I would argue that this turbo is well within hiflow territory so it is probably redundant (on cost/benefit but not useless or hopeless) before it starts.

To me the compressor isn't flowing significantly that early like the other more slippery GTX compressors so a smaller housing would give a bigger shove down low. Nothing earth shattering in saying that. The peak power everyone was expecting was within reach of the 0.63 so it makes sense this might have been the best housing to try.

Why Garrett chose this compressor has a lot to do with someone wanting 450hp, 500hp, 550hp (or whatever steps each compressor offers) after all they all run the same turbine wheel.

What none of us know is how it drives, dyno prints aren't worth a lick of spit for telling you how much better this thing runs on the road for rolling response hence I would just drive the thing and enjoy it.

I would certainly check for boost leaks because they are highly parasitic.

If I was made of money I would consider buying a 0.63 housing and a retune but given the time taken to get it to this point there would be no urgency on my part unless the workshop was going to make up the cost of selecting a less suitable actuator and seemingly putting work on the car on the backburner.

Ultimately all our conjecture is sometimes a bit of a downer for the OP and we can lose sight of the fact that we will never be 100% happy with a setup because someone else will tell you they can always make more power, come on earlier or do something better.....drive it enjoy it and see how you feel in 6mths.

[M@&k]

Once I compared the results with GT3071 of similar setup I felt much better about the whole thing. Seemed to be in line with what you would expect.

The problem is I was always after a turbo with better response than a GT3071. I had in mind something that could produce full boost, say 18 psi, before 3,500 rpm. When picking the GTX3067 I was of the understanding that the 0.63 housing would be used. When I found out it was being put together with the 0.82 I started to think maybe the trade off for more power might be worth it. Plus it was getting to the stage that I didn't want to cause any more delays by pushing for a switch.

For me, it's not so much that it's not performing, more that it's not what I originally set out to achieve. Having said that I will follow Wolverine's advice and just drive and enjoy it for the moment and see if my sentiment changes in a few months.

On a different note; went for a drive yesterday and Profec controller readout told me I hit peak boost of 144 kPa. The Interwebs tells me this is about 21 psi. Now I would guess the Profec is not the most accurate boost gauge but seems quite different from the 16 or so psi it was apparently tuned at. Should I be worried?

Edited July 9, 2014 by M@&k

[discopotato03]

In my previous post I was disagreeing with the black art bit .

To be fair it was always going to be a choice of 0.63 or 0.82 turbine housing on this turbo , and the majority of GT30 turbos .

This turbo was always going to be a GT28 based one that grew an optional GT30 turbine so there's something in that . Most know that there doesn't appear to be much flow difference (if you can believe the turbine maps) between a 0.86 GT28 and a 0.63 GT30 hot side . Now the two litre four cylinder people worked out ages ago that the GT28 0.86 housing wasn't much of an upgrade on their GT28XXR turbos , I think a higher boost threshold and nothing earth shattering up high .

I have read a few accounts of people going from 2867s to 3067s and one about a feller going from a GT3071 to a GTX3067 , they all used the 0.63 AR turbine housing . The feeling was that the 3067 had just a tad more lag than a 2867 and then better off everywhere , the GT3067 was said to spool earlier and be as good or better everywhere than the GT3071R .

I don't think Garrett marketed the GTX3067R very well which is silly because the only variables with them are turbine housings and actuators . The 2L four cylinder people were obviously on the right track with the 0.63 T housing and it's a flow improvement over the usual 0.64 AR GT28 hot side , I wouldn't have gone past a 0.63 housing with this turbo on a 2L four either .

The playing field changes with 2.5 litres and six cylinders because these makes more engine only torque and need more air to fill their lungs and make positive pressure . With no info to go on you can debate either way which T housing size should be better but in the end only trying both tells the story . As I said based on just a few accounts from people with 2L fours the 0.63 seems to work pretty good .

All I can suggest is do the leak tests and then strongly consider a higher rated waste gate actuator , I opted for a 1 Bar big can version and had a pretty good idea that my 0.82 housing wasn't going to give early rapid boost rise so better boost control loaded up .

If leak fixes plus the actuator don't help and the tune and exhaust aren't letting you down then it's pointing towards the turbine housing not giving high enough gas speed through the turbine to spin the compressor fast enough early enough . The common ground with positive GTX3067R results seems to be the 0.63 housing , someone was always going to go larger to see what would happen and I'm genuinely sorry if I gave you reason to think it'd be the go .

With a bit of fault finding and possibly a 63 housing your results can only get better IMO .

Intercooler wise I made 271 wheel wasps with a smaller GT28 turbo admittedly on E70 but if the Blitz return flow had been a major restriction ethanol wouldn't have made the situation any better .

A .

[Lithium]

Agreed. The result isn't exactly bad, its not a lag monster and it is reasonable power.... but I also feel that there is room for improvement, which hopefully means there is an encouraging outlook for m@&k where he can hopefully pick up some response AND power if he starts feeling the urge for a bit more of both, going a more aggressive actuator and a smaller a/r turbine housing.

The unfortunate thing is none of us at this stage can guarantee an outcome, we are all in the same boat - it's all hypothetical, unfortunately. I hope you enjoy the car, and I hope we're right that there is more to be had when or if you decide you need more.

[M@&k]

I don't regret any of the advice I got on here at all. I knew full well this was uncharted territory, which was part of the appeal, and it's hardly a disaster just slightly off what I was chasing.

Just the question of how different it might have been with the 0.63 housing is a compelling one...

Damn you people and your encouraging comments

[Lithium]

Oh - in terms of the boost spiking you referred to earlier, I'm not sure whether it's safe or not - not necessarily ideal, though. How unideal it is depends largely on the tune. What rpm does it happen, do you know? My old GT30R setup I had a crap boost control setup and ECU so I just left an ultra soft tune in it and put up with boost spikes as it felt better than a really soft boost curve, after it was all said and done. I never hurt my motor, or turbo - so I guess it depends on how steadily it is holding that kind of boost. I actually don't expect you'd be able to hold that boost up to peak power so again... depends on the tune or how much it bothers you.

My view on the world in terms of this kind of stuff has definitely been spoilt by being exposed to cars running G4Link ECUs + boost control, makes it all a lot more usable.

[juggernaut1]

Once I compared the results with GT3071 of similar setup I felt much better about the whole thing. Seemed to be in line with what you would expect.

The problem is I was always after a turbo with better response than a GT3071. I had in mind something that could produce full boost, say 18 psi, before 3,500 rpm.

Here's a dyno of a GT3071 with a .63 hsg on an SR with small Unigroup cams (pretty sure it is BHdaves setup). Full boost arrives at 3,600rpm and made 285rwkw. It would seem that the GT3071 would have fitted your requirements. Nevertheless, I would change the rear hsg on your setup to a .63 for a better match to the compressor and desired use of your car. Perhaps a different tuner may also be able to fully exploit the potential of your setup.

[M@&k]

I thought BHdave ran the 0.63 GT3071 on an RB? I remember reading the thread where he was struggling with boost control. ^^ That does seem to be a nice result.

If I do go for the change I reckon I'll look at going to someone who can do both the mechanical work and tuning. In hindsight it was definitely a disadvantage to not be in direct contact with the tuner about the result and get their feedback. Live and learn.

Edit: Also, I don't want to knock Galvsport at all. I was very happy with the quality of their work, just that I think they're probably suited to much bigger projects.

Edited July 9, 2014 by M@&k

[juggernaut1]

He must have also run one on an SR......it was a 3071 .63 with external gate and Unigroup cams.

Posts 83 & 85 (where Dave chimes in).

http://www.nissansilvia.com/forums/index.php?showtopic=413229&st=60

Edited July 9, 2014 by juggernaut1

[Dale FZ1]

Lithium I'm pretty much on the same page as Wolverine. I felt the expectations of a powerful response monster were unrealistically elevated, given what the other GT30 combinations were/weren't able to produce.

Gut feeling on why Honeywell would release this spec for sale? Marketing and being seen to cover bases without a significant outlay since the compressor components had been developed. They are in the business of making money from selling turbochargers. Results are far from poor and there's no way this 3067 could/should be judged from a dyno sheet. It's got to be driven, and the best measure would be to find a mate with a Skyline running a different spec GT30 and try them back to back. Outpaced by some current offerings but IMO not outdated, old faithful GT3076 may not set the standard but direct comparison with it would shine a light on transient response characteristics which is what the 67 should be about. Lots of positives about the HTA units, maybe that's the spec that sets the bar and I'd personally like a run in one of those myself.

If Adrian's information about housing options is correct, it pushes thought on Honeywell's marketing strategy to the GT30 range as being a performance item that is designed to be controlled via external gates. That step neatly shifts responsibility for boost control systems to the aftermarket and choices that individual owners make.

How to improve this one? The consensus was indeed 0.63 housing on this unit, but I wouldn't go with one of those in IW form. Unconvinced that they offer the sort of efficiency/boost control capability that's needed. It's only my view, but if the choice had to be 3067, it would be with a 0.63 running a large external gate. Perhaps run the gate off the housing for cost control, but I'd rather see a fabricated manifold and then the changes and costs become fairly high for a 350rwhp install. Starts to push things away from the low-key visuals of this install which will be viewed by many as a priority to avoid road transport compliance issues.

Never an easy or simple answer to cover bases.

[Dale FZ1]

It's been a while since Dave's Skyline went into the wall and was replaced with an S13. He was very disappointed with boost control capability of the 0.63 IW setup, really did not stack up for him despite efforts to resolve it.

[discopotato03]

Personal opinion based on the force is that these GT30/35 IW housings are expensive to make and make Garretts GT BB IW turbos expensinve/hard to market . Note lots of suppliers quoting turbo price minus turbine housing . Maybe the huge American market thinks they'll get away with ext gates on road cars ...

It's been a while since people put much thought into Garrett's older GT3071R and I always though a lot of its problem was the plain non shrouded comp housing . Some boost and surge related problems are avoided with port shrouding . Think for a minute about say a GT3076R/GT3037S with a HKS 0.61AR turbine housing on it , they didn't tend to have surge or boost control issues - the power delivery was light switch .

Anyway for whatever reason 3071Rs were never a fantastic unit , never seemed to match the 2835 Pro S which used the same compressor and housing size - plus port shrouding . I think all GTX turbos have port shrouded compressor housings and HKS went to the trouble and expense of having their own made if Garrett didn't have anything suitable . No business does that without a justifiable reason .

Also something I think people need to think more about with turbine hosing size is why there even is a range or choice . Its not just about how much exhaust flow you can get through the poor old turbine and what the absolute power number will be just before the rods go into orbit . Its about having the compressor spinning fast enough to make power where you want it in your engines rev range . Turbochargers are not perfect things , a radial inflow turbine at one end and a rotary air pump at the other . Neither turbine or compressor performance is linear and making a hopefully small system restriction in your engines exhaust creates other issues too . There is no perfect hair dryer .

I don't like going out on a limb because opinions are subjective but ...

Obviously with this GTX3067R you have to spin its compressor fast enough to shove enough air down 25s throat to make positive pressure . You have to have enough exhaust gas energy to spin Mr GT30 fast enough to make the compressor boost , and aside from the throttle your foots attached to the only other means of dialling in exhaust energy to boost threshold ask is T housing AR options .

Now something else to think about is that were are having past opinions on GT compressor antics and basing some opinions on what we predict GTX ones will do .

Most realise that the Xs pump lots of air and have better high pressure performance than the GTs do - BUT you have to spin them fast enough to get the best performance out of them .

Uncharted waters . We know some people had issues with GT3071Rs GT3076Rs and 0.63AR GT30 IW turbine housings . The 76R has port shrouding where the 71R doesn't and the 76s don't surge AFAIK in this form . You probably get a lower detonation threshold on petrol but boost issues I don't know about esp with the later larger waste gate flap .

To me it's pretty obvious that you have to spin lil X67 a fair bit faster than 76mm 56T GT to get 500 odd Hp's worth of air out the compressor housing and at 0.82AR T housing size that's probably a big ask with an 84 trim GT30 turbine . Especially if you want it to do something on a 2500cc engine at a bit over 3000 revs . I reckon if you aren't going to spin GTX67 fast enough to run mid 20s boost levels then you probably aren't gonna get 500 Hp or need 500 horse powers worth of exhaust flow capacity . So , with the mass exhaust flow down a bit the only way to get the turbine response up is to increase the velocity of the available exhaust gas by using a smaller volute passage .

Real world I can only repeat what I've read from others using these turbos , response is good boost control seems good and they don't get compressor surge . This is with a 0.63 AR IW turbine housing .

What happens with GT3071Rs GT3076Rs etc is irrelevant because the GTX67 compressor and T04B compressor housing are different animals , apples to pineapples . Different sizes mass and speed ranges/speed matches .

A .

Edited July 10, 2014 by discopotato03

[Lithium]

Thanks - and nice DP, I am way too busy to get too into this at the moment but you covered a couple of things I was hoping to.

Trying to work out where Dale FZ1 stands here, he (and Wolverine) seems to be saying stuff or implying he thinks we're off the target thinking this turbo is potentially a good thing but he just liked your post - Dale?

[M@&k]

I feel like all four of you are saying the same thing but coming at it from different ends. Disco + Lith from compressor side saying the GTX series (and possibly this one in particular) needs high rpm to make boost so would benefit from 0.63 housing.

And Dale (+ possibly Wolverine) are saying the response is primarily driven by the turbine with compressor having minor impact, so needs 0.63 housing to change performance.

What I'm starting to wonder is, if you go smaller housing to get compressor rpm up earlier, do you defeat the purpose by adding restriction to engine flow so it still can't do it's thing. Kind of leaving you in no-mans land...?

[Dale FZ1]

Lithium I'd think this 3067 is potentially a half-good thing and I'll try to explain the thinking.

I liked Adrian's post because it shows logic in the issues of shaft speed for this and the other GTX series, and also acknowledgement that their primary benefit over the older GT series is flow efficiency at higher pressure ratios. But there are also some aspects of his post that are a touch wide of the mark too, so a like just means "yep I get what you say", even if we're not totally in agreement.

Andrew's view that the 3067 is within the realms of some high flows and other offerings is IMO accurate, principally because this is not a high hp unit. Adrian's comment suggesting 500hp capability for that 67mm compressor unit is IMO inaccurate or misinformed. The performance map is saying it's pretty much tapped around the 43lb/min, so it's going to (has to) fall short of the GTX71 and GT76 units for max hp. I'd say it "generally" falls within similar flow capacity as the GT3071 0.5 a/r compressor unit, and the GT2871 52T 0.6 a/r compressor unit. I'm guessing a spread of 40hp would cover them. They have differences in flow efficiency at different pressure ratios, and need different rotational speeds to achieve their best. And they will weigh different amounts, and push air through differently configured housings, so they will have different spool/transient response capability for a given turbine combination. So not exactly the same, but they are rubbing shoulders in a small room.

And this means that they will each have turbine combinations that work best for them. Trouble is that's basically limited to housing changes, and internal/external gate setups with their inherent compromises whether that be cost, physical packaging or boost control.

Adrian's comment that all turbine/compressor combinations are compromised is 100% accurate, and I'll suggest his recent move from the GTRS to GT3076 setup on his R33 has strongly reinforced that point. (bet he's happy he made the change though... ) And his regularly repeated comments urging for different spec options on the GT30 rotor aren't at all bad, except Honeywell haven't and probably won't released anything like that.

I do think this 3067 would give its best if excited early into higher speeds. The few owners who ran the 0.63 I/W turbine combination reported boost control attributes I don't think are optimal, and although that is the most easily fitted from a cost efficiency and packaging standpoint, I wouldn't run with or recommend one of those housings.

Where IMO it would shine is by specifying a package that is able to control boost to whatever level is wanted, and avoid spiralling turbine inlet pressures (aka back pressure) because of a sub-optimal wastegate flow-path forcing most exhaust gas to exit via the turbine rather than bypassing. That IMO means using a good design/combination of fabricated manifold and decent size wastegate. And a decent design of post-turbine merge for the wastegate dump. Screamers just aren't acceptable, especially when we are expecting a likely max hp in the 250rwkW range.

I'm seeing costs mounting for a relatively narrow focus combination that may be reasonably matched by the highflow alternatives retaining IW setup and close to stock appearances.

So yes, I agree to a point that this thing is potentially a good thing but the cost and effort probably outweigh the benefits when you see how Mick's GTX3071 spanked most setups when teamed up with E85. Given his results - and I was fortunate enough to pedal that car so I speak from experience - I couldn't urge someone to go down the path of manifolding and EW instead of retaining stock manifolding and budget for fuel system changes.

There is plenty of room for variety and opinion. Mine is not necessarily the best view but I'm sharing it and open to alternatives since we are always learning. I appreciate Mark's had a good go and he still got a good result.

[Dale FZ1]

Phew sorry for the long post.

[AngryRB]

I have a perfectly good 0.63 internal gate housing sitting here if you want to swap, you could have it done over the weekend.. I cant use it for my RB30 build ,

Even a GT3076 can be limited with a 0.82, mine was knock limited at 280kw so not all engines will make what they even should with a said combo, Mafia has proved the 0.63 can do 320kw with water/meth

and Matt showed 360 odd kw with the 0.82 so its kind of a no brainer really that the 0.82 is well beyond the GTX3067's capability, so 0.63 and some water/meth would see maximum response and

results.

It would probably jump out of its skin with a 0.63 and water/meth, and a tuner that will put more time into perfecting it.

Edited July 10, 2014 by AngryRB

[wolverine]

Lith, I thought we were all on the same page more or less.

A 0.63 with the right wastegate flap and actuator should be fine or am I missing something? Ext gate would be better but surely boost control has been sorted on these things.

[M@&k]

If you need to go to the effort of new exhaust manifold and external gate, then I don't think it's worth the effort as you have pointed out Dale. Received a follow up from Galvsport yesterday to see how it was all going, and I asked what they thought about potential switch to 0.63 housing. Will see what they come back with.

AngryRB the possibility of a swap is looking more promising, but probably wouldn't happen for a little while. I'll send you PM to discuss.

Another question for you gents. If this turbo is only ever likely to do 250-260kW, is there any benefit of going the GT30 turbine over GT28? I decided to go this way based on understanding that with the GT30 hotside it should be more efficient and make more power (like Pro S vs GT-RS although I appreciate Pro S also had larger trim compressor).

I've noticed KudosMotorsports offering kit with GTX2867 with RB25 dump pattern, studs, HKS outlet elbow etc for ~$2,300. That's only $100 more than I paid for just the GTX3067 & turbine housing. GTX28 version seems more compelling value if it can do pretty much the same power with better response, for cheaper.

Thoughts?