Gt2871 Based Hi-flow Into Rb25

[RS500]

Dragging up the original thread, it's been a while as I learned about tuning through the Datalogit + wide band input.

I'll let the graphs largely tell the story, but add the following:

boost control with the $30 boost bleeder is not super smooth - fluctuations are evident, and there is a gradual drop away at the top end. I will look at changing boost control in the next couple of months, as a means to improving that torque dip @ 4000rpm

The torque dip is also going to be attacked by trying an adjustable exhaust cam gear. I fitted it just prior to the dyno run but there is no way I could get in to adjust things while it was all hot. Happy to do that another day.

AFR work was all done prior to dyno runs, with the only changes made being ignition timing. The richness at the very top end was intentional, to promote longevity. My target above 5600 was 11.8, so there will be a little work there yet to be done. Might pick up a couple of hp there, and with some effort I hope to see the power curve remain flat rather than nose over.

Finally, my maths proved to be pretty good. I had actually calculated 290hp @ 14.5psi @ 6100rpm @ 12:1 AFR, using Sydneykid's observations of typical 2wd drivetrain losses. Further checks of my logged injector duty cycles suggested somewhere between 285-295hp, again using the rated flow of injectors and the power levels that would support. Overall satisfaction is still high

Before/after shows the results of 9.5psi boost vs 13psi + some tuning.

Dale,

they are awesome results you have there, well done.

looks like all that R&D really paid off giving you a nice responsive & efficient combination.

it should feel even nicer when you get rid of that mid-range dip and raise the boost to 15-16psi.

this looks like the ideal combo for my R32 RB20DET, although i wonder how much improvement there would be by using the HKS T3 exhaust housing from the GT2530 instead of the RB25 housing.

keep up the good work and well done once again.

[discopotato03]

Lithium I don't seem to remember that but anyway .

I say my latest research because new turbine housings are in the wings overseas , T3 dual entry turbine housings that take 60 - 68mm turbines and in high temp materials ie Inconnel are not yet common . Wastegating is also an issue because unless you isolate the sides of a divided turbine housing and manifold much of the pulse separation can be lost . Going to external gates is generally required and the best results seem to come from two gates , one for each side .

Occasionally you do see a properly made divided turbine housing with integral gates (two parallel flat valves to two separate vent paths) but not common . From memory Toyata did it with one of their CT turbos (CT20 ?) for one variant of the GT4 Celica and Mazda may have as well with one of the 13B turbos ie S4 RX7 ? I think Garrett have also done it to some of the GT3576 bush bearing truck turbos .

From what I read turbine housing A/R size is absolutely critical with divided housings because its very easy to have more restriction not less . People testing these things are saying that divide the volute passage and halve the area of each side . They are going larger than they would for a single passage housing and comparing the turbine inlet to compressor outlet pressure to try to get an ideal ratio . Reading between the lines I think they are trying to maintain the same expansion ratio while dividing the exhaust pulses and cutting out reversion . I think the effort is to pull up the bottom end torque via exciting the turbine earlier and reducing pumping losses at lowish revs .

Anyhow I have seen pics of the real GT3071R with the T3 flanged split housing that ATP Turbo sells , like me others overseas suspect its a Garrett GT3267/GT3271 style turbine housing thats been reprofiled to suit the BB GT30 turbine/CHRA . You can see these turbos at turbobygarrett under GT3267 and GT3271 . Be wary of the wastegated ones because some diesel housings only vent one side of the housing not both .

More as comes available , cheers A .

[Lithium]

You actually wondered why the dude bothered with a twin scroll on a 4G63 as they spool GT30-based turbos well anyway - after looking back on it, so I guess not necessarily knocking the setup persay...

[discopotato03]

Mitsy 4G63 Turbos have a slightly different set of circumstances mainly due to their std reverse rotation turbocharger . To fit a Garrett GT BB turbo means either one of several fabricated exhaust manifolds or ATP's cast one which is a bit of an awkward design to package a turbocharger that turns in the opposite direction to the std one . Actually I'd have to go and have a look because the ATP gadget may have been an adapter for a Garrett turbo and an external gate . The propper solution is a fabricated manifold with the right type of collector . If I was looking into that app I go to Full Race Motorsport and talk to Goeff Raicer . He's had a fair bit to do with Honda and Mitsi engines and is now working with RB25's and 26's as well .

Cheers .

Late Edit : Its actually an ATP exhaust housing not a manifold or adapter , bolts to std exhaust manifold .

Edited February 27, 2007 by discopotato03

[Dale FZ1]

Off topic gents...

[Dale FZ1]

it should feel even nicer when you get rid of that mid-range dip and raise the boost to 15-16psi.

this looks like the ideal combo for my R32 RB20DET, although i wonder how much improvement there would be by using the HKS T3 exhaust housing from the GT2530 instead of the RB25 housing.

keep up the good work and well done once again.

Thanks Gorjan. I'm likely to run with a replacement actuator rather than an EBC. Obviously each has its own advantages/disadvantages but I feel the actuator route will work well for me in this application. Choices are of either a 14 or 17psi setting so I will review the maths and maps to see if there is any real gain (without compromising reliability) to be had from pushing the extra boost. Gut feel is that it will nicely handle 14psi and crank some more power but that will be about it. It's about the size/capacity of the compressor impeller and I don't see the sense in over-speeding the rotating assembly.

If those HKS housings were available economically I would be prepared to try one, but this really was intended as an off the shelf style high flow with cost containment being an important aspect. Similar issue with why/why not to use the RB25 comp housing as per an earlier post in this thread. Overall it was more expensive than a plain-bearing unit like Slide's (no, I don't want to get into comparisons), but it is showing itself as effective.

I'd suggest this spec should mate well with an RB20, depending on performance targets.

Meanwhile I'll work on smoother boost control and the cam timing to see just how good a result I can get.

cheers

Dale

Edited February 27, 2007 by Dale FZ1

[benl1981]

Hi Dale,

Reading this thread with interest.

I'm aiming for about 220rwkw like you had I think (295rwhp) - but with a GCG turbo. I see you kept the stock AFM..did you also have the stock injectors and stock FPR but upgraded fuel pump?

What sort of inj duty did you have at 180rwkw? and what max AFM voltage (e.g. 5050mV?)

And what inj duty do you see at 220rwkw?

I have 172rwkw and see around 70-75% max inj duty. I have seen up to 5.1v but mostly around 5040mV

Thanks for the help with this. Is 295rwhp the furthest you have taken it? Is it very quick?

Cheers

[Dale FZ1]

Still very much a work in progress. Yes, it's fairly brisk but not mind-blowing. No idea or interest in seeing what it will do at a drag strip, but it covers ground quickly without feeling overwhelming. I aimed for smooth progressive power, which it has. Still running the stock clutch, with no signs of slippage. A larger trim unit would probably be more sudden in its delivery and harder on the clutch.

I've acquired a 14psi actuator, which will be fitted tomorrow to cure the 2psi oscillating drop from 14-12psi that was logged on the dyno runs. Between better/more accurate boost control, slightly less rich top end fuel, and playing with the exhaust cam timing I expect to see a slight power lift. I will update the thread after dyno later this week or early next.

Here's what my logged duty cycles look like:

That is using the S15 injectors. Stock fuel pump could not keep up the flow and AFR started to lean out at the higher rpm.

The stock AFM really is marginal, and clearly hits max voltage output @ 5115mv. With extensive data logging of AFR and tight fuel control I think I can get away without it at this power level.

Edited March 5, 2007 by Dale FZ1

[benl1981]

Thanks Dale. Is N16 about 6000rpm?

I'm surprised your AFM doesn't max out a lot earlier. Mine with the stock turbo at 10psi hits about 5.05V at about 5500rpm. That makes me question whether I might have problems..

Good point about the clutch. I think I might only go to about 220rwkw as I dont really want to have to do my clutch aswell. Will cost a lot more to get to 250rwkw..

Sounds like you have a good spread of useable power

The bloke that has 226rwkw seems to max his out at 4000rpm here..in the hpi article

http://img223.imageshack.us/img223/195/z32p1dv6.jpg

http://img220.imageshack.us/img220/5568/z32p2pp8.jpg

http://img301.imageshack.us/img301/7972/z32p3np2.jpg

http://img300.imageshack.us/img300/7904/z32p4lr2.jpg

Edited March 5, 2007 by benl1981

[Dale FZ1]

N15 = 6000rpm. The scale is in 400rpm increments.

I didn't know what to expect with the AFM until I checked the logs. From memory when running stock turbo + boost I hit max voltage @ >6000rpm with 223 rwhp.

Torque delivery (ie sudden sign-in) and quantity combined will determine what the clutch can withstand. I've been lucky to date, but expected to have to budget for a replacement.

It is definitely useable and accessible power, though the nose-over beyond 6000rpm may not be to everyone's liking. Doesn't bother me - there is plenty of torque easily on tap when you pick another gear.

More in a few days.

[Dale FZ1]

Been a little while getting back with this one, but I had to play around with the cam gear setting to address what seemed to be a mid range torque dip.

There was conflicting information from various sources on the board, from those who said that advancing the exhaust cam would boost low-mid range, while others said that retarding it would give a boost all round.

Basically the theory on naturally aspirated engines suggests retarded exhaust cam will boost top end at a cost to the bottom end, while advanced will do the opposite.

I accepted that once the aspiration is forced, the rules of what works changes to some extent.

Those who advised to advance were doing it from a base of experience in chasing the same outcomes as I was, while I think the proponents of retarding were looking for more top end improvements.

Knowing that things are just so damn hot once it's been running, I knew there was only going to be one setting I could try per session on the dyno without wasting time and $$ to prove what would work. In the end I went with 4 degrees retarded, and give it a run. On the road any difference seemed difficult to detect.

Well here are the results of a retarded exhaust cam:

The curve has become much more civilised and smooth, rounding off the edges being shown. All in all a "nicer" curve showing useable power. I'd say the cam timing change produced a worthwhile improvement even though the peak number is lower.

What was causing some intrigue was the tendency of the curve to peak around 5500 and then drop after 6000. Suspecting top end breathing restriction of some type, we decided to remove the bolt-in restrictor in the Blitz cannon muffler, giving 75mm then 80mm from the split dump back. No changes other than the restrictor.

Here's what the dyno showed:

Not bad, and the note is very fruity. Good to listen to, but man talk about drone when running at low load / off boost. The lift in power was a big lesson in just how stock systems cost power in the name of noise restriction. The slight dip right at the top end is ignition related, as I had pulled a couple of degrees out to sort the odd reported knock event. Otherwise, it had more (and more) from 3000rpm onwards.

At this stage I am now waiting on the 14psi actuator cannister after a supply problem from Garrett. I think the oscillating boost control is definitely costing some power because it drops from 14psi peak down to 12psi by 5500. Once that is sorted I will post up the results. Again, this is a mild upgrade but may be of interest to guys with similar aims to what I have.

Edited April 14, 2007 by Dale FZ1

[wolverine]

hi dale

i know you have been chasing a "fruity" sound. congratulations.....lol.

a shame you couldn't try 2 degrees retarded as well as this might fill in the mid-range dip (or not). keep us posted if you change the gear setting.

[skylinecouple]

I too am watching you.2deg. would be interesting.Stick a 12" hotdog on that exhaust and be amazed how the drone will be gone.Personally I think the RB25 had to pass emission laws and received an exhaust cam that would help it acheive exactly that.A 256 exhaust cam might help as timing wheels are a bandaid on an open wound.Or for that matter, a set of cams.

[Dale FZ1]

After the dyno session, I adjusted the exhaust cam timing to 2 deg retarded as a compromise and will run it that way for a time. I have no doubt emissions engineers are responsible for a big chunk of the production design parameters, though at this power level I think cams are a bit of over-capitalising.

The exhaust system already has a resonator in it ex-Japan as part of the Blitz cat-back system, so it's basically a case of either cope with the drone, or take the extra power. Personally my ears and tinitus win the argument.

[gtst25]

Hi Dale,

Here's one tune and the next on my gtst. The first tune (341hp) wasn't the best as it was running pretty rich. I see you've been playing with the vct cut in a bit - we ended up moving the change over point to 5000 rpm and it made a bit more power where it dipped before and the dip is much less pronounced than before. Might be worth mucking around a bit next time on the dyno as it takes hardly any time with the datalogit.

Cheers

[benl1981]

Hey Dale - It looks like you used an R32 exh housing rather than the R33. I can't see the 21U on it and I can see a shaved off rib like on my one here that i think is R32. Let me guess it had "04U" cast on it?

Also,

Do you have the inducer and exducer dimensions. Also, why did you go for the larger comp housing than the nissan one? It is a little wider is it?

Thanks