Gtx3582r

[|CL|NIKI]

Many people had brocken chra's too (with precision). Wheels too. I will use comp like many.

[URQguy]

Like all turbos, you only have a certain length of powerband that is typically 3-4k rpm if VE is good. Older compressors tend to have poor aero and are fairly heavy compared to modern billet wheels. k24 7000 tend to run from 2800-6000, k24 7200 run from 3300-6500 before falling off, 3071 3600-7000, 3076r 4000-7500, 3582r 4500-8000. The GTX3582r is a beast of a turbo from 4200-8500. If you are not built to rev that high, or you do not care to, the GTX3582r is not for you. I hate the word LAG, as it is basically just a description of a user not having the engine in the powerband of the turbo/setup, and the turbo gets blamed.

The catch with the gtx3582r is that OEM audi turbo cams will have a hard time extracting the power out of a 3582r above 6800rpm, and 7a/nm cams won't do much better. Even a mild cam with a tad more duration at 7a lift levels would help tremendously.

[my2000apb]

Like all turbos, you only have a certain length of powerband that is typically 3-4k rpm if VE is good. Older compressors tend to have poor aero and are fairly heavy compared to modern billet wheels. k24 7000 tend to run from 2800-6000, k24 7200 run from 3300-6500 before falling off, 3071 3600-7000, 3076r 4000-7500, 3582r 4500-8000. The GTX3582r is a beast of a turbo from 4200-8500. If you are not built to rev that high, or you do not care to, the GTX3582r is not for you. I hate the word LAG, as it is basically just a description of a user not having the engine in the powerband of the turbo/setup, and the turbo gets blamed.

The catch with the gtx3582r is that OEM audi turbo cams will have a hard time extracting the power out of a 3582r above 6800rpm, and 7a/nm cams won't do much better. Even a mild cam with a tad more duration at 7a lift levels would help tremendously.

so when are you buying a cam grinder?

[Rocco]

Like all turbos, you only have a certain length of powerband that is typically 3-4k rpm if VE is good. Older compressors tend to have poor aero and are fairly heavy compared to modern billet wheels. k24 7000 tend to run from 2800-6000, k24 7200 run from 3300-6500 before falling off, 3071 3600-7000, 3076r 4000-7500, 3582r 4500-8000. The GTX3582r is a beast of a turbo from 4200-8500. If you are not built to rev that high, or you do not care to, the GTX3582r is not for you. I hate the word LAG, as it is basically just a description of a user not having the engine in the powerband of the turbo/setup, and the turbo gets blamed.

The catch with the gtx3582r is that OEM audi turbo cams will have a hard time extracting the power out of a 3582r above 6800rpm, and 7a/nm cams won't do much better. Even a mild cam with a tad more duration at 7a lift levels would help tremendously.

so hearing that is a perfect turbo for my mechanical head with high lift cams good for 8700rpm as the builder said

did you say 8500rpm and still pushing strong? problem would be stock oil pump?

[URQguy]

No, 8500 in a .82 AR housing will be the limit of the gtx3582r in a 2.2L 20vt with above average headwork. 8500 is actually kind of pushing it, 8200 will start to show signs of dropping off. When I say dropping off, I mean that it won't sustain 35psi to 8500. It will likely taper to around 33 by 8500. If you are running less than 33 psi all the time, it will hold 33 psi to redline and not drop off. We typically get used to just running as much as boost as it will make, say 40psi at torque peak, and then ride the compressor till it stops making power. This is called overspinning a turbo. lol I think the Garrett's by far are the most reliable in overspin/no wastegate situations.

GTX3582r is a 675-700whp turbo on e85 or exotic gas. IMO, it is the ultimate turbo for stock oil pumped 5 cylinders, as any bigger will not be able to take advantage of the rpm due to oil pump limitations. I have never had an oil pump "explode" like many claim here, but I wouldn't push it above 8500. I have seen wear on 8500 rpm race engines.

As far as the other turbo options, I am not affiliated with any turbo manufacture out there, but I am very partial to Garrett products due to the low failure rate. I am a big fan of the new PTE CEA turbines, and think they have pretty awesome products, it is just their customer service and relatively high failure rates that turn me off. I would love to try a new CEA 6262 at some point. Comps are in the same boat except you just have to stand over them with a whip to make sure your turbo is built the way you want it to be built. I like aspects of the EFR 8370 and have one here for testing, but that whole EFR fiasco was the most epic product development failure in aftermarket history.

[prj]

URQguy, this is mostly nitpicking, but...
You can make a powerband as artificially wide as you want, but it comes down to the fact that you probably could run more boost.

Specifically, if you look at how the OEMs map the cars, then they tend to have a huge powerband, because they run low boost at onset and ramp it up towards redline.
This makes the motor feel more like a N/A, while still maxing the compressor map at high altitudes, but it costs torque... and the car is not as quick.

3071R can hold about 26-27 psi until 7k, but it's tapering rapidly at that point.

With a 30R I find that they are out of puff at the stock rev limiter of 7200 rpm.
On a bone stock head, it'll do about 30 psi from 4k to 7k and then start falling off rapidly or producing stupid IAT's.
Yes, you can rev it to 7.5k, but your powerband is effectively gone after 7k and the pressure just starts falling off.

With a GTX35R you just add 600rpm to these numbers. If you run it at 37-38psi.

My main point is, you can map the motor to have a more linear feel and have a larger "flat" area of torque, but you will lose peak tq and depending on the turbo, peak hp.
Or you can go all out, and stick with a 3k powerband.

As for cams - I think they are over-rated. It's not a N/A motor, it's FI. You tend to be able to a quite large extent to compensate for lower VE with more boost, however this comes at the expense of efficiency.
Most people when comparing the "gains" they have gotten from improving the engine breathing measure the flow (or horsepower if you will) at the same pressure as before improving the breathing, and then make wild claims. However, what really should be compared is efficiency at the same flow...
Had you maxed out your turbo's flow before by running more boost, after improving engine breathing the boost pressure will fall off, because the turbo won't be able to flow the same anymore.

It's too bad people tend to forget that engines are all about flow and pressure is just a measure of resistance. This is not aimed at anyone specifically.

As for specifically 7A cams on GT35R - it's more or less okay, but you have to run high boost pressure down low, to be able to take advantage of the turbo.
However, the cams have no overlap... Although the car will breathe significantly better up top with better cams.

[URQguy]

I respect a lot of your work prj, but you are nitpicking! Is that not what I just said reworded? I sortof alluded to what you are saying in my post when I said a gtx3582r will hold 33psi to 8500 if you keep it to 33psi. It will not hold 38psi to redline 8500, especially with a VE above a stock head/cams(again, just reworded). I suppose I could run 10psi on my 3582 setup and have power from 3300-8500 and claim I have a 5000 rpm powerband, but that is not why people run aftermarket turbos. I don't see a lot of OEMs that raise the pressure across the rpm range. I see them doing the opposite, that is maximizing torque with higher shaft speeds at torque peak and then protecting the turbo by tapering to redline. I'd say 95% of oem turbo systems are designed that way to make them feel NA.


You are assuming a lot with A/R selection on hotsides in your assumtions, as well as fuel choice and compression ratios. A 3076r turbine with a 1.06 can move nearly as much mass as a 3582r housing with a .82 housing. The .63AR has no place in a 3076r and the GTX3076r really itches for either a 1.06 housing or a 35r turbine with a GTX76 wheel to get the turbine flow required to push 64lbs/min the compressor is rated at.

I do share your belief that most people do not need upgraded cams. Anything smaller than the billet 3582r turbos out there will not overwhelm ports/7a cams before they run out of steam. The GTX3582r is not one of those turbos though. No OEM cam has enough duration to support the cross section of a 20v port above 6800. I have never seen a realistic dyno plot that has made power past 7500 on stock cams, and most drop off dramatically after 6800 depending on cam retard. Retarding the cam will mask this slightly at the expense of powerband down low. If the air cna not move through the pump(ie the engine) it does not matter how much pressure you run, it will not get through the air pump. No matter how big the turbo is, it physically can not fudge VE, regardless of the PR.

I use 7a cams directly as perscribed on a HTA3582r. I run 42psi at torque peak around 5000 and taper to 32 by 8500 on a 160mm rod 2.6L talldeck with stock ports. I virtually do not use a wastegate. I'd like to get some mild Cat cams with closer to 270d of duration at around 12mm of lift to carry power to 8500 with a 1.06 housing. I have played with a 1.06 on the current setup, but the exta 400rpm worth of powerband shift is not rewarded up top because the cams(not the turbo) limit the airflow through the system.

[prj]

I don't see a lot of OEMs that raise the pressure across the rpm range. I see them doing the opposite, that is maximizing torque with higher shaft speeds at torque peak and then protecting the turbo by tapering to redline. I'd say 95% of oem turbo systems are designed that way to make them feel NA.

This is the opposite of what is going on in the OEM world.
Take any newer Audi and see how it's mapped. They don't even have a target pressure, they have a target filling.
And filling is kept constant through a band of up to 5000 RPM. So boost pressure has to go up due to IAT going up and breathing getting worse up top.

For the rest - we are saying the same things, and I agree with you
And I did start my previous post with saying it was mostly nitpicking! I covered my bases

[Rocco]

So you guys prefer full boost theory over flat torque true?

OEM cars looks for more flat torque figures true? Maybe to take more care with transmissions?

[Nik0las]

guys, what you think about GTX3076R with .82 A/R vs GTX3576R .63 A/R on 2.2t with 7A cams - engine spin to 7500~8000rpm tops!

[araipilot]

turbo GT30

yes he could buy my 575hp setup (GT30R,Repo tubular manifold and injetors)

Are you selling your set up? How much are you looking to get for the full monty

[Rocco]

Are you selling your set up? How much are you looking to get for the full monty

at the end of this week i will know 99% sure if the theorical owner will hold these parts,if not the whole kit would be sold for 1300€

[URQguy]

This is the opposite of what is going on in the OEM world.
Take any newer Audi and see how it's mapped. They don't even have a target pressure, they have a target filling.
And filling is kept constant through a band of up to 5000 RPM. So boost pressure has to go up due to IAT going up and breathing getting worse up top.

For the rest - we are saying the same things, and I agree with you
And I did start my previous post with saying it was mostly nitpicking! I covered my bases

Fair enough! I haven't calibrated anything newer than me7 2.7tt. I'll take your word for it, and the logic is there for maintaining a flat powerband. Like you said, you can widen a curve by boost strategy to make my mother feel like she has the same power everywhere she drives. Fortunately there are people like you that can ditch filling targets and work with keeping turbo shaft speeds slightly under the limit for the maximium torque on any given setup baring fuel and surge limitations!

As to the other question about the gtx3076r in a .82r vs a gtx3576r in a .63. You are missing the point. Neither one of those configurations will allow for enough corrected gas turbine flow to match the expense and flow capacity of the 64lbs/min gtx76r wheel. You will have a very spikey power delivery as you crossover from the thresh hold of enough flow capacity at around 5000 rpm to lack of flow capacity around 6500. If you are following prj or my own comments, the 30r turbine seems to fall off at about 7000 rpm to bellow 450hp in a .82ar housing. This is mostly turbine flow related. It does not matter how big the compressor is, it will still drop off there if the turbine/housing is kept constant. You would be better off with a gtx3071r in a .82 housing if you want to maximize torque under the curve.

[prj]

Fair enough! I haven't calibrated anything newer than me7 2.7tt. I'll take your word for it, and the logic is there for maintaining a flat powerband.

ME7 2.7TT, K04 factory:


2.5k-7k.

ME7 2.7TT, K03 factory:


1.8k-5k.

[jade]

Nice, I have a Precision 6262 that I'll be using on my stroker. This turbo is really popular and has delivered great results for alot of people. Can't wait to try it out

Hi, not to but in, I found when I ran 62 62 with a comp turbo I had a had surge problems I now run 58 62, no problems with surge at all

This was on a 2.2 not 2.6l like you