The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support Fuel Injector Clinic
Please Support Morrison Fabrication

GT3782 vs HX35/40

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

r5baker

Probationary Member
4
0
May 5, 2022
Blacksburg, Virginia
No DSM here but this forum seems full of useful relevant knowledge. I have a '91 MR2 with a 2nd gen 3SGTE. For years I had a Garrett GT3271 on the car using a tubular divided manifold. Spool was very fast with this turbo and I had it pretty much maxed out at 25psi and 400whp.

The turbo was rebuilt once with a 360 thrust bearing but finally crapped out this year with oil coking up the bearings. For 'fun' I found an Ebay Garrett GT3782 that was on a Navistar DT466E and cut the exhaust housing flanges off and welded on a divided T3 to the T4 1.11A/R housing along with a 3" Vband. Nickel 55 rod with lots of preheat, post heat, and slow cool.

I put it on the dyno and it put down right at 500whp with very slow spool compared to the GT32 which I expected. Don't get me wrong, I like the additional power, but on the street fast spool is much more fun. The dyno plots below are both 3rd and 4th gear runs on the same plots.

I just swapped out the GT3782 project for a Tims's Turbos built HX35/40 with the 7-blade HX40 60x84mm compressor and the HX35 12-blade 70x60mm turbine. I have not returned to the dyno but spool is nearly identical with less top end. Any insight on whether this is expected with the Holset hybrid?

I am just puzzled that the 12cm^2 divided housing driving a smaller 70x60 turbine spools up nearly identically as the 1.11 A/R T4 housing driving a 72.5x66 turbine on the GT3782? In 4th gear datalogging shows only a 200rpm advantage to the Holset. In 1st through 3rd there is no advantage and like I said seems to run richer at the same boost 28-29psi with no fuel changes and speed density. Fuel is E85 (likely more like E50-E60) that I have not measured the percentage yet.

You must be logged in to view this image or video.


You must be logged in to view this image or video.


You must be logged in to view this image or video.


You must be logged in to view this image or video.


You must be logged in to view this image or video.
 
Welcome to the forum r5baker!
We have a resident person that should be able to help out on this.
Hey Justin @JusMX141 , can you give some input to this question or shed any light on what is being compared?
Thanks!
Marty

Justin is our forums turbo man, if anyone will know, he will. :thumb:
 
Apart from it being difficult to compare response between two turbos on an unloaded dyno, there's no way to pinpoint exactly where twin-scroll T3 becomes a restriction and the turbine flow yields to T4...it seems to vary greatly by setup but I typically use 550whp as a rule of thumb for the "stopping point" of twin-scroll T3 on our platform.

What I can tell you, with confidence, is that nobody I know has ever put a HX40 compressor on a HX35 turbine and run it in the stock twin-scroll T3 housing with miraculous results- it almost always seems to be a disappointment. That particular wheel mismatch is typically reserved for diesels and/or small-displacement applications in the 1.5L to 1.8L department unless your setup is really geared toward handling excessive turbine flow in stride. The higher the compression/displacement or the poorer the fuel you're trying to run the sooner you're going to run out of turbine flow.

Somewhere on here a member ran a straight HX40 in the factory T4 17cm2 housing into the 9's on a stock converter...I'll see if I can find the thread and edit. So apparently spool and response was pretty stellar on a full HX40 rotor group in the correct housing(s). So HX35 as T3 twin or HX40 as T4 twin- just don't try to bridge the gap between the two.


Clean welds on that v-band install, by the way...awesome work!
 
Apart from it being difficult to compare response between two turbos on an unloaded dyno, there's no way to pinpoint exactly where twin-scroll T3 becomes a restriction and the turbine flow yields to T4...it seems to vary greatly by setup but I typically use 550whp as a rule of thumb for the "stopping point" of twin-scroll T3 on our platform.

What I can tell you, with confidence, is that nobody I know has ever put a HX40 compressor on a HX35 turbine and run it in the stock twin-scroll T3 housing with miraculous results- it almost always seems to be a disappointment. That particular wheel mismatch is typically reserved for diesels and/or small-displacement applications in the 1.5L to 1.8L department unless your setup is really geared toward handling excessive turbine flow in stride. The higher the compression/displacement or the poorer the fuel you're trying to run the sooner you're going to run out of turbine flow.

Somewhere on here a member ran a straight HX40 in the factory T4 17cm2 housing into the 9's on a stock converter...I'll see if I can find the thread and edit. So apparently spool and response was pretty stellar on a full HX40 rotor group in the correct housing(s). So HX35 as T3 twin or HX40 as T4 twin- just don't try to bridge the gap between the two.


Clean welds on that v-band install, by the way...awesome work!


That was beans and his laser. One of my idol DSMs was his setup. The spool up time on his divided t4 hx40 was so awesome he could stall it up on the stock converter. Still my dream setup for sure.
 
Thank you for the detailed responses. I guess that was not really the answer I wanted to hear about the HX35/40. What I was looking for was the fastest spooling 500whp solution so it sounds like a straight up HX35 would be better. I also really don't want to go over a peak of 30psi of boost, at least right now anyway.

I like the journal bearing turbos for their simplicity which is why I turned to the Holsets and the configuration I have is what was recommended. I think the hybrid has more potential if I ran a higher boost pressure, but that would require a good deal more invested in supporting modifications like larger injectors, pump, IC, 4-bar MAP, higher torque capable clutch, reliability issues.

What are you guys seeing with 2L setups for spool on a HX35 and what configurations work the best? There seem to be many many options when it comes to the permutations of the Holsets. I am running a divided T3 manifold if that helps. Maybe I'll see if Tim can rework my turbo for better spool from your recommendations while still keeping that top end rush. I do like an occasional autocross.

Here is a link to the dyno video in 3rd. MR2 turbo dyno April '22
 
The stock HX35 compressor is 60lb/min capable but it's not likely to happen at 30psi or below unless your setup is highly optimized...although the response will definitely be there, especially if you're running a well-built small-runner manifold to keep velocity up.

The 35/40 hybrid will absolutely not have more potential at higher boost if you're already out of turbine flow, which is what the dyno chart is indicating where there's a substantial power loss above 5500 despite having a 60mm compressor that I know first-hand is capable of 70 lb/min. It will do the exact opposite, actually...like over-thrust the turbo and potentially lift and/or torch the head on the car, and if EGTs are high enough potentially melt or eject the turbine as well.

This is comparable to the SRT4 platform where the owners mate the largest compressor wheel they can get their hands on to a stock TD04LR turbine because that's what the factory turbo is limited to with the manifold and turbine housing being one piece. So they go out and make 400whp before 4000rpms at max boost/torque and then the graph anchors away to like 220whp at 5500+ because it's completely out of turbine flow from a 2.4L trying to breathe through a 41mm turbine and the wastegate won't stay shut...but they carry that 400whp dyno sheet like a diploma even though that much torquey power is 100% useless on a FWD.

Apart from knowing that turbine flow is key to response and reliability, there is no real "correct answer" when it comes to adapting aftermarket turbos to certain platforms....this is why Precision makes both a 6262 and a 6266, and people will argue to the death about which one is better despite using the same compressor wheel even though the 6262 user in the argument is on a stock block with E85 at 30psi and the 6266 user is on pump gas with a built 2.3L stroker at 30psi and if they swapped one would lose response and the other would lose power then they'd both hate it.


Regardless, a capable builder should be able to take the information you have here and put something together that meets your requirement even if it's slightly inadequate in the total power production department. There are TONS of available Holset compressors in the HX/HE size frame that will fit and work without ruining the powerband based around your desired boost level. The conversation gets more-technical when discussing compressor trim, blade count, and overall compressor aero and never be afraid to come up with something that wasn't necessarily offered from the manufacturer- like using the 56/78mm HE341 compressor wheel over the stock 54/78mm HX35 compressor to bump up the trim calculation a bit without increasing the overall O.D. (and rotating mass) of the compressor and adding unnecessary lag...and so forth.
 
What are you guys seeing with 2L setups for spool on a HX35 and what configurations work the best?
Here is a link to the dyno video in 3rd. MR2 turbo dyno April '22
2.0L with an Hx35 on a twin scroll Morrison Fab header HERE. I think it was in a stock Holset hot side too. Car would have ran a 10 if he hadn't melted a piston at the 1000' mark because of too much timing.
Its my son ;) He's in the far lane in his 95 Talon TSI. He still won too! ROFL
This is what happened.

You must be logged in to view this image or video.
 
Last edited:
I have the 7 blade hx35 and dynod 530. The turbo spools quick but does fall off at higher rpm. I see full boost at 33-35 by 4500-4800. I'm on a true twin scroll setup. I also tried two hotsides and moves the powerband 300-500 rpm difference. I can't speak about the other turbos mentioned but wanted to share my experience with the holsets. I do want to try a hx40 to get into the 600whp but unsure at the moment. The holsets are a great budget option but turbo technology is way better these days if you can afford it. Xona rotor turbos is also a good option.
 
What I am getting presently recommended as the fastest spooling Holset for 500whp is a 7-blade HX35 compressor with the 0.63 A/R undivided T3 exhaust. Anyone run this combo?

How would an undivided 0.63 A/R spool faster than the 12cm dual scroll? Seems counterintuitive but if it works it works.

Too bad about the melted piston, BTDT.

So how does Garrett get away with there newer G25 series making over 600hp on such a small turbine?

Garrett G25-660

I mean a 55x49mm turbine that can flow 600hp worth of exhaust yet an HX35 turbine which is 70x60mm can only flow enough for 550?
 
Wish I could give an answer. I just happened to use what worked for us. Justin should see this and maybe he can give us the lowdown.
 
Hey 1990, what specs are you running on your HX35? 12cm^2 divided with the 7-blade 35 compressor?
 
2.0L with an Hx35 on a twin scroll Morrison Fab header HERE. I think it was in a stock Holset hot side too. Car would have ran a 10 if he hadn't melted a piston at the 1000' mark because of too much timing.
Its my son ;) He's in the far lane in his 95 Talon TSI. He still won too! ROFL
This is what happened.

You must be logged in to view this image or video.
How much timing was too much timing?!?!
 
19*
Setup was on 15*, got bumped up per driver mod. Motor was NICE!!!.
Anything over 16* is just testing to see if it runs better or worse since E85 won't show knock at all until its too late. Stay between 15 -17. 18 max is what I personally ran but it made no difference so I backed down to 17*.
 
So how does Garrett get away with there newer G25 series making over 600hp on such a small turbine?

Garrett G25-660

I mean a 55x49mm turbine that can flow 600hp worth of exhaust yet an HX35 turbine which is 70x60mm can only flow enough for 550?
.92 a/r housing designed specifically for that turbo compared to .49 (7cm2) on a typical DSM housing or .55 on the Bullseye housing which was an adaptation.

The first key ingredient is that nothing about the HX35 turbine is performance-minded in terms of flow for it's size, but it works. It has thick blades, a bulky hub, and a high blade count but can take a foreign object strike in stride.

It's worth mentioning that I don't know anyone in our platform who has made anywhere near the peak rating of a G25-660 using that turbo; it's also worth mentioning if you could somehow adapt a G25-660 to a .49 7cm2 DSM housing it would likely perform comparably to any other 7cm2 turbo with a TD05H turbine in terms of response and peak flow.
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top