When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I see that's pretty interesting. I would assume that the 1.06 divided wouldn't be much different from the 0.83 V-band.
Down the line, would you consider upgrading to the G35-990 (62mm) or G35-1150 (68mm)? Flow has been improved quite a bit from the G35-900 and G35-1150 despite being the same size compressor and turbine wise. I imagine with the G35-1150 you would hit 900whp no problem with similar boost levels.
I see that's pretty interesting. I would assume that the 1.06 divided wouldn't be much different from the 0.83 V-band.
Down the line, would you consider upgrading to the G35-990 (62mm) or G35-1150 (68mm)? Flow has been improved quite a bit from the G35-900 and G35-1150 despite being the same size compressor and turbine wise. I imagine with the G35-1150 you would hit 900whp no problem with similar boost levels.
i guess when you look at the actual space/area the twinscroll 1.06 is actually smaller than the .83. So that explains it choking up top.
not sure if I would want to try the gen2 1150. If I had to change this turbo I may consider the gen 2 990 to hopefully get more response down low and pointe tidally the same power up top
I see that's pretty interesting. I would assume that the 1.06 divided wouldn't be much different from the 0.83 V-band.
Down the line, would you consider upgrading to the G35-990 (62mm) or G35-1150 (68mm)? Flow has been improved quite a bit from the G35-900 and G35-1150 despite being the same size compressor and turbine wise. I imagine with the G35-1150 you would hit 900whp no problem with similar boost levels.
so if you look at the actual area space the divided 1.06 housing is smaller than the .83 v-band housing. That’s why it allows it to breath better up top.
looking at the gen 2 compressor maps it seems that they sacrifice a little down low or are not as wide in return for top end. I would like to see these in action to really see if I would like to change in the future. The g35-990 would interest me if it provides better low end and about the same power
looking at the gen 2 compressor maps it seems that they sacrifice a little down low or are not as wide in return for top end. I would like to see these in action to really see if I would like to change in the future. The g35-990 would interest me if it provides better low end and about the same power
If you compare the G35-990 to your G35-1050 there’s definitely less flow at lower boost, but that’s because the compressor wheel is 62mm on the 990 vs 68mm on the 1050.
Looking at the gen 1 62mm vs gen 2 62mm there is essentially no loss. I am curious how the Gen 2 68mm compressor map will look when Garrett releases it. All the other gen 2 maps have essentially no loss of power on lower boost as well.
What I am curious about, is how are G35-1050s surviving making 800+whp. 800whp is a theoretical 900-950 crank hp and the compressor map tops out ~45psi of boost and flows only 75lbs/sec (~750 crank hp). Only way to flow enough for 800whp at the high boost levels evo need is be overspun.
Not sure if you have enough spare digital inputs on your Haltech, but it would be interesting to add a speed sensor down the line to test Garrett’s compressor map vs reality.
If you compare the G35-990 to your G35-1050 there’s definitely less flow at lower boost, but that’s because the compressor wheel is 62mm on the 990 vs 68mm on the 1050.
Looking at the gen 1 62mm vs gen 2 62mm there is essentially no loss. I am curious how the Gen 2 68mm compressor map will look when Garrett releases it. All the other gen 2 maps have essentially no loss of power on lower boost as well.
What I am curious about, is how are G35-1050s surviving making 800+whp. 800whp is a theoretical 900-950 crank hp and the compressor map tops out ~45psi of boost and flows only 75lbs/sec (~750 crank hp). Only way to flow enough for 800whp at the high boost levels evo need is be overspun.
Not sure if you have enough spare digital inputs on your Haltech, but it would be interesting to add a speed sensor down the line to test Garrett’s compressor map vs reality.
I went looking for the compressor maps again but could only find the g35-900 vs 990. I know I have seen the 1050 vs 1150 before.
I do plan on adding a speed sensor to monitor that in the near future. Also keep in mind that on the dyno we didn’t go past 42psi.
I add the compressor map to show what I meant by the they sacrifice down low and is not as wide in some parts but gain more flow. It looks more peaky. I’m not really sure how to factually interpret the compressor maps so I may be thinking of this all wrong haha
I went looking for the compressor maps again but could only find the g35-900 vs 990. I know I have seen the 1050 vs 1150 before.
I do plan on adding a speed sensor to monitor that in the near future. Also keep in mind that on the dyno we didn’t go past 42psi.
I add the compressor map to show what I meant by the they sacrifice down low and is not as wide in some parts but gain more flow. It looks more peaky. I’m not really sure how to factually interpret the compressor maps so I may be thinking of this all wrong haha
Sorry for the giant explanation below haha.
But ya, I hope they release the G35-1150 map soon. And nice that you plan to add a speed sensor. It will definitely help you know how far you are on the compressor map when your car is producing peak power. If you are at sea level, 42 psi is around 3.8 to 4.0 on the pressure ratio axis on the left. So on your G35-1050, the turbo will flow around 80lbs/min before it overspeeds. I plotted where you might be on the G35-1050 map I posted below. Point 1 is theoretical max of the turbo at 42psi before overspeed. Point 2 is where you might be according to your power levels (~95lbs/min). Rule of thumb is 10lbs/min for every 100 crank hp. This changes based on factors like efficiency, DA, back pressure etc.
As for the map, ya it can be confusing haha. You are right that it is more peaky. This tall and narrow map is actually a good thing for small displacement motors like on our evos.
How far left the map goes basically shows the least amount of flow a turbo can make at a certain boost level before surging (flow going backwards because it is producing more than the engine can eat). An example I experienced where a further left boundary is needed is when I was playing with rolling anti lag at lower rpms on my Supra. It basically created too much boost and the motor wasn’t at a high enough rpm to suck it all in.
The Gen2 G35-990 actually looks to be a mild upgrade to the Gen1 G35-1050 in your application due to the very high pressure ratios you're running. Don't expect the spool-up to be any better because the Flow/PR for a given compressor shaft speed is very similar between the Gen2 76mm exducer wheel of the 990 vs the Gen1 84mm 1050 wheel. The transient response should be a bit better though for the G35-990 due to the wheel being smaller diameter = less inertia. If you want more significantly more power though, you'll need to step up to a G40-1250 I think. But, that's a new manifold setup, etc.
But ya, I hope they release the G35-1150 map soon. And nice that you plan to add a speed sensor. It will definitely help you know how far you are on the compressor map when your car is producing peak power. If you are at sea level, 42 psi is around 3.8 to 4.0 on the pressure ratio axis on the left. So on your G35-1050, the turbo will flow around 80lbs/min before it overspeeds. I plotted where you might be on the G35-1050 map I posted below. Point 1 is theoretical max of the turbo at 42psi before overspeed. Point 2 is where you might be according to your power levels (~95lbs/min). Rule of thumb is 10lbs/min for every 100 crank hp. This changes based on factors like efficiency, DA, back pressure etc.
As for the map, ya it can be confusing haha. You are right that it is more peaky. This tall and narrow map is actually a good thing for small displacement motors like on our evos.
How far left the map goes basically shows the least amount of flow a turbo can make at a certain boost level before surging (flow going backwards because it is producing more than the engine can eat). An example I experienced where a further left boundary is needed is when I was playing with rolling anti lag at lower rpms on my Supra. It basically created too much boost and the motor wasn’t at a high enough rpm to suck it all in.
G35-1050 Map
nice! Thank you for the clarification. I was kind of thinking around the same thing but was a little confused I guess. Looking at how you plotted it I am definitely pushing the limits on this turbo haha. I need to get that speed sensor soon
Originally Posted by spdracerut
The Gen2 G35-990 actually looks to be a mild upgrade to the Gen1 G35-1050 in your application due to the very high pressure ratios you're running. Don't expect the spool-up to be any better because the Flow/PR for a given compressor shaft speed is very similar between the Gen2 76mm exducer wheel of the 990 vs the Gen1 84mm 1050 wheel. The transient response should be a bit better though for the G35-990 due to the wheel being smaller diameter = less inertia. If you want more significantly more power though, you'll need to step up to a G40-1250 I think. But, that's a new manifold setup, etc.
I definitely do not have interest in going any bigger. With how reliable Garrett turbos seem to be I hope to stick with this for a few years haha
Ah, yeah, Robert of Forced Performance. He's so full of BS. Yeah, he knows some stuff, but he says stupid stuff too. Somewhere back in the 2008-2010 timeframe, he made a post here saying the visible machine tool path on point-milled CNC'd billet compressor wheels were an aerodynamic feature. Yeah, no. Newer designs, and I imagine just about every billet compressor wheel these days used in high-volume OEM applications, is made by flank-milling which is much faster (i.e. cheaper and also able to crank out more parts in a day) than point-milling. There are potential design compromises because the blade surface of a flank-milled design can't be as complex as point-milled, but I'd say the aero engineers are pretty good at making the performance targets these days with flank-milled designs. Flank-milling leaves basically a smooth finish on the blade surfaces as opposed to seeing all the tool lines flowing along the blade surfaces.
Now, I didn't watch the video beyond the first minute because I have better things to do than sit through 2 hours of Robert chatting. He was right about one thing and that's Pulsar buying Garrett turbos, scanning them in, copying them basically exactly, and Garrett can't do anything about it legally for whatever reasons. Specific to 4-cylinder engines, it sounds like it was in reference to very high horsepower builds. It should be noted Garrett is the OEM turbo supplier for 4-cylinder engines like the 2.3L Ford Ecoboost, 2.0L VW engines, the twin-turbos on the BMW 4.4L V8s (so one per 4-cylinders), etc. And the Prodrive P25 uses a Garrett G25-550. And a couple decades of VNT turbos for 4-cylinder diesel applications common in Europe and Asia. Anyway, if he was referencing very high horsepower 4-cylinder builds, i.e. small displacement, that implies compressor designs that are targeted for higher pressure ratios. Like the old Gen1 GTX wheels were a bit higher pressure ratio design than the Gen2 GTX. It looks like the Gen2 G-Series in the G40 and bigger frame size is a higher pressure ratio design wheel than the Gen1. Garrett has many families of wheel designs for specific applications, across all flow ranges and pressure ratios. The Aftermarket team will grab one of those wheels and put them in all their aftermarket turbos. The turbos are not marketed as being targeted for higher or lower pressure ratios, but they exist within the different offerings. As an example, high-output 4-bangers typically need a wheel designed for higher pressure ratios. Same with diesels. Big, gas V8s typically need a lower pressure ratio biased compressor wheel. And then there's the turbine side of things, but that's a whole other topic.
Ok thanks for weighing in on this topic. I know you have more experience with garrett turbos. Building a custom Garrett turbo is not something that I would believe a customer could request unless I am mistaken. Anyways a couple years ago AMS built a test rig with an Evo and iirc they wanted to produce results for the current line of garrett turbos with that test rig. I wish they published the results of that testing since I signed up for the email chain to be notified when the results were finalized.
It would have been nice to have that. As we could characterize the compressor maps with dyno data to see how close each turbo was in a theoretical vs empirical way.
nice! Thank you for the clarification. I was kind of thinking around the same thing but was a little confused I guess. Looking at how you plotted it I am definitely pushing the limits on this turbo haha. I need to get that speed sensor soon
I definitely do not have interest in going any bigger. With how reliable Garrett turbos seem to be I hope to stick with this for a few years haha
Absolutely get the speed sensor if you're going to push the turbo to the edge. And from rough numbers, you're already overspeeding the turbo. Keep in mind the compressor pressure ratio on the compressor map is measured as the ratio of compressor outlet vs compressor inlet. The pressure at the compressor inlet drops due to intake (filter) restriction and the compressor outlet pressure is significantly higher than intake manifold pressure (pressure drops from intercooler and plumbing). As an example, you're seeing 42psi at the intake manifold. So, that's maybe 46psi at the compressor outlet gauge pressure, which is about 61psi absolute pressure (14.7 psi atmospheric, rounding up a bit). And your inlet pressure has dropped about 3psi, call it down to 12psi absolute. 61/12 = 5.08 pressure ratio.
The turbos will hold up for a bit if you over spin them, but not forever. The compressor wheel is designed to be strong enough to 'last forever' at the max rated speed. If you spin it faster, you're going to stress the wheel more and shorten its life. It's like if your engine has a redline of 8k rpms, but you spin it up to 9k. It'll hold up for a bit, but eventually blow up.
Ok thanks for weighing in on this topic. I know you have more experience with garrett turbos. Building a custom Garrett turbo is not something that I would believe a customer could request unless I am mistaken. Anyways a couple years ago AMS built a test rig with an Evo and iirc they wanted to produce results for the current line of garrett turbos with that test rig. I wish they published the results of that testing since I signed up for the email chain to be notified when the results were finalized.
It would have been nice to have that. As we could characterize the compressor maps with dyno data to see how close each turbo was in a theoretical vs empirical way.
Yeah, Garrett has no capability to make custom configuration turbos. It takes forever to get a configuration into Production due to all the testing/qualification stuff. Burst containment qualification being a huge one. And these things are mass produced with a lot of automation, not hand built. I mean, I guess you COULD get custom ones made by the Motorsports division, but now you're talking like $8k-$10k+.
Yeah, would have been cool if AMS released the data. But I get why they wouldn't, being a distributor for Garrett. AMS probably just used for their internal programs. If you didn't see this article way back in the day.
Just some thoughts on turbo's
Garrett- hard to kill, makes lower power, great for road course reliability
Precision/Xona- makes more power, quicker spool up but can break here and there if really pushed. . Hard to be a NG 6466, or xona 9569 /6869 for 900whp+
Pulsar- hit or miss but may be getting better.
Other stock frame then FP. slower spooling and less power.
