yes, I saw your post over on the STi forum about it, but just as there you put forth words without providing full explanations and details. If it was sent in for warranty then they will send back a document explaining why it was rejected. What did they say?

Because the one thing that will cause that problem is overspeeding the turbo. Unlike some other turbos, the efficiency on the G-series at the RH map limit is such that the turbo won’t stall or fall off, but will keep pulling hard on past the limit. Especially if too small of a turbine housing was used. After which it will eventually start to fall/taper off, but the damage is done then. A speed sensor and software strategy are highly recommended if the turbo will operate anywhere near the limit.

So we can’t really blame either the turbo or manufacturer for that, but again we only know what is provided, which is very little. So without the details and information it’s just as hard to pinpoint anything on what you’re saying any more than understanding the photos we’re looking at. Because as you know, in that same thread an STi with G25-660 made 500 whp without any trouble at all.

That owner provided all the info and full details on his build though. This turbo does peak at 60 lbs/min, but only right around ~2.55 Pr. Anywhere above or below that peak then the limit is less than 60 lbs/min. The info I’ve seen indicates that G25-660 0.92 turbine outflows the much larger EFR7670 0.92 turbine. The compressor limit is a bit lower though, but with better efficiency at the limit. So it shouldn’t be too far off whatever an EFR7670 will achieve in a much smaller package and with much better spool.

The divided T4 0.92 IWG housing is due out any day now.
.

 

The first, which was unused, was denied warranty. The two turbos did not mic out remotely close, the castings were so different, the piping no longer fit, and the first would not even spin by hand.

I have the logs, 24 PSI (~ 2.3 PR) did not overspin the turbo. Turbine housing was a .91 and this turbo did fall off... I have the dyno graphs and VE maps to show this. The 500whp dyno also spooled up 1,000 RPM sooner and didn't fall off horribly. I've tuned countless cars, including EFR's and Gen 2 Garrett's... I've never had a turbo let go in almost 20 years of doing this. I've had a 7670 make more power and spool earlier, without blowing itself apart, same generation of car.

We are still waiting for a response from anyone. A big reason on why we have little faith is the results from the first turbo.

 

The first turbo, was it from when the turbo was first released? I know they had turbine housing casting supplier issues and some rubbed right out of the box. A turbine wheel rub pretty much explains the behavior of the first turbo and eventual destruction. As you have seen, the new turbine housings are from a new supplier.

 

The first turbo was never fired up. It rubbed the housing from the beginning. I do no know when he purchased it, however, they left him out to dry with the purchase (unused turbo, with picture proof, and measurements).

New supplier is fine... If all the mounting parts are in the same location, which was not the case.

 

Wow thats some shite luck. Knock on wood my turbo hasnt had any issues so far. I did free hand spin it off the car and it didnt contact the housings at all. Hopefully wont have any issues.

 

I'm not fully understanding this, but how would the assessment be off, would the g30 660 be making less power by your take?

 

You are correct, there is more to impeller size Turbine efficiency is a function of both turbine housing A/R and also blade speed ratio; blade speed ratio is a function of the compressor diameter paired to the turbine diameter. For a given turbine wheel, it will have a different value for peak efficiency and also a different location for peak efficiency vs turbine pressure ratio for different A/R turbine housings. Leaving the turbine wheel and housing A/R the same, the peak efficiency will change depending on the size/flow of the compressor wheel on the other end of the shaft. There is a sweet spot. As you get further away, the blade speed ratio gets worse, turbine efficiency drops, and the turbo gets laggier and makes less power. To go to a bit of an extreme, you wouldn't make a G25-900 as the compressor to turbine match would result in horrible blade speed ratio and therefore horrible turbine efficiency. If you're a cyclist, think being in the biggest gear up front, smallest in the rear, and trying to move from a standstill.

A 60mm diameter compressor pairs really well with a 54mm turbine. That's why the G25-550 is a really great overall turbo. Pairing the 67mm compressor to the 54mm turbine is a bit of a mismatch. The old GTX analogy is a GTX3067 will make 25whp more than a GTX2867 with the same spool if put in the same turbine housing casting. So the GTX3067 turbine actually flows more in the same casting than the GTX2867, but the two turbos will spool the same because the GTX3067 has a better blade speed ratio resulting in better turbine efficiency. Turbine power is a function of both efficiency and pressure ratio, so the GTX3067 can make the same turbine shaft power at a lower pressure ratio because of higher efficiency.

The interesting bit with the G25-660 in the 0.92 A/R and the G30-660 in the 0.83 A/R is they actually flow about the same in the turbine stage. I figure the G30-660 w/ 0.83 A/R will spool faster and make more power than the G25-660 in the 0.92 A/R.

 

That depends on the engine and what it needs. Because the flip side is that the 0.83 might be spooling faster, but then choking early with high emap and making less peak power. If the engine is small enough and boosting to 30+ psig then your concept there might be in play.

I haven’t run the numbers, but based on everything I’ve seen it isn’t going to play out that way on an S2000 engine. The 0.83 is either an open T3 or v-band, which is just a lighter, more compact open T3. The divided T4 is going to be offsetting some of that by keeping the pulses separated despite the higher A/R.

Perhaps you also overlooked the general concept of having one size difference between an open vs divided housing. And also that the G30 has a larger/heavier turbine wheel with more MOI than the G25.

It seems to me that your perspective hasn’t thoroughly considered how that will play out in reality.

.

 

that depends, full out on an S2k engine, my feeling is that the G30 0.83 housing will choke early compared to G25 0.92 housing. The G30 1.01 housing is going to outflow the G25 0.92, but the low end response will definitely be off some seeing it’s both higher A/R and an non-divided (open) housing compared to the G25 divided housing.

It may run out of compressor flow before a 1.01 turbine housing can justify the low end response loss. Again, I haven’t run the numbers, but have seen a few S2k EFR7670 results. So I’m only projecting feelings based on my perceptions rather than facts.

The way I see it is the the G25-660 0.92 divided T4 will be what most people had hoped for out of the EFR7163. It essentially is the 7163 compressor flow with 7670 turbine flow, but in a smaller package than the 7163.

But that 0.92 T4 housing has been expected since being announced in Feb/2021 and is still not released yet. So the option doesn’t actually exist at this time. Just Garrett is saying it’s coming any day/week/month now.

honestly I don’t much care for either of the G30-660/770 turbos. If you compare them against the G30-900 then you can see their short coming. My view is there’s sort of a poorly filled gap between the G25-660 and the G30-900, but I can appreciate Garret wanting to avoid the cost of making a tweener size option there instead.

.

 

I still think the 7163 is a physically smaller package. I wanted it for that reason, however, the exhaust side wasn't big enough.