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.
welcome, just going to tell you I almost joined up over there in some of those threads to point out that the reason the Garrett G25-660s blew up were due to a classic case of overspeeding. That’s not Garrett’s fault or a warranty situation. It was a mistake to blame Garrett. It’s happened on a few different G-Series models and in every case there was no speed sensor installed. The reason is because the G turbo compressor wheel assembly runs out of centrifugal strength before it and the turbine run out of efficiency if pushed off the far right side of the compressor map.
Garrett warrantied the turbos as we provided them logs and data to get an idea of where the turbo speed was at (you actually can do some math here, which I do anyway, being at high altitude), and could not find anything to deny the warranty. You're more than welcome to show me through the process where it was over-speeding. Using a dyno to find out where the turbo made power, we never found the turbo to live up to its rated capability. It is a "bigger" turbo than the 7163 and the 7163 easily made way more power, more easily. Additionally, when the capability was shown on the dyno, we turned it down quite a bit.
Originally Posted by TeamRX8
You’re not getting there on the 7163 yet because the turbine housing is too small, coupled with the lower efficiencies results in it stalling out. So you also need more turbine A/R, except then you’ll likely start seeing the ceramic turbine catastrophically destructing without a speed sensor and overspeed limit software strategy. You can’t guess at over-speeding if operating near the limit, this is where the current technology is going now.
The car that had the G25-660 is getting an EFR 7163 and a speed sensor, which I would have liked on the previous turbo, but I don't own the car. Being on a stock ECU with Cobb, we won't have the ability to use the input for any safeties, just data to view and log.
Both of the Garrett turbos were very early production runs and we would have liked our second one to be a later batch. We also talked to someone else, using the same turbo on their time attack car and had the same issues. He eventually upgraded to a turbo with a bigger shaft.
I couldn't fathom how many cars I have tuned in probably close to 20 years, these are the ONLY two turbos I've ever had blow up. I consulted with quite a few people during this process, so this wasn't going in blindly. Additionally, I had another tuner there (the in-house tuner at the dyno I use) who had tuned a few of these turbos and was just as stumped/irritated as me in how it was performing. So if you say I can't "guess", how are you to "guess" without even having any of the available data nor have been there for the process of dialing it in?
On the old gen 1 GTX3576R, the T04S compressor housing of the GTX3582 was worth 2 lbs/min more flow than the standard T04E housing which was more compact. So not surprised at all to see SX-E comp housing on the EFR7163 make more power.
well Jaz, I could have sworn it said the damaged G25s were not warrantied, but those were my thoughts and opinions based on what I had seen on it compared to other similar failures. However, I don’t have any problem saying they’re only my opinions and observations of what was presented on an internet forum. So not having the first person experience as yourself, I accept your accounting of it. However, without actual compressor speed data then pushing out to the right on the G-series compressor map is always going to be the equivalent of Russian roulette imo.
it is an EFR7670 thread, but like any internet forum it can get into a few side tangents on similar but related variations
so heres the EFR7163 variation; 0.80 A/R IWG V-band housing on left, 1.25 A/R EWG divided T3 right
a bit more fitting for a 13B application, kind of like a straw vs a garden hose:
Over time I’ve concluded that V-band turbine inlet housings are not quite matching up to the superior flow claims of round vs rectangular/oval inlets. Divided housings still offer benefits to certain applications and circumstances.
well Jaz, I could have sworn it said the damaged G25s were not warrantied, but those were my thoughts and opinions based on what I had seen on it compared to other similar failures. However, I don’t have any problem saying they’re only my opinions and observations of what was presented on an internet forum. So not having the first person experience as yourself, I accept your accounting of it. However, without actual compressor speed data then pushing out to the right on the G-series compressor map is always going to be the equivalent of Russian roulette imo.
it is an EFR7670 thread, but like any internet forum it can get into a few side tangents on similar but related variations
I get how forums work... However, this is really a different subject from another board, trying to point out something for being my fault without having all the information. That's fine, if you like to take digs like that (still seems that way, too). It doesn't really bother me, but it doesn't mean I won't address it and you probably should have signed up on NASIOC/S2KI/IWSTI to post it on a thread actively speaking on said subject.
As far as turbo speed, how long have we been sizing turbos and using compressor maps to figure out what turbo we can use? How long have people been easily able to monitor turbo speed like we do these days (and most don't do it anyway). A little bit of math is all that is really needed to have a pretty good idea of what it's going to do... Hell, it's what they do for engineering these things I've intentionally overspun turbos over the years without having failures and for each for those two G25-660 turbos that came apart, they only went when being in sustained use, like an autoX (last one was an autoX that never got out of 2nd gear - video below where you'll see exactly when the turbo let go) and the logs on the last one showed that boost never exceeded 12 PSI on the runs (low gear, not a lot of time WOT).
I won't get into comparing an inefficient Subaru motor vs a motor with so many combustion events per rotation. They will certainly have different requirements. The test we did was a Subaru and a compressor cover comparison.
EDIT: What you read being as denied warranty was the first turbo (technically, we've gone through 3). The first turbo was never used as the compressor hit the compressor cover when installed. It would not spin by hand. It was denied warranty (which was another drama/issue I don't prefer to air), until we had the next one blow up (BTW, the castings between the two turbos were so far off, the piping had to be re-done). Since we had someone in our group that had some company information from either working there at some point, or something similar, the e-mails went directly to the top as everything else got ignored (which went on to serve another Garrett issue, later). So yeah, there is a lot you're not privy to, which is something I would expect being in similar shoes.
Franky, you arguing against an actual wheel speed readout would make me question you more than anything else. Because I didn’t figure out who you were and only being the tuner until after your reply. My comment had nothing to do with that, just in response to your own vocal commentary against Garrett.
So it is what it is, yet you continue commenting a novel length reply on something you also claim to be off topic. I get your feelings and all, but it will just be better to let it go and move on. I shouldn’t need to say it again, but will; I recant from my own opinions and accept your explanation. The reality is that outside of you and me, nobody else here probably gives a toot about it.
Franky, you arguing against an actual wheel speed readout would make me question you more than anything else. Because I didn’t figure out who you were and only being the tuner until after your reply. My comment had nothing to do with that, just in response to your own vocal commentary against Garrett.
I'll keep it short for you this time. Below is what I said. Also, my first post said that I was the one that ran the test; first sentence actually.
Originally Posted by yamahaSHO
The car that had the G25-660 is getting an EFR 7163 and a speed sensor, which I would have liked on the previous turbo, but I don't own the car. Being on a stock ECU with Cobb, we won't have the ability to use the input for any safeties, just data to view and log.
Originally Posted by TeamRX8
So it is what it is, yet you continue commenting a novel length reply on something you also claim to be off topic. I get your feelings and all, but it will just be better to let it go and move on. I shouldn’t need to say it again, but will; I recant from my own opinions and accept your explanation. The reality is that outside of you and me, nobody else here probably gives a toot about it.
As I have also said, I will address things and I like to be pretty thorough. I know plenty of people who care as we've had many discussions on how they/we want to do builds based on what has been learned. I save all my data from any and every car I have ever tuned, and I liked to be able to go back and reference. My intent her was to help pass any information that someone might be interested in... Oh well.
Fair enough; I again concede to all your points and also regrettably my being a bonehead in how and where it was addressed; that’s entirely on me. It will be interesting to see the results if someone tries the 1.0 A/R housing over there.
Kind of assuming you understand the difference in flow vs whp requirement between a rotary and reciprocating piston engine? As this thread demonstrates, even an EFR7670 hits a low-mid 50ish lb/min compressor flow wall between the combination of waning compressor efficiency and turbine flow limitation using the largest 1.05 A/R housing that BW offered. Of course this same 1.25 housing can now be fitted to it as well. Yet even if it can reach say 60 lb/min, which equates to ~450 whp on a 13B rotary, it’s operating at the limit and all the fire-breathing it generates there.
I’m not so sure that going to SX-E cover on a 7670 will provide as much gain as it does on the 7163, but I suppose somebody will need to prove it out. All that’s intended and can really be hoped for with a 7163 is a super zippy, more compact alternative with the same basic limitation. Within the EFR range, the potential will shift to the new 8370 with one of these higher flowing aftermarket turbine housings once BW can get past the manufacturing issues and actually supply the product.
Which is a bit funny because, and I’m not so comfortable putting this “O” word out in front of you again, they claim to have combined the 83 compressor with the 70 turbine to limit Overspeeding issues. Even it wanes on compressor efficiency out to the far right and in general relative to where a 13B will operate in the 2.0 - 3.0 Pr range, but should at least get in the mid-upper 60ish lb/min range needed for ~500 whp … “with a turbine housing that can support the flow requirement”.
can’t even get to the center efficiency island below 2.1 Pr at that.
.
Probably one of the last of the EFR7163 posts by me here in this 7670 thread, but there was a prior discussion about the BOV that seems disproven in my view. At least on the compressor covers for this particular EFR turbo.
I ordered the SX-E cover from Hayward, CA on Saturday afternoon with standard shipping from the only place that seems to have then in stock and somehow received it Monday afternoon in Dallas, TX; must be a USPS record or something.
Its a bit interesting. I can see that the scroll appears larger externally and am assuming this transfers over to it being bigger internally. Yet they more or less kept it within the same basic profile. Stacking them directly centered on top of each other they are close to identical externally. When matched up on the same mounting surface the SX-E cover appears to be slightly taller, but not much. When the outlets are lined up to each other the SX-E cover outlet is about 0.25” taller (wider to the inlet side) or so. Yet the end of the inlet is about 0.37” shorter (just eyeballing it with a straight-edge across them). So it really doesn’t seem to take up any more space like I was expecting it to.
Supposedly a larger A/R, but no markings on any of them. It actually takes up less space without the integral BOV and solenoid valve, though the SX-E inlet is 3.5” dia. vs. 2.5” dia. for the standard EFR cover. So a larger CAI supply tube is required, but would have done that anyway with 4” and just reduced it at the inlet.
After looking at how the BOV is configured in the cover I’d conclude that that the openings in the outer scroll and inlet disrupt airflow relative to not having them in the SX-E cover, and certainly don’t contribute any benefit to keeping the the turbo spooled any better, as was suggested by someone in an earlier conversation. It just dumps the recirc air straight in from the side at about the center length position of the inlet perpendicular to flow and imo only serves as a convenience over an external BOV. Especially if it needs to be recirculated back to the inlet if a MAF is used rather than MAP.
....
In addition to no BOV ports it doesnt have the anti-surge compressor inlet which also increases peak compressor flow.
I thought the "anti-surge" also increases max flow. So the map gets wider to the "left" and "right".
At high flow speeds it catches additional air through the "anti-surge" ports...bernoulli effects and so..
Anti surge slot broadens the Left side of the compressor map at the expense of a very small amount of peak flow.
It is a popular mod to delete the slot with a machined ring to gain power if your application isnt pushing the compressor into surge.
Just having a smooth less turbulent entry into the turbo inlet will improve both total compressor flow and spool.
Yes, may seem counter intuitive but the anti-surge slot hurts spool a bit. Its there to keep surge at bay on a compressor that is already spooling too hard for the engine.
Large as possible inlet piping, straight as possiblw or the correct direction bend (pre-counter rotation of air column), a velocity stack and as large a filter (or none) that can be fit also play a big rioe in spool. Especially on the very low rpms in my experience (data logging before and after).
Some of my responses are regrettable, but no excuses; I’ll try to do better.
Honestly I didn’t think to look for one of the Indy covers. Those turbos in general seem to have all dried up, but I haven’t looked around in a while and maybe not in the right places either. Other than not having the anti-surge, it appears to be pretty similar.
I’d rather not have the anti-surge for that reason, and between the usual BW casting and machining sloppiness it’s simply not the best job at accomplishing it either. At the same time it might be necessary on a 13B. I can see it potentially coming on sooner at low rpm than a 7670.
The compact simplicity of how it’s now configured is nice. With the SX-E cover, aluminum CHRA, and cast iron turbine housing it comes in at just under 16 lbs total on a good scale.
Anti surge slot broadens the Left side of the compressor map at the expense of a very small amount of peak flow.
It is a popular mod to delete the slot with a machined ring to gain power if your application isnt pushing the compressor into surge.
Just having a smooth less turbulent entry into the turbo inlet will improve both total compressor flow and spool.
Yes, may seem counter intuitive but the anti-surge slot hurts spool a bit. Its there to keep surge at bay on a compressor that is already spooling too hard for the engine.
Large as possible inlet piping, straight as possiblw or the correct direction bend (pre-counter rotation of air column), a velocity stack and as large a filter (or none) that can be fit also play a big rioe in spool. Especially on the very low rpms in my experience (data logging before and after).
That it's bad for spool seems logical to me, with the theory that it catches additional air only at high flow speeds, it might not at low flow speeds, and even air can escape through the ports.
In university they told us that it behaves like this, but that was also 15 years ago...
With my 8374 I had surge when getting out of the corner, being like half on the throttle and already boosting like 12psi at 3500.
I fixed this by making the boost control be affected by the TPS signal, like half throttle -> half boost control duty
throwing this put as a lower cost 450-500 whp option:
Received the hybrid “S262 SX-E”; a Borg Warner S257-SXE converted to the S362 SX-E compressor wheel along with a 1.22 A/R divided T4 housing, except it’s a T4 flange with T3 ports. There was nothing on the turbo specific page indicating this. In the turbine housing only section it states this along with offering to port it put to full T4 ports for an additional charge. So waiting on them to respond about getting it swapped out with the full T4 ported housing. I see BW lists it in their documentation though.
So the S362 compressor wheel in the S200 cover is supposedly good for low-mid 70-ish lb/min flow; which theoretically is low-mid 500 whp on a 13B and the 70mm turbine wheel with 1.22 A/R housing is on par with an EFR8374 0.92 A/R turbine flow to support that (~28.5 lb/min), yet be responsive still at low rpm. Looking forward to see how it pans out overall and in comparison to the G30-770.
The compressor is on par with the EFR7163 SX-E cover I posted elsewhere, but the turbine housing is quite a bit larger and heavier. The full assembly weighs just under 24.5 lbs total, or about 50% more than the hybrid EFR7163 assembly I’m working with on another project.
I have the IRL 6758s on my LS and there is no info of parts for the Indy-specific compressor inlets. It's also a stepped adapter, so not even as straight forward as getting a 3-bolt flange made. I ended up just welding 2.5" AL tubing to the inlets to extend them forward where I needed.
The IRL turbos are incredibly light, though. The investment cast v-band housings, aluminum CHRA, and slim comp cover weigh 11lbs total IIRC.
Not all caught up on the thread, pardon me if covered before... Custom build turbo or purchased though a place that already did the configuration? Would love to test this configuration vs my S362!
throwing this put as a lower cost 450-500 whp option:
Received the hybrid “S262 SX-E”; a Borg Warner S257-SXE converted to the S362 SX-E compressor wheel along with a 1.22 A/R divided T4 housing, except it’s a T4 flange with T3 ports. There was nothing on the turbo specific page indicating this. In the turbine housing only section it states this along with offering to port it put to full T4 ports for an additional charge. So waiting on them to respond about getting it swapped out with the full T4 ported housing. I see BW lists it in their documentation though.
So the S362 compressor wheel in the S200 cover is supposedly good for low-mid 70-ish lb/min flow; which theoretically is low-mid 500 whp on a 13B and the 70mm turbine wheel with 1.22 A/R housing is on par with an EFR8374 0.92 A/R turbine flow to support that (~28.5 lb/min), yet be responsive still at low rpm. Looking forward to see how it pans out overall and in comparison to the G30-770.
The compressor is on par with the EFR7163 SX-E cover I posted elsewhere, but the turbine housing is quite a bit larger and heavier. The full assembly weighs just under 24.5 lbs total, or about 50% more than the hybrid EFR7163 assembly I’m working with on another project.
.
what can I say; it blows not having a proper shop to work in. Finally got my generator running and the turbine housing T3 inlet ports above ground out to the T4 port size. It turned out the that the internal passages open up and expand larger an inch or two in past the inlet ports.
Sort of a half-hack job using what little I had to work with and the conditions, but it will suffice. The exhaust gasses flow will definitely transition better into the turbine with the T4 ports. It doesn’t really make sense for BW to not be casting T4 ports in the first place imo, at least not with the larger A/R housings.
Just need to get this cleaned up and the car back together after a few other things are also worked out (RX8-REW), verifying the tune/revising as necessary, and get it onto a dyno.
This maybe demonstrates better how shrouded the internal flow path of each divided housing side is from T3 port size. Because the housing widens considerably after the flange. The outer edges of the T4 port are as wide as the outer housing surface just past the flange. So the outer T4 port edges can’t just be ground/ported out straight, but have to be radius transitioned some to keep from breaking through or making the housing wall thickness too thin. The bottom picture doesn’t really do justice compared to looking at it in person.
took the SXE8370 hybrid out for the first initial spin the other day. Considering it’s on a BP’d junk yard engine that’s less than ideal for this turbo, it felt pretty good rushing up to 20 psig on a short run or two
hopeful of getting it on a dyno in the not too distant future for some data/results.
02-07-22 | 11:17 AM
