Given that you don't know what your talking about, s3 turbo is fine, the other turbo let go at 26psi, that being said sometimes turbos just let go, had nothing to do with anything. And nobody said I knew how the turbo failed so not sure where your getting you false info and all your comments were speculation, all I said was I have tuned 2 and 1 failed and 1 was in an S3 and that it ran 11.5
I have tuned 2 pmax 2260 turbos, the one thing they both had in common was boost control took quite a bit of work getting under control, more than most turbos, you have to treat these turbos like an is38 as far as the boost curve and what not. Out of the two, one of them failed, the comp wheel nut literally came off, the other one runs good in a full weight S3 with a 290lb driver and 180lb passenger did 11.5@120 on the street. Treat them like an is38 and they will do good, as hoon said, the spool is good and so is response but imo should not carry the Garrett name.
Compressor nut departing the shaft isn't something that 'just happens,' it's indicative of a quality problem if operating within parameters. Forgive me for not knowing what I'm talking about, I'm asking these questions because it was impossible to get the most relevant information out of ITGUY; who tuned the car. Since you tuned it, I'll ask: did you bother to install the shaft speed sensor to see what kind of RPM the turbo was at when making the requested boost in your tune?
26psi would put you at a pressure ratio of ~2.8 with no system depression. Depending on the losses in the system, the PR could approach 3. ITGUY's failure occurred in "3rd hitting 6k," but prior excursions to redline could've contributed if they caused an overspeed. At 6800rpm cramming 26psi of boost into our 122cu in engine, we're pushing just about 45lbs/min. Maybe it's a little less or little more depending on density altitude, ambient temperature, and intercooler efficiency. Likewise, a real world system loss would raise the pressure ratio up to ~3, and there's a good chance this tuned engine may rev to 7000rpm. Of course, this is all just theoretical.
45.05=((14.7+26)*(0.99)*(6800/2)*122)/(639.6*(460+120))
I'm not accusing anybody of anything, but it is at least theoretically possible to overspeed this turbo with 26psi of boost. If you've put more than 43lbs/min through this turbo at any pressure ratio, you've oversped it. We should expect some safety margin, but given all the unknown unknowns, and your own admission that boost control is difficult with these turbos, it's not hard to see potential for inadvertent operation outside the allowed envelope, especially during untested circumstances. In particular, Garrett's advertised output numbers for this turbo are right at the very edge of its operating range.
I will reserve judgement until we see OTS tunes. Garrett is in the unique, disadvantageous position of selling a bolt on turbo for this car without offering its own tuning solution. It will take some time for people to figure out how to run this thing safely and efficiently.