it goes to high for a couple of milliseconds when I turn the key to off position because I have an apexi turbo timer

 

The differential is whistling. Again.
It's like the fifth in a row, this time it lasted 16 days

Any advice?
I think I have two choices:
-overbuild with a stronger differential but without the ppf
-lower the pressure
-change the turbine with something bigger (less torque)
-sell the car

Since option #2 is not possible what would you do?

 

Find out what you are doing wrong that is destroying the diffs.

And it's not just sheer torque.

Bundy just posted about how his is still fine and he blew up 15 transmissions from pure torque and super hard track use, so I highly doubt your failures are from just too much power.

Either something is mis-aligned, or they keep getting put together wrong (if even taken apart, that is), or bad fluid, or something.

 

Excessive wheel hop?

 

This disntigrates diffs, not make em whistle. (Ask me how I know )

 

I'm guessing incorrectly set tolerances or it's angled too much or janky fluid.

Just guesses

Shock loading will violently shear things, not "wear it out" like OP seems to be experiencing

 

Agreed. Something isn't right here. It's not just power or whatever.

 

I'm honestly demotivated.
I done what I could, but the differential assembly isn't a thing that I want made myself, and this time I want to be sure is the last one.
What if I give the differential to another machine shop and it brokes again? I don't want this happen.
The only thing that I can do is unbolt the old unit and put in another one. It's pretty much straightforward, I don't believe that is my fault... I even torqued everything to mazda specs...
The last time I reinstalled the oem suspensions, trying to exclude even the uncorrect shafts angle (it was too low)

I honestly don't know where to smash my skull

 

oh, and by the way, holding the brake in sixth gear I can reach 7psi (0.5 bar) since 2500 rpm and 21 psi (1.5 bar) since 3000 rpm.
Yeah, yesterday I was bored during a 400km trip

 

Can't blame you there -- this would be seriously demotivating.

Would you please summarize your diff adventures into one post? Lots of info spread over several pages.

Tell us what diffs broke, when, and how you fixed/replaced them.

 

first: mazda oem unit 3.636 + torsen, never opened.
Good until I tuned to 21 psi and it started to whistle. (from here I still ran 21 psi)

second: mazda oem 4.2 open, never opened. Two days and the left shaft escaped about 10mm out from the diff, letting the oil exit from the seal: wasted the third day. It was installed perfectly, I don't know why happened.
Bought new elastic rings for the shafts, oem of course. Raised the car with the oem suspension to exclude the excessive shaft angle.

Third: my first case with torsen, new r&p, new bearings, new seals and new crush sleeve.
One day after installation started to whistle.

Fourth: same differential, same case, just adjusted the tolerances (local machine shop)
lasted around six hours and started to whistle again

Fifth: same case, new r&p. The bearings had less than 250 km, we leaved them. This time I sent the case to another machine shop far away (very good reputation here).
It lasted 16 days. Now it's whistling again.

This saturday I will dyno on a mustang dyno, just to see how much torque I'm putting down. Sure enough I'm under 300hp, but the torque is real

 

Third, 4th and 5th COULD be errors in adjustments or parts. Second lost its oil, so it makes sense why it went pop. First diff... how was it a 3.636 but never opened?

I think it's worth a try to snag a low-ish mile, OEM Torsen that's truly never been opened and give it another shot.

But a Getrag upgrade is balleur.

 

Aussie mates.

I agree. Getrag

 

the first was the original of the car. Mine is (was) a "miracle", selled with 1.8, 6 speed and 3.636 + torsen

 

Did you put it together with new bearings and a new "distance piece" or "crush sleeve"? If the bearings are worn it could make the diff not keep its tolerances.

 

yes, the last one had everything new inside: all the bearings and new crush sleeve (oem mazda replacement)

I want to try the last time. Oem 4.1 open diff with just 40.000 km on it. It's the less used one I found

It seems to be in great shape:

 

Are you taking the correct measurement on the end caps with a 7-8inch micrometer? Then setting your backlash or are you setting your backlash and then taking the cross measurement?

 

I've never done a rebuild, but every machine shop I went for has the mellens printed from me, so mazda specs and assembly instructions

 

That "1 bar" sensor it's actually the mpx4250 (0-250 kPa) , not the 0-110 kPa one