Icedawg

Of the many broken 5-speeds I have taken apart for parts, the gear teeth strip in pairs, main and countershaft together.
There is always too much damage to sleuth out if the main or counter gear failed first.
But note the SM twisting problem is on the countershaft, which sits in the oil.

In non turbo Miata and rotary RX7 racing ~80% of failures occur when shifting, 20% randomly when in gear. Third gear strips teeth about 70% of the time, and 4th the rest.
I think the lower % failure modes are fatigue, while the higher % are from marginal mismatches in engagement during shifting, which instantly strips the teeth.

Mis-match could come from load induced twisting of the casing, harmonic vibration induced mismatching during shifting, worn bearing induced twisting along the main shaft, and over heating induced mismatch. Front and rear main shaft bearings have more wobble and play in broken 5-speeds than new ones; within spec for a street car, but maybe not for a race car. The "pointy teeth" on the sleeve of the clutch hub shifter assembly and the engagement points on the gear can sometimes be worn, but the greatest wear is on the shifter fork where it sits in the groove in the sleeve, usually well beyond spec in a used stock street transmission. That allows a lot of rocking and twisting of the sleeve during a gear shift, and that could easily lead to gear mismatch on shifting.

A gentle shifting technique, no slamming, with the briefest pause in the neutral position extends the racing life in normally aspirated cars a lot, and I think the potential for cocking the shifter sleeve in a slammed shift has a lot to do with that, as does letting stress on the casing relax during that pause.

But I suspect an oil cooler would also greatly reduce the rate for the shifting induced failure mode with NA engines. And why else would Emilio have thought it was worth building one for their endurance racer? But it probably would not help as much with the fatigue failure mode, or a turbo car where raw power simply overwhelms the 5-speed gear strength.

emilio700

iTrader: (3)

Our implementation of a trans cooler was an attempt to keep the aluminum shift forks in the AZ6 from expanding and binding. What kills 5 speeds is more minor manufacturing flaws, insufficient internal lubrication, casing flex and inadequate strength materials. Pretty common for boosted AZ6 transmissions to bind up and refuse to shift when hot but OK after they cool down. Aluminum has roughly 2x the thermal expansion rate as steel. Something that Mazda and Aisin didn't care about as the box was never intended to see those temps with 120whp. The trans cooler helped the AZ6 issues we were having, 160° vs 270°

sonofthehill

iTrader: (1)

In my experience, granted I have only broken 3 5 speeds, 4th gear has no teeth to strip. I was always able to use 4th gear in a broken 5 speed since 4th locks the input and output shafts together. All my transmissions broke while in gear, not from hard shifting. The typical failure mode every time, where the teeth on the input shaft break and power can no longer be transmitted to the counter shaft. Might have an old counter shaft in the garage somewhere, I will have a look later today, don't recall about the 5th spacer, wasn't really relevant with a broken input shaft.

codrus

Yeah, that was my experience with the 5-speed. Downshifted to 3rd for turn 11, went through the corner, fed the throttle in and as soon as the boost came up ZING all the teeth came off.

--Ian

bbundy

Here it is again how do you explain this gear teeth stripping in 4th on the 5 speed? Besides zero load going through the secondary shaft in 4th gear there is also zero load going through any gear set in the transmission in 4th. selecting 4th gear is literally just a selector ring locking the input and output shaft together like there is no transmission at all. Power loss and heat build should also be minimal in 4th. Ive driven my car home more than once with only 1st and 4th working.

Icedawg

how do you explain this gear teeth stripping in 4th on the 5 speed? ^^

Well ... it simply happens. And when it does, of course, no other gear works.

4th can fatigue, chip and crack just like any other gear. The same worn selector sleeve and shift fork that engages 3rd also engages 4th. But in my experience when 4th fails with an NA engine in a race car (SM or RX7) it most often simply lets go while driving, not in conjunction with a shift. So metal fatigue or worn bearings are the most likely explanation. I have seen badly worn roller bearings on the tip that goes from the main shaft into the 4th gear input shaft, which causes some wobble. And some input shaft 4th gears that simply wobble, probably because they have worn against an aging roller bearing. And no question the input shaft bearing wears and gets a lot of play, also allowing wobble. But not all stripped 4th gears have wobbly bearings. All out sprint racing is hard on everything! Mazda did not design for it.

sonofthehill

iTrader: (1)

So are you saying that the teeth that engage the selector ring break? I haven't seen this before, even so why would that prevent 1st, 2nd and 5th, reverse from working?

bbundy

So what you are saying is the 3-4 selector breaks. There is no gears involved with 4th. If you are saying the input gear breaks then yea Ive done that multiple times when that happens you loose every gear but 4th which still works. I've broken over 20 transmissions 5 and 6 speeds and have never broken one on a shift only after it is in gear and torque is applied either shear teeth off or some form of thrust force failure. In 4th there is no gear teeth or thrust loads from the helix angle on gear teeth though. Only other odd failure I have had a Quaife gear set weld 5th gear to the main shaft once this happened while doing a burnout preparing to stage at Prosolo. Third time I broke teeth off the input gear on the Quaife gear set two times and second gear once before I gave up on using it.

For what its worth the first picture in this post shows a failed input gear with the teeth stripped off. that gear can fail with the transmission in any gear besides 4th. When you shft to 4th there is no load going through that gear set or any gear set in the transmission. It goes straight down the mainshaft. 1 to 1.

sonofthehill

iTrader: (1)

That must be it, I clearly don't understand.

Midtenn

iTrader: (15)

I wonder how much of this is to reinforce the box for bump drafting. Since they are allowed to reinforce the diff housing, that impact load has moved further up the system.

jedd

I bought a brand new 5 speed from mazda for my spec miata and it came with oil in it. I put it in and ran a test session (was gonna put redline mtl in it after that session) and it failed going into fourth near the end of the 20 min session. Still wondering what fluid it shipped with, that was a stupid idea. They sent me a new one free of charge and required the original back to tear apart. Mazdas great! I dont know what failed but id say that failure was fluid/heat related to fail that quickly in a 125hp car.

emilio700

iTrader: (3)

Transmissions do benefit from being broken in.

LeoNA

I’m running a new 5spd from Mazda with a turbo. I have 3 track days and 8k street miles on it so far. I drained the factory oil out and ran it lightly for 500mi with mtl. Then switched to a 50/50 mix of mt90 and lt shock proof. It’s holding out but I do think it will die eventually. I change the fluid before every track day.