Tranny Upgrade Parts

[Custom_V6_Limited_SE]

I'm wondering where I could get upgrade parts for my GS tranny to handle large amounts of horsepower and how much they would cost.

[Yoda]

You would need to find a tranny used with the 2.0L RF-T in the 323 made after Feb 2002 to Sept 2003. Mazda really beefed up the internals of the G series to handle the torque of this engine. This tranny is useless as is because the bell is case for a 4 cylinder so you will have to mix and match the stronger 4 cyl second section of the gear case with the weaker V-6 first section and bolt on a gridle kit to keep the older V-6 half of the case from flexing and crapping out the tranny.

The ultimate FWD tranny for handling a built engine would be to some how covert to the new Ford FWD ATX with the beefed up Motorsport parts This things are taking over 500ft/lb of torque without fail. GM also has a FWD ATX that will take over 500ft/lbs in near stock condition.

[Custom_V6_Limited_SE]

I wanna use a tranny that will handle 500hp just in case I actually end up boosting a built engine that high. How plausible would it be to try to use those Ford or GM trannys?

[PATDIESEL]

Not very. They wouldn't be close to the G Series most likely.

You could also take the gear sets out and have them hardened or get a custom one built from either Quaife or Tri-Point Mazda. I cannot remember which one builds them, but one of the companies will make a gear set with custom ratios. The cost is about 4000 (if I remember correctly), but that is the end all if you want max performance.

[Custom_V6_Limited_SE]

How much power would it take to break those gears?

[JWMotorsports]

The actual gear set is milled from solid billets if it is the ones I'm thinking of. So they'll hold about anything you can throw at it as long as you keep the bellhousing from twisting. If the bellhousing twists then your lash tolerances become severly skewed resulting in broke gear teeth or bent/broken input n output shafts. We are currently looking into to these problems our selves as my car will be DESTROYING trannys shortly w/ 35psi boost and Nitrous on top of it. We are beliveing the gear n shaft damages are a result of the tranny twisting with the engine block more than anything, with the exception of weak OEM differentials...at least for the 4cyl. Pro-LX boxes. This reasoning derrived from the DSM world's experiances. Under high HP the engine blocks actually twist which on DSMs is so bad they spit the freeze plugs out. Looking at the engine mounting positions all the torsional loads are being placed on the middle diameter of the bellhousing and the only torsional support the engine has is the single mount on the driver side for the 4cyl engines (I'm assuming the V6's are similar?). This puts the input shaft in a highly torsionally loaded position which can skew your gear lash clearances causing gear and or input/output shaft failure.

Yoda...what is your take on this?

[Yoda]

I have to agree there is a issue with torsional rigidity. The two primary functions of any gear case are to rigidly mount the bearings of any gear shafts, such is that the gears maintain accurate meshing under the parting loads induced as they transmit torque, to keep dirt out and to keep the lubricating oil in. The weakest link of all the Mazda FWD trannys is the gearbox casting. Like everything else days the casting is made as thin as possible to save weight. The casting technology at that time the gearboxes in the MX-3 were produced is based on early 80's technology which leaves the possibility of porosity and variable wall thickness, due to problems of positioning the cores accurately during the casting process. Todays lost wax technology produces a much better casting. I was looking at one of the spare G-series trannies in my collection the other day and comparing it to a Ford MTX-75 (Mazda G35M-R). The MTX-25 casting is much stronger where is counts due to the lost wax technology allows for its design includes a lot more ribs and gussets designed into the casting to triangulate key points like around the bearing cups for the primary and secondary shafts and mounting flange to the block. Heat is another issue the operating temps can be anywhere from 100C to 130C. This heat is generated by the gears but it is also transferred by the exhaust and air flow threw the radiator. Temps over 100C can cause the gear case to stretch/ flex more that when it is cold. This as a lot to do with the expansion rate of aluminium compared to other materials such as steel which is more dimensionally stable. When mount a turbo over the tranny and routing the exhaust next to and under the tranny is actually contributing to the failure of the gears compared to an engine making the same power normally aspirated. Those that have turbo charged a V-6 have the deck stacked against them right away. The couple of gearboxes I have opened up it appears that the gears have single and multiple impact damage which indicates that gear mesh below the PCD causing the teeth to bind and other gears show signs of "fold over" which means there is to much backlash which indicates worn teeth, lack of lubrication or a shift of the relationship of the shafts with in the gear case.

Looking at the OE gears, the suggestion to harden them will not make the gears any stronger and the guys out in our gear shop say it would it would actually have the opposite result. The problem is that not all steel alloys are capable of being harden. The gears are of a cast design then carburized (case harden). Trying to hardening the gears again will result in having them fracture when put under a hard load due to the alloy used in the gear casting and you will varying levels of hardness because of the previous hardening processes. The other thing is the hardening process caused distortion in the metal which means the gear will have to be rectified on a gear grinder with the end result being the gears will now be under pitch diameter when will cause even more issues. In some cases it is necessary to have one gear harder that the other to compensate for wear over a long period of time. One such case in the differential. The pinion gear is made harder than the ring gear.

I do agree that a turned gear from a solid billet can be much stronger but again it over all strength is doing to be determined by the alloy it is made of. An alloy like 9310 would give you one of the strongest gear that can be made without going into the really exotic metals. It also appears that the gear pitch used in the tranny was chosen more for quietness of operation rather maximum power transmission. A courser pitch gear would allow for more power to be transferred at the expense of gear noise. Changing to a spur gear may also be an idea as well Technically a spur gear is not as strong as a helical gear like the stock gears but most racing trannys use spur gears because all the power is being transferred in the rotational direction and gear noise is not and issue. Helical gears like the stock gears transfer the power not only in the rotational direction but also in a angular direction as well. Going into detail here is probably beyond the scope of this forum.

There is a slight difference in the way the engine mount attaches to the from of the block between a 4 and a 6. On the 6 the engine mount is just above the thrust line of the crank and attached where the block is the strongest and being a V it is inherently dimensionally stable on the other have with the 4's Mazda sort of took a step backward from the older generations of 323. The engine mount was a yoke that attached to both sides of the engine and wasn't offset as it is now to that attachment point on the frame. The older SOHC had either some pretty beefy bolt on braces or a cast girdle that tied the bottom of the tranny below the adaptor plate to the sides of the block which distributed the torsional forces over a greater area, stiffening the engine/ tranny assy as a whole. The DOHC's lost the external bracing in favour of a bolt on extension to the oil pan skirt or a flange cast in the oilpan. Due to the way these attach to the engine it really doesn't offer that much additional torsional rigidity. With the B series engines another problem is the tranny is really just bolted to an adaptor plate attached to the block. The strongest part of this is where the plate is sandwiched between the block at the top and tranny the sides of the tranny are only attached to the steel plate which is bolted to ears cast to the sides of the block which is sort of allows the tranny to flex off the back of the block.

Making the stock tranny handle more torque that when it was originally manufactured is going to require more that just making the gears harder or stronger. The real issue is more likely to be heat and the casing of the gearbox. It seemed that Mazda beefed you all there tranny in the 2002 model years so as much as many of you don't want to here this but a new F series tranny used in a 2002/2003 Protégé is probably now as strong as a G series tranny made when the MX-3 was still in production.

[Yoda]

I was given a factory G25M-R manual recently. Just notices that there are no thrust bearing in the design. Helical gear transmit the power not only in the direction of rotation but also in a angular direction perpendicular to the the angle of the teeth. This mean that the gear are acutally using some of the engines power at the crank to push the gears sideways The gear on the primanry shaft and secondary shaft are angled in opposite directions so one shaft wants to go left and the other wants to go right. Nornally this would mean that a thrust bearing would be needed on one side of the gear to keep the gears from walking sideways and reduce power losses. The bearings on the Primary shaft are a taper roller bearing so it can take some of the angular thrust but the bearings on the secondary shaft appears to by only on radial bearing which are not designed to take a lateral load. Under stock power levels this is probably acceptable for the service life of the average car. But if the engine torque is increased this will increase the loading on the bearings. Elevating the under hood temps with a turbo increase the temp in the tranny allowing the tranny case to stretch and break down the lubricating film on the bearings and gear teeth.

[JWMotorsports]

Thats what we were thinking. Currently the best luck we've had so far is using Lucas 80W90 Gear Oil with a bottle of Lucas Trans Fix and the gear box in my car has amazingly taken 24-26psi on a Big 30R turbo for over 10,000 miles. I know if you go to thick on the fluid it will generate more heat due to the drag created by an overly thick film. It never dawned on me that the tanny would need thrust bearings in addition to the tapered roller bearings. I guess with all the tolerance stack up under heat they would change drastically, but doesn't heat make them expand closing up the excessive end play? I was thinking that the inner case that 5th and reverse sit on acted as a trust washer with the taper bearings it houses? If so then additional spacers can be swapped out under the taper bearings to adjust end play? I'm assuming the only way to gauge that clearance properly is to plastigauge it with the input and out put shaft nuts torqued back to spec, then take it back apart to read the plastigauge and only able to do this one clearance at a time? I agree absolutely about not hardening the gears because it would make them extremely brittle, as you said that material is being pretty well pushed to its limits. If you were to take the stock gears out and remove .003"-.005" off the gear faces (total lash opened up .006"-.010") and go to and extremely heavy gear oil like a 140 weight, would that possibly be a simi fix next to replacing them with dog gears or is there not enough gear face lash to allow such modifications? I guess the ultimate would be a 6 or 7 speed dog box utilizing Grade 5 Titanium dog gears & shafts, with a Carbon Fibre housing covering a Grade 5 Titanium tube n plate sub structure with ceramic ball bearings and thrust washers. Then utilize a front mount oil cooler and pump with Lucas 80w90 Gear Oil or Red Line 80w90 Gear Oil. Also utilize a Kaaz Limited Slip setup. I'm thinking about adding a pump and front mount oil cooler to my stock tranny housing due to the NDRA Street Class rules. It states that you must use the stock tranny housing, but....ANYTHING that fits inside it is legal to use!

[PATDIESEL]

Could I take a minute of your time JW or Vaughn to see if I am understanding all this correctly.
(PS, I love a thread like this. It is not often we see an engaging thread anymore and I love to learn new things)

So the basics are this: (correct me if and where I'm wrong please, I'm sure I missed something)
This also ignores all the info Vaughn gave on the actual structural design of the tranny and its mounting positions per motor)

A. Issues:
1. the case is not heat tolerant due to poor casting methods, causing differentiating levels of thickness and porous metal. Also, it lacks sufficient webbing to displace stress on certain areas.
2. the gear sets are not firmly planted with proper bearings to allow for high torque.
3. The gears are not strong from the factory due to weaker metals and older heat treating tech.

B. Problems caused from Issues:
1. the heat builds up, the tranny casing starts to get soft and varies torsional stregnth due to different thicknesses
2. the softness allows the bearings to move from the natural helical forces placed on the shafts
3. the shafts move from lack of stregnth in the bearings and flex from casing. Then they are out of the designed positions and place undesigned forces on the gear teeth faces
4. the gears hardness cannot withstand the friction and load and will either wear down or fold

This is all about the gears and nothing about the diff. (a whole nother problem)

So, I guess if I'm am correct or atleast close, my question is this. At what heat and torque are we talking about? I'm sure a BP at 30 PSI is a ton of power, but the limited runs of a drag shouldn't yield too much heat. (or does it, I don't follow drag racing much) Meaing that other than the torsional issues of the bearing design the heating and flex is not an issue. Or is it from the lack of webbing and support from the frame or motor?
I'm more concerned with the sustained build-up of heat from a track raced car (like mine). I could use a trans temp sensor to tell me what temps I'm getting on the track. This would allow me to know what kind of trouble I'm looking at if I knew what the danger levels are. Then in order to render the problem fixed I'd need to take several measures to reduce the A. heat and B. torsional flex of the casing.

The fixes discussed are:
1. Grind the gear teeth faces to open the lash a few 100ths and run a thicker dope and perhaps a trans fluid pump and cooler.
2. get a new gear set and somehow machine the case to fit the proper bearings.
3. use straight-cut gears to remove the helical gear force issue and hope that the torsional regidity of the case is sufficient to hold it in place.
4. build a F-1 style gear box with custom titanium gears and CF covered titanium box with substructure that will cost more than I make in 10 years let alone finding someone to make it (PS, how does this not destroy your fitting into class ) Wouldn't it be easier to buy a 20,000 dollar Quaife gear box?

Oh, just a note. I run a V6 and G series (of course).

[Flinch]

A ford escort GT tranny should fit on there with no problems and its good to about 475hp

[rebel2k4]

Very good info guys. Funny because I was just talking with my machinist about transmission solutions.

We basically came to the same conclusion after talking about mindiola trannies for dune buggies. It is essentially nothing more than stock VW trnas with a beefed up final drive, but used stock gears with a different case. The primary difference in the case is more gusseting. This is, of course, for their lowest model upgrade.

So, Ive been thinking about making aluminum braces out of 1/4" plate and having them welded to the case to increase the rigidness and perhaps prevent gear failure. I have broken 3rd gear twice now, even with street tires and a relatively soft engaging clutch.

What do you guys think? And once again, GREAT info!!

[Yoda]

We have to remember when the F & G trannies were designed and the technology of the day. They are both products of the late 70's early 80's.Sand casting was the primary technology of the day. A production tranny is a catch 22. It has the be quiet and smooth in operation, strong enough to handle the peak torque of the engine not add a lot of weight to the car and be cheap to produce.

Actually the heat issue has no bearing on the casting technique use. That is more of a design. Given that cars at the time made at most around 150-160ft/lbs of torque the engineers probably made a trade off case stiffness in favour of weight savings. The 626 at the time was smaller and lighter than a Protégé or M3 today. The temperature of the case never get even close to the annealing point of aluminum so the case doesn't become soft. Aluminum does have a greater expansion rate that steel so the case can "grow" as it becomes hotter. Thermal expansion and lack of reinforcing webbing can cause temporary core shifts. Add this to the torque of the rotation of the crank and the tranny is more securely mounted to the frame of the car that the engine. The other issue of the "growth" of the case is that the steel shafts don't expand at the same rate as the case so it is quite likely that shaft end play is constantly changing. I believe the critical temp is somewhere around 100-120C. You should be aware that much of this heat is radiated heat either from the exhaust or airflow through the rad and internally generated heat. I have known a few people that have ceramic coated there gearbox to keep the radiated heat out with success.The installation of a low pressure pump and cooler is not a new idea it and been used in Mazda touring cars for at least 20 years. Webbing cast into the sides of the case has another advantage as well in that increases the surface area for radiate the heat away.

It doesn't appear that the case can be machined to accept a different bearing the case has barely enough material to hold the bearings now. Also adding a thrust bearing/ washer is also problematic. Looking through the bearing catalogs this morning I see that there are taper roller thrust bearings that can do the job of a radial bearing as well. One day when I have time I'll have to look at the specs of these bearing and see if it is possible to replace the OE bearings with these.

Increasing backlash opens up a whole new can of worms and will lead to even more tooth wear of breakage. The gears should always make contact at the pitch circle diameter where the tooth can transmit the greatest amount of power. Under cutting the teeth will only move the contact point on to the addendum cause the crown to fold over in the drive direction. Also since the teeth of the gears are shallow case harden regrinding the gears may make the depth of the harden surface very shallow. As far as OE gear material and hardness go it is correct for the intended purpose of the original design of the tranny. Heat treating technology has not changed much since before world war II. Not all alloy can be harden and those that can some can only be carburized (caseharden/ surface hardened) As far as I am aware titanium alloys don't make good gears. I have never seen it out in the gear shop or at the gear shops we farm work out to and never seem any of the AGMA newsletters. On the net the only references to Titanium gears are for R/C cars and servo motors. According to our production people with titanium - you can expect twice the deflection as a steel part under the same load. In the aerospace industry the strongest gear are either 9310 harden & nitride or Nitralloy-N. Other that heat treatment Cryogenics is also another option for strengthening the gears. If fabricating new gears,change to a course pitch gear should be considered to handle more torque,Beefing up the primary and secondary shafts should also be considered. Removing the syncros and converting to wider spur gears may also be a consideration as well.

As far as I can tell the G series trannies made up to 2/2002 were designed handle up to about 200ft/lbs torque and I know people that had no issues with the up to about 300ft/lbs. Mazda's 2.0L RF-T makes 230ft/lb @ 2000rpm so there must have done something right on the G series made after 3/2002. The latest version of the RF-T makes 263ft/lb @ 2000rpm through the new 6 speed tranny.I have been talking to a guy in France that is making 310whp and over 500ft/lb of torque from the RF-T through the new G25M-R seems to be holding up. Since neither one of these trannies are used in North America or Japan it is not likely that we will find one easily. I have one of these trannys just waiting for transport the next time there is a container coming over.

Ultimately a tranny case made like the Ford MTX-75 would be ideal.

Compared to a G25M-R (4 cyl. version)

I also see that the aftermarket has made a kit to replace the stock helical gears and removed the syncros with wider and courser pitched gears

Combined with the design of the MTX-75 gearbox this should hold near 500ft/lbs. The only problem with this setup that gear whine with be pretty loud and it appears that the gear ratios have been closed up and non-overdrive so would not be streetable or 1/4mile friendly.

In terms of what can be done to the tranny is going to depend on what engine you have, your budget limitations and what form of motorsport you are in. There are a lot more options for a 4 cyl that the V-6. The most obvious ones for either engine are to keep the tranny temp as low as possible, reduce the radiant heat under the hood, reduce wheel hop, and reduce engine/ tranny twist. After this if you have a B-block and F/I making more than 250ft/lbs considering importing one of the newer beefier G25M-R from a 3/02+ BJ 323F with a 2.0L RF-T may be worth the S$3000-4000. Not only is this tranny stronger but the gear ratios are very turbo drag friendly. Since the transmission mounts to an adaptor plate on the B block it might be possible to adapt a Ford MTX-75 (most of the fab would be for engine mounts, converting to a cable shifter and driveshafts 240mm flywheel and with 25mm input shaft and 27 spline clutch hub). Another possibility myself and other O.G MX-3er have talked about is the Mazda H series FWD tranny used in the 88-92 MX-6/ PGT this tranny also as a 25mm input shaft and uses a 240mm 27 spline clutch. This may require a lot less fabrication than the MTX-75 because the motor mounts are already in similar locations and it has a rod shifter. The V-6 you are pretty much limited to what you have with out going to custom one off gears and shafts. The aftermarket racing tranny such as the Quafie and the Hewland EGT & PCT are still not designed to handle more that 250-300ft/lbs so robbing a bank may not get you any further ahead.

[PATDIESEL]

So basically we are screwed into a lot of custom fabing work.
The Ford tranny doesn't seem like a great alternative since so much of the tranny is going to have to be reworked in order to get it to A. fit i nthe egine bay and to the motor B. work with our exhisting axles, crank and shifter.
So the real answer for those looking for a better trans for under 5,000 is to source a newer Mazda G25M-R from a 323f (Lantis) 2L turbo and have it shipped over.
However, for running close to 300 fl/lbs we should consider ceramic coating the trans case, add a trans oil cooler and pump and maybe heat wrap the exhaust until after it passes the engine bay.
Geez, it is really too bad that Ford got so envolved with Mazda and basically took over the engine and trans manufacturing. If not then maybe Mazda would have stuck with the G series trans a bit longer and worked on improving its structure.

[rebel2k4]

Even so, what about stiffening up the case by welding aluminum gussets to it?