roadster axle and transmission upgrade

this started out as work i did not want to do, but glad i did. the 3rd gear synchro on the T96 transmission died on the return leg of a road trip. easy highway driving, and i shift this weak transmission easy. i rebuilt it myself in 2007 or thereabouts, parts were scarce and expensive and it's a lousy transmission. the only transmission that fits this particular bellhousing is this early version of the T96. the T14, a far more "modern" transmission fits the most crucial dimensions -- input shaft, bearing retainer, and two upper bolts -- but the lower bolts hang below the bell, and it of course is a different length. it's also a non-overdrive transmission, so i'd have to re-gear to accommodate a non-OD top gear. and given that the early American's "big nut" axle itself has unobtainable parts (pinion shaft seal) and none of the ratios available are suitable, it was time to modernize. i took the hint from current practice and got a 1998 Mustang axle. now the axle is of nearly identical design, but it's too wide, so i had it narrowed. it also comes with disc brakes, so instead of retrofitting drums to it, i bought a Scarebird kit to put disc brakes on the front.

not only. i'd already ditched the leaf springs, and after an initial design with enough faults that it is not documented here, i'd fit a fairly workable rigid wishbone to the Rambler 'big nut' rear, which worked quite great, with some minor shortcomings (mainly, non-adjustable). so for the new Mustang axle i redesigned the wishbone system and made it adjustable and stronger. this being a unibody car, basically a second-generation Nash unibody, to anchor the front wishbone joint i'd made a "spider" that supports a big heim joint under the front U-joint. this spider too was replaced, and made adjustable, stronger also, and i added support for the very long T14 transmission tailhousing; the Nash design allows the transmission tail to bob somewhat eg. going over big bumps. given the closeness of the added wishbone pivot there was occasional clunks (which after most of a year did little more than nick the paint, but was unnerving) the fifth engine mount seems like a good idea.

upon disassembly i found, unsurprisingly, that one of the rear motor mounts tore, and needs replacing. but of course they are no longer available -- from any source -- and N.O.S. rubber is pure foolishness. so yet another task was to redesign the rear mount system. and of course the clutch had 8 years of use on it; though it was still in good shape it seemed foolish to not send it out for rebuilding. and so it got sent out to Tennessee Clutch and Supply.

obviously one thing led to another... and so the broken brass ring ended up with a substantial upgrade/overhaul. while i was plotting, drawing and accummulating parts, during the last months of the school semester, i continued research into suspensions, found the wonderful suspension calculator software (see bottom of the main Roadster page) and realized that i could possibly eliminate much of the notorious Rambler understeer by... drilling two holes. later on that.

the first task was to provide the lower mounting bolt targets to the bell. the bellhousing is tiny. the bottom half is cut away, with a steel cover affixed after installation, that provides access to the nuts for the two rear motor mounts (note that this is a "four point" mount system, there is no transmission crossmember) which won't be replaced since none are available. the front mounts are common as dirt, since they have some industrial application, i have a bunch of them, and so i made the transmission adapter also accept a pair of front-type mounts. this ended up being super clean, easy, and i think better than factory. however they are less compliant, so my guess is driveline stiffness was surrendered for a quieter, softer car.

the new Problem Solver/transmission-and-new-mount-adapter fits neatly into place. the T14 is hefted into place, sans clutch, to check fitment and to assemble the rest of the driveline and rear suspension. at this point the T14 dangles off the back of the bell as did the T96. even though it's longer it's lighter without the R10 OD unit.

wishbone rear suspension

next is the wishbone build. a lot of planning and plotting went into this one, which belies it's simplicity. this one is even longer, nearly 70 inches from axle centerline to front pivot, forward of the front U-joint. the wishbone is two lengths of 1.25" DOM, 0.120" wall. the struts are prefab swaged tubes from, with left and right threads.

the wishbone is an old design, but not inherently a bad one. it solves a lot of problems that more-popular rear systems, like truck arms and four-bar links, have. it's easy to make it have zero bind due to body roll. pinion angle change is very low. it requires some care to make it stiff enough, but no worse here than truck arms. it can be thought of as truck arms extended forward until the arms meet, and share a single joint.

a wishbone has dynamic geometry similar to the torque arm rear suspension, but the wishbone prevents lateral axle rotation and doesn't need the two lower arms shown in this page. it does of course still need a panhard bar to lateral location. the wishbone i constructed has pinion angle adjustment built in. discussions on the AMC Forum (thanks Red Devil) point out that this current (single tube) design will probably deflect somewhat. i suspected my previous one did, hence half the reason for the adjustable struts; this will certainly flex less. if it ends up being unacceptable it would be a day's work to make a trussed, dual tube wishbone of arbitrary strength. but for now i'm going to drive what i built.

the axle was narrowed by Cook's Machine Works here in Los Angeles. they cut from the middle of the tube and butt weld, the whole assembly held in a large lathe to keep it true. i had the pinion centered. i retained the OEM four-link brackets to mount the wishbone, and extended them upward a bit to add the anti-wrap braces, which also serve as geometry but mainly pinion angle adjustments. though these joints are rigid in operation i used heims to accommodate the angles involved; also rotating the assembly about the axle axis changes the triangle angles, so some way to accommodate the geometry change is required.

the wishbone was aligned using V blocks and a cheapo laser. the front plate (which will later receive the front pivot heim joint) is scribed with a centerline and the V blocks centered on it with a mechanic's square. the triangle was then tweaked to put the dot on the pinion nut. the rear heims will provide a little bit of adjustment for squareness on the car. the bottom picture below shows the final strut assembly and the air spring perches welded to the axle tube.

since this is a unibody, and there is no transmission crossmember handy to bolt things to, i made a spider that positions the wishbone's front pivot below and forward of the front U-joint. this conveniently passed under the T14's rear mount (the T14, a Jeep transmission, was adapted by AMC for Rebels etc in the 1960's, by which time all AMCs had switched to threee-point mount schemes). conveniently, the transmission mount is very compliant, so it won't be doing much more than keeping the transmission tail from bobbing and klunking the necessarily close-tolerance spider and heim joint. the white marks on the wishbone pivot mounting surface (on chassis centerline) shows the engine offset to the passenger side. it's dangling from one mounting bolt so it's hanging at a funny angle.

all of the design and above work was done with the car in pieces; now the completed axle and wishbone (minus axle shafts, minus front heim) is placed under the car, the air springs located correctly over their axle tube pads, and the chassis dropped onto it in full collapse, eg. no air in the springs, on the "bump stops". whew! everything clears nicely first time. this is worst-case, not operational.

got the rear under the car, got it approximately squared (within 1/4") so that i could measure, cut and weld in the heim joint at the tip of the wishbone. it tips upward at the very end to increase driveshaft clearance without decreasing ground clearance. the driveshaft runs less than an inch from the wishbone tip since the excursion is a fraction of an inch. still not happy with my reference points up front. i think i'm going to weld up a tape measure reference onto a Zerk fitting and screw one into the trunnion at the end of each lower control arm, since it is the center of the virtual kingpin. it's close to the tire and nicely low to the ground and clear everything. at the moment the air springs are empty, and the axle is positioned at maximum bump, so it' unrealistically high but it made things easy to get and and adjust, since all i was doing here was squaring the axle to the chassis and locating it longitudinally (eg. wheelbase). for pinion angle and driveshaft length, and panhard mounts, i'll drop it to ride height.

also welded up the huge hole that had been necessary for the old Rambler Twin-Stick shifter that fit the T96 OD so nicely. it was actually a great shifter, but not compatible with the T14. i banged out a piece of sheet steel and welded it in place. long ago i'd scrounged a N.O.S. Hurst Indy 3-speed shifter, finally got to use it.

panhard rod

the mount for the axle end of the panhard rod ended up being a pain. there was no fully satisfying point in close to the axle, due to the shock mounts and the rear cover. the panhard runs along the axle axis, but spaced off rearward about 6" from the axle centerline, which means the side to side forces are translated to the axle at a 45 degree angle. this oversized-seeming bracket spreads the force to two places; the anti-groan bracket to the brake flange, and a fat bracket up near the center housing. the shock can just barely be installed (but the shock stud itself can also be removed).

with that complete, and enough parts arriving, the axle and axle-mounted stuff is assembled, except for brake lines and e-brake cable. i'll get the correct fasteners this week (at the moment all the heims are held with junkbox fasteners). all of the fasteners pinning the wishbone heims are in double shear.

final assembly

assembly followed the construction above by a few weeks, as i awaited the arrival of ordered components. no serious rework necessary, luckily. i swear i had squared up the axle when i placed the axle-side panhard mount, but it's located about 5 degrees up from vertical, somehow. i should have, but didn't, make the panhard fixed mount height adjustable, so i will fix both in a few weeks once i've got some hours on everything.

the clutch came back from Tennessee Clutch and Supply just fine (actually, two clutches, i sent them three core sets of these old Nash rare-but-low-value 8" clutches), got back two, one's on the shelf. they threw in new throwout bearings. this went in just fine, and i took the opportunity to completely replace the vestige of fabric-support for the clutch inner Z-bar pivot, replacing it with a heim and a nut welded to a stub of the old fabric hanger. got the car up high enough that i just hefted the transmission into the hole. bolted the wishbone spider into place, which takes the "fifth engine mount", on the transmission tail, that will dampen it. most of the mass is supported by the rear motor crossmember and mounts.

the axle assembly, including wishbone, slip into place easily. i just stuck it on the creeper and slid it under, lifted the tip up to pin the front joint into place, then lifted each axle end onto a jack stand and aligned it by eyeball under the chassis. from here i was able to jack, shim, and position the axle precisely, squared to the front trunnions. from here i could build the panhard chassis-side support, and cut the panhard to fit. i jacked the axle so that it's relationship to the chassis was identical to riding on the ground. the springs were attached and inflated, and chassis lowered onto the axle via the springs so everything was loaded. with this panhard measurements would be correct.

panhard installation

the chassis-side panhard mount i smartly (this time) made removable. which is good, because i already need to modify it (to make the panhard adjustable). here the panhard itself is a piece of thinwall and cheapo tube adapters, awaiting the good stuff from for the 1" DOM tube.

parking brake

the parking brake was surprisingly straightforward. the old rambler system was a ratched pull under the dash, rotated counterclockwise released the catch. this pulled a cable that pulled a jack-lever, about8" to the fulcrum, mounted on the rear motor crossmember, to translate stroke into torque. at about the one-inch point was attached a secondary pull rod that had an adjustable yoke on the end to which attached the cables that ran to each rear brake. drums of course. the mustang system was identical. i made a new adjustable yoke and drilled a new hole in the jack-lever. worked first time! well almost, it's too stiff at the pull handle and stroke is only a few inches, so i need to move the secondary pull rod closer to the fulcrum. no big deal, i already drilled an extra hold.


when i first started i thought this was going to be a big deal, but it turned out to be absurdly simple. the old Nash system, with the T96, used a very peculiar and tiny U-joint, but the "modern" T14 used a standard SPicer 1310, and the mustang wants a Spicer 1350. so a Neapco combination 1310/1350 solved the U-joint problem, and in my iron pile i had a driveshaft that was about 3/8" longer than ideal, so i increased wheelbase by about 1/4" (that front wishbone heim). with suspension in full compression there's about 0.5" of yoke stickout.

rear (hydraulic) brakes

again no big deal. i don't flare my own lines anymore. i have the tool, the common cheap kit that does double flares, but it's a PITA to get them really square. i now simply build with stock lengths of pre-flared tube. i've never had more than a single couple, and often none.

i've got the typical hot rod setup of the master cylinder in the floor, mounted lower than the wheel cylinders/calipers, but so far so good. no bleedback, so no need for 2 psi checkvalves, yet anyway. i temporarily have the odd combination of front drums and rear discs. i have a Scarebird kit for the front, used 'em before i know it goes in just fine, but i wanted to minimize problems and do it in stages. i added the rear manual proportional valve. it sticks up through the floor just under the front seat.


(whatever "done" means.) looks same as before, except the axle is pushed back a tiny bit, and there's those telltale mustang axle ends instead of the big castellated nut and cotter pin. underneath nothing dangles lower than the axle housing and crossmember.

T14 transmission

after disassembly, and three serious washings, i found two broken parts; the tip of the shifter fork and the 2-3 interlock plate. the transmission might have been dropped on a lever or been in a crash. found the parts on bought the rebuild kit (for the Jeep T14) from it rebuilt otherwise straightforwardly.

first generation rear rigid wishbone

i took some photos of the first pass wishbone before i dismantled it. i made the embarrassing assumption that the engine and transmission is centered in the chassis -- it's not, it's offset about 1" to the passenger side. needless to say i had alignment issues. i'd kludged it parallel, but the rear was offset, and there remained a U-joint error that manifested at 65 mph (only). i was lucky i got off so easy. though this wishbone isn't adjustable, i got the pinion angle within a degree. problems were solved in the current system, mentioned above.