Part One described the design features and internal components of the Tipo 61 gearbox, and this entry covers setup and known failure points. The gyrations involved with getting it in and out of the car are also covered.

Special Tools

Maserati liked to use ring nuts, which require a special socket to properly apply torque. These tools are typically made by selecting an impact socket of the right size, then turning the open end to the diameters required. An operation in the mill cuts the appropriate number of slots so that what remains are four or six drive dogs that fit the nut.

The double row ball bearing that is held into the left side cover is secured by a very large ring nut, and instead of a socket, a spanner is made using flat stock that reaches across the nut and engages two slots, welded to a large nut in the center.

It’s worth considering how tight that bearing-retaining ring nut should be. The nut is secured with a setscrew that is drilled and tapped to intersect the thread of the ring. When you fit the new bearing, heating up the magnesium case makes it an easy drop-in fit. Oil the mating face of the bearing and nut, and use blue Loctite to lubricate the thread and backup the setscrew, then tighten the nut. Does the notch in the thread line up for the setscrew? How much has the case cooled off?

If you heat the case again, you can probably get it to turn another quarter to half a turn, depending on the temperature. If you were to assemble it cold and the set screw lined up when it felt appropriately tight, then when the case was at operating temperature on track, the bearing would likely be loose in the cover. This bearing controls the side-to-side location of the shaft, so affects bevel gear lash as well as gear alignment.

This issue of the magnesium case growing with heat can affect many setup variables inside this transaxle. It is also why the mainshaft and layshaft both have roller bearings on one end, so that the cases can grow while allowing the shafts to float at one end.

In addition to the ring nut tools, Maserati made the main cases with tight counter-bores that the main case studs fit through. The 14 mm nuts that fit inside the counter-bores need a very thin wall socket to access. I used a 12 point socket that was turned down to the minimum wall thickness possible.

Inspecting and Evaluating Parts

The beginning of a transaxle service includes a thorough cleaning and inspection of the parts. Magnaflux crack testing is mandatory and often results in rejecting gears, dog rings, and shift forks. Parts can also be rejected due to worn dogs, sloppy shaft splines, and gear tooth damage. The main gears one through five are rarely worn on their teeth before the dogs are shot, but the bevel gears and the final drive have a very hard life.

The Maserati Birdcage parts pipeline is surprisingly robust, considering that fewer than 30 cars are likely to exist, including replicas. Crosthwaite & Gardiner and Steve Hart Racing are the sources for most of the parts I’ve fitted to the Tipo 61s that I’ve worked on.

There are a number of things that need to be checked when new parts are unboxed. The gear hub needs to be wider than the gear by a few thou for end float, or the gear will be pinched by the neighboring parts and bind up. There should be couple thou radial clearance between the hub and the bronze bush in the gear. The oil drillings should be inspected to make sure they are clear, and de-burred if they have sharp corners.

New dog rings need test fitting into their shift forks to make sure they don’t bind up, and the same goes for new forks. I have also had new shift forks arrive with a different shape that would interfere with the magnesium case, so work was needed to make clearance. In one gearbox, the case casting’s internal shape prevented the new 4/5 fork from engaging 5th gear, and much material had to be removed.

Each time a part gets replaced, like a shift fork or a dog ring, the spacers will need to be changed. None of the parts I’ve received for one of these is an exact match to the old part. New shift rails will likely have the detents in slightly different locations.

All the bearings should be replaced on each service, regardless of appearance. All are easily available apart from the two splined-hub roller bearings, which are a specialist item. The mainshaft bearing next to the final drive is particularly important and should be the highest spec you can find. At Phil Reilly’s, we used a full-complement bearing that had no cage and an extra ball.

The Mainshaft Spacers and Splines

Each of the five gears on the mainshaft has a splined hub that the bronze bush in the gear spins on. All of the gear hubs, and the three dog ring hubs, are squeezed up tight by the big ring nut on the end of the shaft, and they have a pretty small mating contact area. Every time the driver lifts off the gas and then accelerates again, the hubs are trying to work against their neighbors, attempting to rotate on the shaft, but restrained by the splines.

As the parts go through hundreds, thousands of cycles of load, unload, the parts begin to wear minutely. The carefully ground spacers lose a few thousandths of width. After a certain point, the torque tensioning the stack together is reduced, and relative motion gets easier, until eventually all the parts are loose and rattling around.

That makes the shifting sloppy and the dog rings begin to get more beat up. The splines start to get worn and loose. It is one reason I believe these gearboxes need to be serviced every season, to check for wear in the shim stack and replace them as required, before parts are damaged.

Mainshaft Setup

When test fitting new parts, start by assembling the shift rails and forks with the detent balls and springs. Then the built-up mainshaft fits into the lower main case half, using the left side bearing carrier’s 6mm bolts to locate the shaft in the case. The two long center case studs need to be removed to slide the mainshaft into place, the shift forks limiting the angle of insertion.

Then I engage each gear and check how much room there is between the gear and the dog ring. It can be measured with feeler gauges, targeting somewhere between .005″ – .010″, but it is easy to get a feel for it.

The way I attack it is to engage the gears by moving the shift rails, then push the dog ring towards the gear. When it is set right, the shift rail will have the ball in the bottom of the detent with that 5 to 10 thou clearance to the gear, so that when pushing on it you feel the detent spring compress as the ball rides up the side of the detent notch in the rail. The dog ring makes a distinct “clink” sound as it hits the gear. The sound becomes more of a dull note as the clearance tightens.

If there is zero free play between the gear and dog ring, you need to move the gear away from the shift fork. That will mean a larger spacer between the dog ring hub and the gear hub, and removing the same amount of material from the shim on the back side of the gear. If there is too much free play, the opposite is done.

Sometimes this is easier accomplished by grinding or turning some material off the gear itself, or one of the hubs. I aim to have a minimum shim thickness of .060″ for durability, and thicker is better. Having a handful of shims for testing is pretty much essential.

To reduce the wear on these shims over the course of a season, we started using a Nitralloy steel that gets heat treated post-machining, with a final surface grind to get it on spec.

The mainshaft itself is positioned in the cases by the left side bearing that lives in the steel housing. It is possible to move all the gears left or right by changing the shims around this bearing.

The Layshaft

Once the gears on the mainshaft are sorted, you need to look at their alignment with those on the layshaft. These gears are splined to the shaft and held tight with the small ringnut on the end of the shaft. The shims are changed or gears shortened as required to get good tooth alignment. But before that happens, the shaft has to be positioned properly.

That is related to the mesh of the bevel set of input gears. The shaft’s position is controlled by the spacers on either side of the gear and the large double row ball bearing that lives in the end cover.

So the first step in positioning the layshaft is to check the bevel gear lash. That requires mocking up the gearbox cases with just the bevel set and the layshaft. A dial indicator with a long stick can go down through the final drive opening to contact a tooth on the input bevel. To establish a reference for the layshaft position, I bolt a steel parallel to the right side of the case and measure to the face of first gear.

In the above image with the depth micrometer, you can see that first gear on the layshaft is wider than the mainshaft gear, and sticks out by around .080″ towards the camera. My experience is that these gears are never the same width, and they don’t have to be exactly aligned.

Things that Break

I experienced a couple of major failures in my time with these cars.

First, a layshaft broke just under the nut that pulls the gear stack up tight. This is a relatively small nut compared to the mainshaft, 14×1.0 mm if I recall, whereas the mainshaft nut is more like 32 mm. Was it over-tightened? How do you choose the torque spec? The broken part showed no sign of the threads being stretched, just a fracture at the thread root. Most likely a weak point and a stress riser at the root, possibly affected by the heat treat of the shaft. This caused the nut and spacer to fall off the end of the shaft, allowing the stack to come loose.

Second, a mainshaft bearing failed, the most highly loaded bearing in the ‘box, next to the output gear that drives the ring gear.

That one was the biggie, which seized the gearbox on the front straight at Laguna Seca, and could have ended with the car in the wall at high speed, but for the skill of the driver who managed to save it. When the bearing failed, balls escaped the rolling tracks and the mainshaft tried to push itself away from the giant final drive gear above it, which was being driven by the car’s inertia via the wheels and half shafts. The main cases were damaged beyond repair and a new set was required, along with many other internal parts.

In what was certainly the most crazy thrash of my life at the racetrack, I was able to swap the broken box out for the spare, with the help of the whole team, I think in between sessions? It normally takes a whole day to fit a Birdcage transaxle, in the shop!

The post-race assessment of the broken box concluded that the single row ball bearing on the left end of the mainshaft came apart, but there was a heated argument as to why. “BEARINGS DON’T JUST FAIL!”, was Ivan’s loudly delivered opinion. I remember that argument as the day that I came closest to punching the old man right in the nose.

My opinion was that every bearing has a lifespan and this one is very highly stressed, so it should be replaced every season. I thought that this gearbox did too many laps between rebuilds. Why else should it have failed? There was no sign of overheating, no loss of lubrication. The box worked perfectly until it stopped. It simply timed out.

The consensus was that the bearing should be spec’d to the highest standard possible. We special-ordered a number of full complement bearings for that purpose, which eliminate the cage holding the balls and have more of them as a result. More balls = greater load distribution, less load per ball.

That failure was in 2008 and it was not repeated after switching to the new bearing.

The bearing Carrier

A related issue is that this bearing is held in a steel case, bolted into the magnesium casting. Twice I have had to replace this steel housing, because the bearing outer race has started to spin, the internal bore of the housing worn out. The cover plate that squeezes the bearing race is quite thin and can’t be expected to apply much force, via the five screws, of 6 mm diameter. More evidence, in my view, that this bearing is worked extremely hard.

One other related note… the oil nozzle that feeds into the main shaft fits through a plug in the end of the shaft. Each ‘box I’ve worked on had a different seal design. One had a spring loaded face seal, like a water pump. The one that blew up with this bearing failure had a brass insert holding a lip seal, and the oil nozzle was broken when the shafts tried to get away from one another. When the new shaft was fitted, a new brass seal carrier went in.

On the next gearbox service, I found that the new oil nozzle had displaced the lip seal and had a groove worn from contacting the brass insert. My conclusion is that when these gearboxes are hot and maximum torque is being transmitted, the shafts move dramatically more than you’d expect. How else does this oil nozzle get damaged like this?

Getting it out of the car and back in (the dance)

The design of the Birdcage chassis made it impossible to simply lower the gearbox into the opening in front of the giant gas tank, or to lift it up from below with a jack.

I’ve had the trans in and out of four different Birdcages, and though each are different, they all require a complex dance to get it into place. Some cars were modified with removable tubes that made it much easier, but those bolt in tubes are obvious and non-original, which some purists frown upon. Each bolted-in tube also reduces the chassis stiffness by some increment.

The dance goes something like this…

• remove the gas tank and place a plywood platform in its place to stand on
• Climb a step ladder and get into the boot area, standing on the wood
• have someone hand you the transaxle, over the fragile aluminum bodywork of the tail
• Twist your torso and lower the box down to sit between your feet on the wood, sitting in its normal forward facing orientation
• Take a breath and stretch your back muscles
• Pick up the box and rotate it 90 degrees around the fore-and-aft axis, so that it will fit between the frame tubes, and lower it through the tubes part way home.
• rotate it another 90 degrees on the other axis, to get it to clear another set of tubes
• rotate it again, another 90 degrees to allow it to drop into its four mounts in the bottom of the chassis. Make sure that the de Dion guide block on the axle has fit correctly into the guide rails on the front of the ‘box.

Joel Finn’s classic book, “Maserati Birdcage – The Marvellous Tipo 60 and 61 sports racing cars” claimed that the transaxle weighed 85 pounds. It ain’t easy to do this dance!

The four bolts that secure the case to the frame via metalastic bushing are large diameter and hollow, with thin 24 mm nuts and split pins for locks. The upper mount was a fairly complicated tube fabrication, sort of a dog bone, clamped to the top of the final drive case, with two more metalastic bushes and two more hollow bolts at the ends.

Once the complicated gymnastics are finished, and the gearbox is basically in place, it’s time to connect the half shafts, via heavy duty U-joints, to the yokes that were already mounted into the top case containing the ring gear. This is tricky, because the half shafts are a fixed length, with the U-joint yokes on the inside, and a large pin on the outer end that holds two needle bearing roller cups, rolling inside a track in the wheel hub assembly. Called a bucket, this allows for the change in length as the wheel moves through its travel.

With the gearbox bolted in place, there is often not enough space between the yoke on the gearbox and the inside of the drive bucket to assemble the U-joint. So you must leave the bolts out and jack the box up slightly to make room. Once the U-joints are together, you can fit the bolts.

When fitting new U-joints, I often ran into problems with the distance between the circlips that secure the cups into the yoke, and with the cup diameters being too loose or too tight. It seems like every car has differences with the yokes, and typically the parts are numbered to make sure they go back in the right orientation.

Another fun part of installing the T61 gearbox is connecting the driveshaft. The flanges use a typically Italian design with drive dogs that engage in a female slot in the opposing flange, so that the bolts only clamp and don’t transfer torque. So they can be small, only 7 mm diameter… but the bolts are designed to go from the rear of the car towards the front, which means they are trapped between the input flange yoke and the seal carrier on the front of the gearbox. I used temporary nuts to pull the bolts up to the flange to keep them secure when working on setup, otherwise they get caught up and bind the input flange. If you were to forget to fit the bolts before bolting up the pinion assembly, you’d have to go backwards. They will not fit after the flange is on the pinion.

Buttery Smooth Shifts to Ruination in a Moment

My time with these gearboxes has proven that careful attention to detail is the key to a happy driver and a reasonably long life. The complexity of the design and the time required to do a proper service means it will never be as easy as a Hewland, and there is no way of knowing what’s going on inside without pulling it apart. When it does go badly, it can be gut wrenching.

Given how valuable these cars are, and how quickly a gearbox failure can end in disaster, I see no reason not to give them the care they deserve, as often as possible. Regularly inspecting the internals also saves money over time, as correcting small issues like worn spacers is much cheaper than replacing major components – even if the labor involved in getting it in and out of the car challenges that point.

It’s a real paradox that this component was responsible for so many DNFs in the Birdcage’s competitive life, while also being praised as one of the best-shifting gearboxes of its day. I think it is safe to say that the difference between the win and the failure came down to who was last inside the unit and how carefully they put it together. With proper attention, these cars will continue to star in historic racing and do the engineers at Maserati proud for a very long time.

If you have enjoyed this look at the details of an old Italian icon, consider subscribing to One Day in the Shop. There’s more to come from the Birdcage files and plenty of other projects that deserve a closer look.