So, where do we go from here? First, it's important to understand what failed in the broken transmission, and why. Next, we'd like to figure out how to prevent another occurrence, because transmission replacement is not an enjoyable recreational activity. Finally, we'll hopefully build a stronger gearbox that can survive 600+ lb-ft of torque on the input shaft and a couple tons' worth of mass on the other end.
*Yes, that was several months ago. Welcome to the real world, where projects don't move in nice, neat one-month increments.
Carnage was expected upon teardown, and we got it as soon as the tailshaft housing was removed.
Fortunately, the magnets located at the rear of the case did their job, and collected a handful of large fragments. So where'd this metal come from?
Once we removed the case from the midplate, the situation was clear - most of the teeth had been sheared from the input shaft. This is the gear that transmits the power to the countershaft in all the gears except 4th (this is a 1:1 direct-drive ratio, so the input shaft is connected to the output shaft and does not route any power through the countershaft). The countershaft didn't suffer any major breakage, but several teeth were showing chips and discoloration from excessive heat.
In all likelihood, the damage was caused by separation between the input shaft and countershaft under load in the first three gears, which was probably the result of insufficient pre-load on the input shaft and mainshaft assembly. We'll correct that the next time around.
So, with two expensive components thoroughly trashed, it was time to see what we build from the parts that we had.
We knew we wanted to use as much of the SSR transmission as possible, because of a few features. First, its gear ratios are lower:
1st: 2.66 2.97
2nd: 1.78 2.07
3rd: 1.30 1.43
4th: 1.00 1.00
5th: 0.74 0.84
6th: 0.50 0.57
With our engine, the deeper and wider ratios will result in a more enjoyable car. Second, it has double-cone synchronizers on all of the forward gears (the F-body box has that feature on only 1st and 2nd). Third, the SSR gearbox uses a stronger one-piece countershaft. Finally, it has a stronger 32-spline mainshaft (the F-body box has a weak 27-spline output, which we'd previously broke and replaced with a 30-spline Viper mainshaft).
In addition to the above desirable differences, there are a few others that caused some hassle. The LT1-style trans used in our Impala uses the same pull-type clutch as an '93-'97 GM F-body, which means that it can (and does) use a shorter input shaft than the push-style clutch and hydraulic throwout bearing that's used in T56s behind GenIII/IV GM engines (such as the LS2 in the SSR). This 7/8" difference in length means that the SSR T56 won't bolt up directly to our LT1.
The first thought was to use an LT1 input shaft in the SSR trans, but you can see in the above two pictures that the length isn't the only difference between the two parts. The reason for this is the double-cone 4th-gear synchronizer assembly on the SSR mainshaft (on the left in the above pic); the floating friction ring requires four "ears" that nest into the rear of the input shaft. More importantly, the SSR input shaft has fewer teeth (29 vs. 33 on the LT1 input) to yield numerically-higher gears.
This left us with three options: 1) Use the input shaft from a '93 T56 (which has the lower 29-tooth count) and swap over to a single-cone synch; 2) cut down the SSR input shaft by 7/8"; or 3) put a 7/8" spacer between the trans and bellhousing to accommodate the longer input shaft. Yeah, nothing is ever easy or simple when it comes to non-stock parts.
Option 1 was out because we wanted to retain the stronger and slicker-shifting double-cone 4th-gear synch on the SSR trans (the 3-4 shift takes place at 95 MPH, which means that it's encountered frequently at the track). Option 2 was ruled out as the SSR input shaft didn't have enough spline length to engage our McLeod twin-disc clutch after being cut down. That left Option 3 as the path, which means that we'll need to make a spacer.
We'll use an LT1 midplate on the SSR trans, as it provides the mounting location for the clutch fork pivot. The pivot will need to be lengthened by the same amount as the spacer thickness, as it's now moving rearward relative to the clutch throwout bearing.
The SSR shaft was 5/8" shorter than our modified 30-spline Viper mainshaft, which works in our favor.
A call to McLeod got us a part number for a 3/4" spacer (normally used with the company's modular bellhousing system). That's enough to allow the use of an unmodified SSR input shaft.
Since the spacer isn't made for this exact application, it of course doesn't fit the LT1 midplate perfectly.
A portion of the spacer was cut off...
...positioned on the other side of the cut-out for the clutch fork...
... and TIG welded in place (we tacked the piece while it and the rest of the spacer were bolted to the midplate, and then removed it to complete the weld).
The spacer was $75, and we spent about an hour modifying it. That's far less time and/or money than we'd spend fabricating one from scratch.
A small block of steel was welded to the clutch fork pivot and then milled down to provide an additional 0.75" of length.
Longer dowels were installed into the midplate to accommodate the spacer.
When mocking up the transmission (using the empty case - it's about 100 lbs lighter and much easier to handle), we stumbled along one more difference - the shifter pad on the SSR tailshaft housing was about a half-inch higher. That means that our B&M Ripper shifter wouldn't fit, and we'd have to buy something specific to the SSR. Or, we could simply cut down the mounting pad on the SSR tailshaft housing.
The housing was clamped into a Bridgeport mill...
... and a series of fly cuts performed to trim it down. This only took a few minutes, and saved us from buying a new $160 shifter.
In our next installment, we'll get this mutt of a transmission assembled, and eventually we might just install it and get a chance to drive the car for the first time in over a year.