Z33 + Z34 6 Speed Gearbox Conversion: Part 1 - Adaption
- If you machine, grind, cut, weld, bash and mash enough then anything will "fit" anything.
- If you don't care for reliability and disregard your own safety then anything will "work".
The Long Story
So where to start? The beginning? Sounds good let's go!
Once upon a time in 2009 a z33 350z 6 speed gearbox was bought to design and engineer the parts required to adapt said gearbox to the Nissan SR20 Silvia - 180sx, s13, s14 and s15s.
By that time we had already completed a tonne of research on a tonne of gearbox options and had already finished the first version of the Z32 Gearbox Conversion Kit “Z32GCK”. My my how time flies... 7 years later we’re already up to version 5 of that thing with zero failures to date - just constant improvements. Yay!
A tonne of gearboxes for R&D
Gearbox conversions are simple. Take one coordinate measuring machine (CMM) and map the coordinates of the Z33 bellhousing bolt pattern, centre point, alignment dowels and other knickknacks. Start your favourite computer aided design (CAD) software and then overlay your Z33 coordinates onto your S14 data and draw a profile.
Following on from there you work the computer aided manufacturing (CAM) programming, fire up the CNC machine, push a button and out pops a super accurate CNC machined adapter plate. My brother’s wife’s cousin’s mate has an uncle that knows this guy that supplies the billet material for a few dollars... our Australian engineers and machines that do the work costs hundreds of thousands of dollars.
Alpha Omega Racing Z33-to-S14 CNC machined billet adapter plate
Done! Now to bolt on adapter plate, bolt on gearbox and proceed into racing heaven.
Too Soon Junior
I jumped the gun and got to the exciting part first. Let's rewind it a little. Alpha Omega Racing does, amongst other things, 3D modelling and accurate simulations to make sure everything works like it should for a long time. Here’s a little snippet of the Z33 vs S14 gearbox input shaft comparison from our CAD models:
3D model of Z33 and S14 gearbox input shafts, to scale
It's modelled TO SCALE and modelled TO EXACTLY how it would sit when installed. What you would like to see in this model is the Z33 splines in line with the S14 splines, however a blind man can see the Z33 splines are way off.
For those that don’t know, the input shaft is the one and only connection between your engine and the rest of your gearbox – every kilowatt of power and every Newton meter of torque goes through this 22mm round shaft. More importantly, all that twisting force is going through 1mm deep splines that are only 45mm long.
3D model analysis
To be precise the Z33 splines are 17mm shorter than it needs to be. That's 38% too short, and I believe the technical term for this huge difference is "a country mile".
So what if the Z33 splines are a country mile off? Well... your clutch plate is literally falling off the input shaft and the splines are literally being twisted off. Bla bla theory bla bla... here’s a real world example:
Input shaft too short: the damage
Input shaft too short: closeup
As a famous commercial use to say “it won’t happen overnight, but it will happen”.
Soooooooooo there’s a few things that can be done to fix your short shaft problem *Insert penis joke here*
Cut and Shut / Weld a Bellhousing
This is a process where you cut the flange off your current gearbox (and your new gearbox) and then weld old flange onto new bellhousing:
Cut and shut example from Google
This was a service we thought about offering. Here is what we found. To save time I’ve used 50c pieces for quick illustrations. One 50c piece represents the face of your old gearbox, and the other represents the new gearbox.
First issue: Rotation: your flanges could be rotated off centre
Flanges incorrectly clocked
Second issue: Misalignment: North/South direction
Third issue: Misalignment: East/West direction
Forth issue: Misalignment: North/South + East/West directions (double misalignment)
Double misalignment: North/South/East/West
These examples are exaggerated so I can quickly show you the issues, but all it takes is one millimetre or one degree to cause issues. One little millimetre or degree doesn’t sound like much right? In engineering terms, it’s huge. Have you ever tried to screw in a nut or bolt that’s slightly off? No? Try it next time. Use a power tool for best results.
Fifth issue: Angled: Left/Right
Off angled left/right
Sixth issue: Angled: Up/Down
Off angled up/down
Seventh issue: Angled: Left/Right + Up/Down (double angled)
<No pic as I couldn't balance the coins on their corners, but you get the idea right?>
Other issues: Jig flex, jig quality, jig inaccuracy, human error, mix of above
"Jig" example from Google
"Jig" example from Google
I’m not saying that all of the above will happen, but it takes just one to cause you issues. Just look at the above jig we found with Google. There are many risks which cannot be controlled in the cut and shut method. None of these issues exist with an adapter plate engineered the way we outlined earlier.
What do these issues equal? Since first offering our “100% CNC machined adapter plate style” Z32GCK in 2008 and it's now 2015, we have 7 years of real world experience. We have signed a number of new customers after they complained about their cut and shut gearbox conversions being harder and harder to change gears, makes louder and louder noises, and clutches/bearings wearing out faster and faster.
“It won’t happen overnight, but it will happen”.
For these reasons we ruled out “cut and shut” bellhousings.
Custom Cast a Bellhousing
We priced it up, the looooooooooong Z33 bellhousing would cost over $1200 to cast and machine plus the initial tooling of patterns and moulds and it becomes really expensive - around $7000. Add on top the gearbox itself and don’t forget the other conversion parts like cross member and shifter and tailshaft and other knick knacks and it defeats the whole purpose of a gearbox conversion in the first place – the purpose of being a cheap upgrade before going to a proper sequential or dog box.
Let’s look at cheaper options shall we?
Use a flywheel spacer to push the clutch plate out to meet the short Z33 splines.
Here’s an example of such a thing, this is NOT DONE BY US as we would never do such a thing, but by an unnamed company who use to make gearbox conversions. They’ve stopped now.
An unnamed company's flywheel spacer
Unsurprisingly it had prematurely worn out clutches – premature as in a brand new clutch wore out in less than a month of daily driving, would grind gears and got to the point where it couldn’t select gears, would make all sorts of noises and so on. It was pulled out and inspected.
An unnamed company's cut down hardware store grade bolts to use as flywheel bolts
Notice the gap between the flywheel face and the bolt? All bolts were fully torqued. Turns out the bolts were cut down incorrectly and were your local hardware store grade 8.8 bolts. To put that into perspective, your OEM brakes are 10.9, we supply 12.9 bolts in the Alpha Omega Racing brakes kits, and flywheel bolts are even higher grade than that. ARP has built an international business out of making "just" bolts because the right grade / right tensile strength / correct yield is critical.
You can’t get the right grade bolts in the right length and diameter and pitch unless you special order a million of them, so if you don't want your flywheel flying off and chopping your legs and feet off then I suggest not using a flywheel spacer.
Machine more off the Z33 bellhousing so the gearbox sits further into the engine.
Have you ever looked at a Z33 gearbox up close? Next time, check out the flange on it *insert vagina joke here*
LEFT: 9mm machined, RIGHT: original
The left bellhousing has been machined 9mm, the right has not been machined at all. You’ll notice the left 9mm machined bellhousing has a tiny little flange left - machine any more off and you’ll have nothing left! Remember, we're trying to fix a 17mm deficit as calculated earlier in our CAD models.
Now would be a good time to mention adapter plate thicknesses. How do you work out how thick it needs to be according to mechanical engineering school? We’re talking strength and thread engagement and stuff like that here. It’s easy. Take the bolt diameter and multiply by 1.5. The Z33 gearbox uses M12 bolts so 12x1.5 = 18mm = minimum thickness. Machine 18mm off the Z33 bellhousing and you’ll have no flange left.
So you can't machine the Z33 bellhousing 17mm to fix the input shaft deficit, and also can't machine off an extra 18mm to make room for the adapter plate. There goes that idea.
- Z33 input shaft splines are too short
- Cut and shut bellhousings have too many risks
- Custom casting a new bellhousing is too expensive
- Use a flywheel spacer if you don't really need your legs
- Can't machine the Z33 gearbox deeper
To be continued...
Copyright ALPHA OMEGA Racing. ALL RIGHTS RESERVED. No part of this article can be reproduced without explicit written consent.