FJ20 Gearbox / Clutch Issue - Help needed

[BigMuz]

I agree with mrbishi I'm afraid. Just eyeing off the bolt holes could be out my heaps..

Why does it take 25 hours to get the engine out and back in?

[BigMuz]

You can run a dial indicator off the face of the flywheel running back onto the gearbox mounting spigot when you have removable bellhousing. Dunno what you do when you can't remove the bellhousing- maybe get an empty case and turn up a dummy mainshaft and bolt it all together.

[MrBishi]

I agree with mrbishi I'm afraid.

Be very afraid.

Are the dowels and bolt holes equi-distant from the centre of the input shaft? I have successfully centred a supra gearbox conversion using verniers and a bit of math, but it's a bit dodgy still.
There would be less than .5mm error in factory I'd reckon. Don't confuse bearing play with machining tolerances. Just because a bearing has a bit of play in it does not mean it will like operating with off-centre loads.

Does this new fangled NEO gearbag have an integrated bellhousing?

[tim510]

I agree with mrbishi I'm afraid. Just eyeing off the bolt holes could be out my heaps..

Why does it take 25 hours to get the engine out and back in?

It just takes that long for me doing it on my own. I can't remove the box in-car, and I don't have a hoist.

To remove the engine , i need to remove both the inlet and exhaust manifolds, radiator and i/c.

I'm obviously being careful as I don't want to dint the guards, scratch the paint or destroy any connectors...

muz post 2: I see what your getting at with the dial gauge. I do have a spare block I've been using for my dummy up, I can fit the box with no clutch to that block and take another look.

MrBishi - yeah this type of box has a integrated bell housing holding all the gears in place. It's possible to remove it, but I'd prefer not to.

Unfortunately I don't have a stock FJ20 gearbox at the moment, my the closest thing I have is the stock backing plate. This block backing plate fits really snug over the existing dowell in the block, but I understand what your saying that it could be out of whack.

[MrBishi]

Is dummy engine bare?
You could maybe attach dial gauge to input shaft with doodle resting on rear main bearing. Put gearbox in gear and turn output shaft and see what runout you have.

Or suck it and see.
BTW I sympathise with the 20+hrs for gearbox removal. My car entails removing radiator, cooler, manifolds, exhaust and sliding the motor forward. I should install with wing-nuts on everything next time.

[jmac]

To give some perspective - 3 thou runout can fuck the front pump of some autos. I doubt ANYONE could tell alignment by eye better than 20thou plus.

I should probably also point out that it might be that the factory block is also a touch out (it certainly happens with some, and as a big wake up call to many people - if you have the main bearing tunnel closed and honed, it shifts it 'upward' a few thou at least.

The only way to do it well is to check the actual box/engine/bellhousing combo you are running.

I know this the hard way, and it cost me a crank *(worth over a grand in todays money, probably a few grand). It's cost other people their transmissions. Now that I think of it I also ended up with a fucked gearbox that would jump out of 4th, because the input shaft was trying to spin in two different axes at once (teh centreline of the box and that of the crank) it chewed out the spigot bearing (not the spigot bush in the crank, the bearing that the input shaft has at the rear where the mainshaft 'nose' fits into it, and then

EVen if it is a nice snug fit on the dowel - it could be nice and snug in the wrong position!

To give you an idea of the likelihood of it, there's actually one valiant parts vendor who claims that it's 'impossible' to mount a torqueflite behind a bw35 patterned hemi 6 because of some magical harmonic issue. In fact it's all down to alignment. There's enough variance in std mopar blocks that the factory workshop manual details the dial indicator process, and states something along the lines of it being a required operation any time you mate a new trans/engine block to the other, only avoidable if they were originally a matched set (adn didn't have main tunnel machining at all)

You might find it next to impossible to use the dial indicator, about the only help I can think of is you might have to resort to using a small mirror and feeding it in through the starter motor hole and slowly reading it from there.

I have no doubt it's a HUGE pain to have to do this, about the only way you can maybe rationalise it is by taking some comfort in the fact that it'll take less time to do it right than to have to do it again (plus ongoing engine damage).

On the topic of some adapter bellhousings I've seen - won't name names, but if you bolted it up and measured it and got less than 15thou runout, you'd have to send it back to them so they could figure out where they went wrong.

re-reading it, I'm pretty sure it's the problem. the fact it made less noise when in motion and the clutch pedal is pressed helps 'seal' the deal. Likewise, there's a little bit more leeway with a sprung centre clutch vs a solid one (not that this necessarily applies to you, just for sake of completeness)..

[PLAYA]

I'm not expert but why are we talking about run out? The shaft isn't bent it is possibly just not aligned to the center of the crank?

[townie :)]

+11ty for the off centre shaft issue,would make the bearings whine in such a similar way,gotta be some kind of way engineers can measure this stuff, perhaps its a problem left to someone who has experience measuring this stuff/fabricating it. Would save you a heap in blown gearboxes, I can't imagine that you paid the regular sfa for this neo box.

[curmudgeon]

There's another method for establishing concentricity that's sometimes useful where access is difficult or impossible. Basically what you do is turn up a dummy shaft or mandrel that fits snugly and accurately into the front and rear gearbox bearing bores. The forward end would be turned down so it's a light tap in fit into the end of the crank where the spigot bush normally fits. The dummy shaft in effect becomes a central dowel. You'd bolt the bellhousing snugly to the block while it's centred and then drill and ream the dowel holes in both parts simultaneously.
BTW you don't need a big chunk of 3 or 4" round bar to make one of these; a couple of discs can be cut from say 1" plate and these then drilled and tacked to a piece of 1" brightshaft or similar. So long as all the od's are machined in one operation (after the discs have been tacked onto the shaft) it will be quite accurate.

[BigMuz]

Playa it's not runout- it's concentricity plus misaligned axes.

I'll find a drawing.

[BigMuz]

http://www.mech-engineer.blogspot.co...andout_13.html


In the case of the gearbox it is relatively easy to be sure that the shafts are not skwede because once you make the bellhousing you skim top and bottom on a milling machine.

So all you really have to worry about is concentricity. This is where the axes don't intersect. There is hopefully never any runout but if you mount an indicator on the outside of one shaft and you run it on the outside of the other shaft that will show the misalignment of them relative to each other.

In reality your best bet is that somebody made an accurate jig that you can use to drill the holes in the flange. And hopefully either by careful measuring and adjusting or some sort of witchcraft the axes of the crank and input shaft are coincidental within fuckall.

[townie :)]

I'm guessing that the misaligned axles would be made a million times worse when the clutch is clamped, as it would be pulling the middle of the input shaft in circles as it would have clamped off center.

I take it there was a few bolt holes that you didnt have to drill out or whatever, that people use as the guide?. Do you have the correct bolts in these, i.e. the ones with the bit of dowel slightly bigger then the thread to keep it centered on the holes?. At a guess also I would think this could possibly be a problem if many people have been doing that conversion with better success then you previously.

[260BC]

But also Richard, if you had the correct carrier, I would not space the pivot I do it as a fine tune as you can not keep 10 different carriers.

Yeh I looked at spacing there once, agree for fine tuning purposes only, the pivot does not have much thread.

On the concentricity question Tim, the box does line up using those points which are spaced the same as the original box. All that has to be done is to realign those few points which do not.

[tim510]

Yeh I looked at spacing there once, agree for fine tuning purposes only, the pivot does not have much thread.

On the concentricity question Tim, the box does line up using those points which are spaced the same as the original box. All that has to be done is to realign those few points which do not.

Thats right mate. The block has 4 points where the gearbox bolts to it. The lower 2 points fit 100% ok. The top 2 and the top dowel recess required re-drilling/adjustment.

Here's a shitty diagram :

[tim510]

I'm guessing that the misaligned axles would be made a million times worse when the clutch is clamped, as it would be pulling the middle of the input shaft in circles as it would have clamped off center.

I take it there was a few bolt holes that you didnt have to drill out or whatever, that people use as the guide?. Do you have the correct bolts in these, i.e. the ones with the bit of dowel slightly bigger then the thread to keep it centered on the holes?. At a guess also I would think this could possibly be a problem if many people have been doing that conversion with better success then you previously.

As per my shitty diagram, the lower bolt holes aren't touched.

Also this time I plan to run the lower gussets to brace the box further. I never put them on in the past as I am lazy + stupid.



So there will be 4 OEM mounting points which anchor the motor to the box unmodified.