Gear box gurus

[Snagger]

Hi folks. I hope someone may be able to help with some issues.

I have rebuilt my gear box after third gear was stripped. The box worked mostly fine on refitting, but jumped out in the over-run in fourth.

Suspecting that the main shaft and primary pinion may not have been perfectly in line (a worn spigot bush having caused the same symptoms when I fit the Tdi), I stripped the unit again and rebuilt it with new bearings for the entire main shaft and a new spigot bush, assembling the entire transmission as standard before hoisting into the vehicle - last time I assembled the transmission in sections onto the engine, ad the play in the pinion bearing may have allowed the bell housing to be fitted slightly misaligned to the flywheel housing. Another new spigot bush was also fitted.

Now, when the main shaft rear nut is tightened the last quarter turn, it locks up the gear box. With it slightly loose, the main shaft turns happily. I suspect that for some reason, the first gear end float has disappeared, so tightening the nut is clamping the first gear between its thrust washers. Any other ideas?

[secondjeremy]

You could try using Loktite on the nut in a position where wverything turns nicely and see what happens.

By and large if the nut is loose a S3 will jump out of 1st and 2nd on the overrun. S2's (non-synchro) don't seem to suffer much - probably as the engagement is by sliding gears rather than the small movements of the synchro units.

[Snagger]

I thought I had put a second post on this thread. Maybe I pressed the wrong button and deleted rather than posted it!

Anyway, I pulled the unit back out ( ) and stripped it back out. The 1st gear bush looks good, but appears to have worn the steel thrust washer where it is in contact with the bush, allowing the bush to sit slightly into the washer and reducing the end float below acceptable limits, binding the gear wheel between the washer and the synchro hub. I have sourced a new bush from Rogers, but they tell me the thrust washer is no longer available, genuine or pattern. I ordered one from an online retailer that said on the phone that they have them in stock, and that it should be with me on Friday. Fingers crossed... If they are unavailable, I'll have to polish the old one back a bit (the area that contacts the gear wheel seems unblemished) and use the new bush with a bit more length to counter the washer's recess.

Frustrating, but I should have checked it first time round rather than just concentrating on the broken third gear and replacing that gear's bush.

[bill van snorkle]

"Jumps out on overun in 4th''. Are you sure you have 3/4th gear synchro assembly the right way around? They are designed with some teeth thicker than the others. the extra thickness is biased to the overun side of the tooth so that they grip the teeth on the gear with greater pressure on overun than if all the teeth were the same thickness. There is recess machined into one end of the splined section of the synchro unit. this recess must be fitted facing 3rd gear.

[Snagger]

Yep Bill, got it the right way around - the flat end faces third and the end with a "nose" faces the pinion; get it the wrong way around and you won't get third engaged because that nose will press against the main shaft snap-ring.

I collected the bronze bush and it is longer than the old one, but closer reading of the manual states that the bush should have some end float. The thrust washer presses against the ends of the main shaft splines, not the bush, and there is no way of adjusting end float of the 1st gear. It makes sense, really - with the compression of tightening the main nut working through to the bush, the bush would want to spread outward and seize the gear wheel again, this time from the inside. So, the fault is not in the bush, just the thrust washer. The wear on its inner face is from it spinning against the splines; I would assume that when the main nut in the Roverdrive coupling slackened off (a recurring problem), it allowed the thrust washer to spin on the main shaft as it was no longer clamped firmly between the ends of the splines and the rear main bearing race. I might assemble it with a light blob of thread lock to bind it to the splines.

The Roverdrive coupling is hollow and much longer than the Fairey part, so uses a thick collared hex nut on the main shaft inside of the coupling, with two grub screws passing through the sides of the coupling (clutch sleeve equivalent) to clamp the nut. The trouble is that the small amount of spline play allows the grub screws to drag the nut loose like a ratchet wrench, bit by bit on each torque reversal and spline movement. Using the really strong thread lock on the main nut and standard thread lock on the grub screws, will all threads thoroughly cleaned, helps. It should have been included in the manual.

[greenmeanie]

The Roverdrive coupling is hollow and much longer than the Fairey part, so uses a thick collared hex nut on the main shaft inside of the coupling, with two grub screws passing through the sides of the coupling (clutch sleeve equivalent) to clamp the nut. The trouble is that the small amount of spline play allows the grub screws to drag the nut loose like a ratchet wrench, bit by bit on each torque reversal and spline movement. Using the really strong thread lock on the main nut and standard thread lock on the grub screws, will all threads thoroughly cleaned, helps. It should have been included in the manual.

I had well over 100K of very hard and very,very hot miles on my Roverdrive before I managed to kill it. Never ever had a problem with that mainshaft nut coming loose though I did take the time to torque it properly.

[Snagger]

I had well over 100K of very hard and very,very hot miles on my Roverdrive before I managed to kill it. Never ever had a problem with that mainshaft nut coming loose though I did take the time to torque it properly.

As did I - it's not hard to set 100'lbs o a torque wrench and do the grub screws up a bit at a time to make sure they're balanced. I think a lot depends on how well the splines engage - any play causes the problem I had. Like I said, though, the strong thread lock seemed to stop it.

[greenmeanie]

The point being you have a worn spline on your gearbox yet think it is the resposibility of the Overdrive manufacturer to include notes in their manual to deal with your problem. The Roverdrive design works so why make your problem their issue? If you have play in the spline it will generate other problems in time beyond just the nut coming loose so fix the root cause you have identified rather than bodging it.

[Snagger]

The point being you have a worn spline on your gearbox yet think it is the resposibility of the Overdrive manufacturer to include notes in their manual to deal with your problem. The Roverdrive design works so why make your problem their issue? If you have play in the spline it will generate other problems in time beyond just the nut coming loose so fix the root cause you have identified rather than bodging it.

Well, the original transfer box gear fits without any play, but not the coupling. There really isn't much wear on either part, and nor would you expect it at less than 40,000 miles on both transmission parts (the main shaft was replaced in the gear box rebuild that kicked off the car rebuild). Besides, I've always been a huge advocate of the Roverdrive - I think it's a very robust unit and it's a huge leap forward over the Fairey. The movement of the coupling on the shaft with the nut loose is tiny, and so it takes a long time to slacken a tightened nut, but it has on occasion done that. The thread lock, like I said, though, seems to deal with it just fine.

[bill van snorkle]

Whilst Fairey overdrives were a pile of junk, the use of grub screws by Roverdrive as a locking medium for the mainshaft nut just sounds wrong. Is there a washer between the coupling and the nut so that play in the splines doesn't work the nut back and forth?

[Snagger]

Whilst Fairey overdrives were a pile of junk, the use of grub screws by Roverdrive as a locking medium for the mainshaft nut just sounds wrong. Is there a washer between the coupling and the nut so that play in the splines doesn't work the nut back and forth?

No. What they have is a a long coupling (roughly twice the length of the original gear or Fairey coupling) with internal splines, narrow on one end to fit the gear box main shaft, and wider at the aft end to fit the Roverdrive's input shaft. The step from one set of splines to the other is quite wide, and is the shoulder that the collared nut presses against. The shoulder of the nut collar and the step are slightly bevelled to help centralise the coupling on the shaft, which suggests that the tolerances of the splines are designed to be a little loose. Two grub screws, aft of the needle bearing on the coupling's exterior (like on the Fairey coupling) go through the coupling and against the nut's collar.

They had to do it this way because the coupling is so long that it extends well past the nut and any lock washer would be impossible to bend over (or remove later). The trouble is that the nut is locked to the coupling, not the shaft and its threads, and if the grub screws don't bite quite enough, then they'll drag the nut loose but won't drag it tight again on the next shaft torque cycle, acting like a micro ratchet, and the coupling's splines are a little looser than the original gear's so that spline movement is unavoidable. The play has got worse because the splines in the coupling and on the shaft have suffered, but it was always slightly loose, and the main shaft was only about 5,00 miles old when the Roverdrive was first fitted. Then again, I do have the first unit sold in the UK after they made their new molds, so this slight play may not affect later units. I did the grub screws up tight enough that they have marked the nut collar quite clearly, but maybe they needed to be tighter. I don't recall a torque being specified for the grub screws and I don't have a torque wrench that would do it anyway - I only have a 3' torque wrench for things like that main nut, head bolts and so on. I didn't want to strip the threads or mangle the Allen key holes, and I used the Allen keys pretty hard. I don't know how much tighter I can turn them without damaging them...

I think it's the one draw back with the unit. If there was a way of locking the nut directly to the shaft, I think it'd be bulletproof.

[greenmeanie]

<p>No washer</p>

<p><a href="http://roamerdrive.com/pdfs/ssparts.pdf">http://roamerdrive.com/pdfs/ssparts.pdf</a></p>

<p> </p>

<p>I used to design rotating equipment and we would run 200KVa gennies up and down the rpm range through full load to zero power transients up to 27000rpm with nothing more than correct torque and typically two to three stakes to hold the shaft nuts in place. In this case you can't support he opposite end of the shaft so you can't stake without running the risk of brinneling the bearing on the mainshaft. Loctite is not very reliable unless you get all the oil out of the joint which can be quite difficult, lock tabs are out because of the depth of the bore.</p>

<p> </p>

<p>Set screws with a hardness greater than the nut and the right point to bite into the nut will work almost as well as staking although the tightening sequence to get it truly straight is a pain. In this case, however, I would be tempted to loctite the set screws as the coupling is easily removed and properly cleaned in the <strike>dishwasher</strike>, sorry parts washer.</p>

<p> </p>

<p>I had no more play in the splines than the stock LR transfer gear so it sounds like you have something out of tolerance there. As the engine turns clockwise (Viewed from front)  it means viewed from the rear the shaft turns anticlockwise. This means the use of a conventional RH thread is correct on the end of the shaft. As a result your ratcheting effect should also come into play as power is applied from take off and most torque is used in acceleration from rest as opposed to power on/of transients once the vehicle is moving so it, in theory, would ratchet tighter rather than looser. Now, of course, if your set screws are not biting hard enough to prevent this they will move and wear a groove in the nut and that opens a whole new kettle of fish.</p>

[Snagger]

I was going to use electrical cleaner aerosol and a tooth brush to clean the main nut threads, so they'd be completely clean before the coupling was slid on )splines on the shaft pre-greased).

The grub screws on my coupling aren't pointed; they are concave and leave circular imprints in the nut. There are also some scrub lines where they have scratched the nut as the nut has progressively unwound. You're absolutely right that engine torque will tend to turn the transmission in the direction that the coupling and grub screws would retighten the nut, but I think what has happened is that under engine braking, over-run and reverse gear, the torque is reversed and the coupling has dragged the nut slightly looser (with no thread lock), which is easy enough, and then as normal torque has been applied, the screws have scraped across the nut because of the higher effort in tightening the nut, and repeated tiny movements of the ut in this way over a period of months has undone the nut by 45 degrees or so until the torque required to do the nut up the same amount as it was backed off is small enough that the grub screws stopped slipping. Hence the ratchet wrench analogy.

I did try using just the thread lock on the big nut and omitting the grub screws, with grease between the coupling and nut shoulders to allow it to slip without dragging the nut, but it only worked for a few months. I then tried fitting the coupling with paint marks to align the same splines and marking the nut through the grub screw holes to cut grooves for the grub screws to nose into so that they couldn't slip around. That worked better.

A production tolerance issue on the shaft or coupling is plausible - if the splines were both at the narrow limit of production tolerances, then together they'd be magnified and would create the play I had on both new components. The mainshaft is now clearly worn on its splines, and perhaps the coupling is too, but the coupling spines have no visible wear pattern on them because they engage on their entire edge, unlike the main shaft which is unworn at the ends of the splines. I wonder how much of that wear is normal at under 40'000 miles, and whether much of the whole problem is attributable to the increased torque (in both directions) and vibration of the Tdi engine driving it all... I can't help but think all that banging back and forth must have accelerate the wear and is at least a contributory factor in the 1st gear thrust washer wear, which can only have happened because the washer wasn't pinched hard enough by the nut at some point.

Of course, once that thrust washer started wearing, it could wear faster and faster, and that would have left the rear nut looser without it even turning, so everything then starts screwing everything else up..

I'll try reassembling as per the instructions, all oiled except the grub screws, and do them up as tight as I dare. Were your grub screws pointed to penetrate the nut collar, then?

[greenmeanie]

Sounds like you need to read ANSI B92 to understand your spline fits.

Some comments

The torque on over run is far less than take up of drive from stopped.

Vibration will undo the nut far better than your cycling theory.

You should consider how you applied counter force to achieve what you think was the correct torque. The further up the drivetrain you jambed it he further out your applied torque will be. Each shaft and joint acts like a spring and you have to tighten against all of those plus the running torque before you get to the locking torque. Consider that running torque accounts for over 80% of applied torque and it doesn't take much to get it wrong. Unless you actually managed to defrom those few threads at he joint the nut will not actually lock and now you are hoping that loctite will do the job.

You should also understand that by applying the loctite you have changed the torque requirement for the nut. It typically reduces the coefficient of friction reducing the running torque meaning the torque you have applied is more likely to have achieved the locking torque than before. Of course, you could have also damaged the threads this way with too much torque but hey, its all a guess unless you know how to play the game properly.

If you really want to prevent a reoccurrence of the problem. Clean up the threads, follow proper torque rules to tighten the nut, mark the nut relative to the coupler and apply the set screws to mark their location on the circumference and take it all back off. Now apply a small flat bottom hole for a blunt tip set screw or a standard drill bit for a pointed set screw to create a bore a couple of mm deep. Now retorque the nut to the original marks and put the set screws in with a little loctite. She will NEVER come loose again.

[bill van snorkle]

Thanks for the explanation on how the coupling and grub screws are arranged guys.

Interesting MeanGreenie that you got high mileage from your unit. probably has something to do with relatively low ambient temperatures. I have read that they don't fare quite so well over here in our generally hotter Outback conditions unless an oil cooler has been fitted, particularly if regularly split shifting the indirect main gears when towing. My only experience with planetaries is with my crawler box and hub reductions on MAN and Merc prime movers, and they do get quite hot for the relatively low speeds they run at. I personally would have liked to see the Fairey concept further developed with twin layshafts and detail work to improve lubrication of the coupling splines.

[greenmeanie]

Thanks for the explanation on how the coupling and grub screws are arranged guys.

Interesting MeanGreenie that you got high mileage from your unit. probably has something to do with relatively low ambient temperatures. I have read that they don't fare quite so well over here in our generally hotter Outback conditions unless an oil cooler has been fitted, particularly if regularly split shifting the indirect main gears when towing. My only experience with planetaries is with my crawler box and hub reductions on MAN and Merc prime movers, and they do get quite hot for the relatively low speeds they run at. I personally would have liked to see the Fairey concept further developed with twin layshafts and detail work to improve lubrication of the coupling splines.

Bill,

I lived in Phoenix Arizona and summer ambient ranged from 110-120°F. It was thrashed up and down the I10 from Phoenix to Tucson and back for 1000miles every week through sun rain and dust storms. Now I'll confess that the Roverdrive got VERY hot but it hung together for a good while. I did spend a lot on oil changes though and played with synthetic for a while which did help although the gearbox did feel more notchy as I think it was just a bit too slippery for the synchros. That EP90 got to smell REALLY bad after a month.

Adding the finned bottom plate for the TC does make quite a difference in temperature and is a must for running the truck hard. Then again, on many of those very long runs on the extremely boring I10 I spent a good amount of time designing an oil cooler system and pump to add to the TC to keep her cool.

Now I'm in the procvess of finishing up a Chevy 292/NP435/series TC powered SIIA 109 for those long hauls across Europe to get back home to Scotland. No, fuel economy is not a concern.

[Snagger]

I applied 105'lbs, as per LR's instructions, to the nut, using 4th gear to lock the shaft up (fitted with the transmission in the vehicle). The nut, shaft and grub screw threads were all fully cleaned up with all traces of threadlock scraped out and the grub screw threads chemically cleaned too before refitting this morning, and the grub screws were tightened as much as I could manage with the 4mm Allen key (alternating as I went), with medium strength thread lock to them only and the main nut was put on dry. We'll see how it goes.

Frankly, I still don't know the cause of the jumping out of 4th, but with all the bearings along the main shaft axis, from spigot bush to rear bearing replaced, the previous pinion (which didn't jump 4th) refitted and with the transmission assembled and installed with the standard rear bearing in place instead of the overdrive coupling leaving it unsupported, I've done all I can. Ideally, I'd like to replace the main shaft and overdrive coupling, but with costs are racking up too high and they're still serviceable, if less than perfect - I should get another 40,ooo out of them yet.

[bill van snorkle]

I mentioned on the other gearbox thread that a misaligned flywheel housing to bellhousing interface has on many occasions been the cause of popping out of 4 th gear. A popular engine swap over here used to be the GM Holden 6 cylinder ranging through 149, 161, 179, 186 and 202 cubic inch capacities. Some of the conversion flywheel housings were truly horribly machined, being up to 0.20'' out of concentricity with the crank shafts. When fitted to used gearboxes, sometimes they wouldn't pop out of 4th gear, but if a gearbox was reconditioned with new bearings, popping out was virtually guaranteed. I would guess that during my career, I would have 'dialled in' around 50 Holden conversion flywheel housings to rectify that problem.

[Snagger]

Yep, I remember that< Bill, and took a close look when I mated the bell housing to the flywheel housing. The studs have less than 0.5mm clearance through their holes, so while it's not as precise as having locating dowels, it should be pretty damned close. That's assuming that the DIscovery flywheel housing and Series bell housing are both meant to be circular with the studs concentric with the flywheel/pinion. The flywheel housing had most of the studs where I needed - just one had to be removed and refitted in a blind hole adjacent, and the four bolts that go through the bottom of the housing into the ladder block between sump and block were replace with countersunk bolts as there were no holes in the bell housing for them (and the flange is bevelled, so new holes weren't possible), with three new studs fitted in the flywheel housing bottom blind holes between those four bolts, which already lined up with the series bell housing stud holes. The only machining or working was countersinking those four bolts and tapping the threads in the blind holes.

It has worked most of the time, the only previous issue being caused by a worn spigot bush. This time, the cause may be something else as the bush seemed OK, but I have since replaced it again anyway. It could be that the other pinion has a fault (could be bent, or have a machining fault), or it could even be an issue with the new clutch plate. That pinion looked to be in better condition than the existing one, which in itself was not bad, but it does look very slightly different - the front end has rounded edges over a much bigger radius than the small bevelled edge of the tip of the original. Both are Gen Parts, so why they differ I don't know (and yes, both are Suffix D with the ecm gears and notched synchro teeth).

[Snagger]

Well, apart from a little backlash if the pedals are used clumsily, it (tentatively) seems to be good. It was very smooth and quiet in all gears with the overdrive coupling fitted but overdrive unit still off. With everything except the prop shafts, it gets progressively more chattery from the transfer box output shaft in third and fourth, but has done that for the full 20 years I've had it, so I think that's normal. It drove well, with no nasty noises, no jumping gears and no notchiness. The stick movement in third has almost disappeared, too.

Given the main shaft and coupling wear, I know this is not a permanent job, but as long as nothing comes loose (like that damned nut), then it should last a good few years. Long term, I may consider a total transmission replacement with a stumpy R380, with all options open - the Ashcroft mating kit to keep the series transfer box, an LT230 or even a Borg Warner unit.. I'm even beginning to consider an auto box, which could be sourced with a Borg Warner from a scrap RRC. The plan is to restore the RRC first and then I can look at more significant mods on the 109 to remove these weak links in the chain.

[Snagger]

Had to drive from Bedford to East Mids airport and back yesterday, as well as to Luton and back the day before. It seems to be running fine. I still think I might upgrade at a later date when I'm not needing the car as a daily driver - a quieter transmission would be rather nice, and trying to get back some of the raised overall gearing would be beneficial; 60mph is an effective top speed for noise levels, even though it's wind itself up to a bit over 80 if I push it hard (I really don't like it though, as it probably isn't doing the engine or transmission much good).

So, the future options will be:

LT77/R380 with Series transfer box and 4.1:1 diffs;

LT77/R380 and LT230, 3.54 diffs;

LT77/R380 with Borg Warner and 3.54 diffs;

RRC 4 speed auto with Borg Warner.

I quite like the Borg Warner units, having one in my RRC. While they lack the option of fully locking the centre diff to get you home if you have to remove a prop shaft or half shaft, and could damage the viscous coupling if you drove in that condition for a long way (and the couplings are pretty expensive), and they do scrub tyres more in daily use (I get half the mileage out of the BFG ATs on the RRC than on the 109), I like the way it drives and the fact that it gives such good performance on roads with slippery patches. I quite fancy an auto, too... Anyone know if they fit a Discovery 200Tdi engine? I assume that some RRC 200Tdis had automatics. That'd be a big job. The easiest option would be the Series transfer box and 4.1 diffs with a 5-speed box - no chassis mods, no custom props, just the Ashcroft conversion kit, some new engine mounts and fiddling with engine piping to move it forwards, and the diff work which seems pretty straight forward, having done it once already.

Anyone know about getting a Series speedo to work with an LT230 or Borg Warner unit?

[TheRecklessEngineer]

Anyone know about getting a Series speedo to work with an LT230 or Borg Warner unit?

The triangular series speedo cable fits in the square LT230 drive gear well enough for it to work.

[lexi]

I love auto boxes and have had loads in cars. I would miss the old cog stirrer on a Series though.

Mine is fun and limited use mind you. Good to see you still driving and thinking your special truck through it's evolutions though.

[Gremlin]

If i had to change again (lt77 now) (not that it will happen soon!) My choice due to the damn traffic here would be Auto, with my lt230.

G

[Snagger]

Anyone know if the ZF has a short version, comparable in length to a short LT77 or stumpy R380, or are they all long for the Discovery and RRC? I assume they fit a Discovery 200Tdi without difficulty, given that most 200 powered RRC had them... Are they different from the V8 coupled ZF (torque converters, bell housings, flywheel housings, etc)?