setting the pinion shaft 1930 Hudson Super 8

[Bloo]

56 minutes ago, timecapsule said:

#16 is the "differential gear"

#17 is the "differential pinion"

#18 is the "differential spider"

Differential terminology is funny, and often used wrong.

 

Differential case: the piece the ring gear is bolted or riveted to

Differential carrier: the outer housing. That would be the removable center housing on a dropout style rearend, the whole housing on a late GM or Dana style rearend, etc. Whatever piece holds the outer race of the carrier bearings and those screw adjusters we have been discussing is the carrrier.

 

Nothing is wrong with the terminology of 16-18, but in modern lingo:

16: Differential side gears

17: Spider gears

18: Cross pin shaft  (maybe its really a spider shaped piece here? It is a shaft in many modern rearends)

 

33 minutes ago, timecapsule said:

You were asking about the axle and setting the end play.  If you look at #38 in the illustration, that is what's called a "thrust button"  So the axles touch each other and over time they wear down that thrust button on the end of each axle.  So the procedure is to add "hard facing" to the end of that thrust button.  Which I did, and now I am at between .008 and .009. 

Funny it should turn out to have thrust buttons. The Chrysler setup with the axle bearing shims on the axle bearing race was replaced in 1965 with a setup that used thrust buttons in the center. Adjustment motion was then fed through the middle, and adjustment was only needed on one side. They replaced the shims on that one side with a threaded adjuster, and used that setup for several more years.

 

29 minutes ago, timecapsule said:

Thanks for that warning of frustration.  That's one of the reasons I may pull it all out and do it on the bench.  The gas tank is right up tight against the back of the differential and I can't even see the adjusting nut on the ring gear side.  So working off a creeper would only add to the frustration I'm thinking.

Oh you DEFINITELY want this on the bench if there is any way to do so.

Edited September 5, 2022 by Bloo (see edit history)

[Oldtech]

9 hours ago, timecapsule said:

Here is that diagram I just mentioned.  #20 is listed as "differential adjusting nut".  But there just doesn't seem to be enough detail in the diagram to figure out what moves when the nut is adjusted.

#16 is the "differential gear"

#17 is the "differential pinion"

#18 is the "differential spider"

We may have missed explaining something. 

Number 20, which it calls a nut, IS a nut. It is a threaded collar that you have to turn to make the adjustment. 

[Oldtech]

10 hours ago, timecapsule said:

Here is that diagram I just mentioned.  #20 is listed as "differential adjusting nut".  But there just doesn't seem to be enough detail in the diagram to figure out what moves when the nut is adjusted.

#16 is the "differential gear"

#17 is the "differential pinion"

#18 is the "differential spider"

We may have missed explaining something. 

Number 20, which it calls a nut, IS a nut. It is a threaded collar that you have to turn to make the adjustment.  These "nut"s move the complete  assembly side to side. 

Edited September 5, 2022 by Oldtech (see edit history)

[Oldtech]

This MAY be of assistance. 

[timecapsule]

2 minutes ago, Oldtech said:

This MAY be of assistance. 

Yes thank you, I'll print that off right now before I head over to my shop. 

[Bloo]

It may... but it may not. Try it and see if it behaves like @Oldtech's chart. Post what you find. One takeaway from the chart that holds true is that you want that load out in the middle of the tooth, not out at the edges where it would have leverage to break teeth, yet still not too far down in the root.

 

You will see charts like this all over the internet that contradict each other. There is a document from Dana corp that explains all the contradictions. It is because there are 2 methods to produce the gears today (hobbing used to be the only way). In the Dana document, there are 2 charts like that, one for each manufacturing method. From memory, in/out on the pinion caused movement of the pattern in the opposite direction, but that might be wrong.

 

Additionally, when a chart like this says add or delete a shim, it is for some particular axle design, and could be opposite of yours. Removing/deleting might have the opposite effect on the in/out position of the pinion. That is a separate issue from the manufacturing differences. For example, adding a shim under your Hudson pinion-and-bearings assembly would move the pinion out. On an axle design that had shims between the pinion gear and the rear bearing to set depth, adding a shim would move the pinion in.

 

I am traveling, otherwise I would find the Dana document and link it. Unfortunately my Chevrolet/Pontiac spiral bevel gears didn't behave like either chart. Let's see what you come up with for patterns on those Hudson gears.

 

 

Edited September 5, 2022 by Bloo (see edit history)

[Oldtech]

Thanks Bloo. That's why I said it MAY be of assistance.

[timecapsule]

I didn't get as far as putting the pinion assembly into the differential assembly.  But I did get all the guts out of the car.  To be honest I'm scratching my head just as much now as I was when I was on the creeper looking at it while it was still in the car.  I don't know if I've done this right but I put the axles back in and secured them with a pair of pipe wrenches so it would keep the ring gear still so I could see where all the movement is ( .060+ th.)

Sorry but the picture is upside down.  I can never figure out why this site insists on rotating most of the picture.

[timecapsule]

Here is what the adjusting nut looks like

[timecapsule]

Just as the illustration shows, and as I suspected.  Unless someone can correct me and I'm not seeing clearly, all the adjusting nuts are doing is preloading the bearings.  The nut is tightening on the outer lip of the nut on to the race.  the rest of the assembly is part of the casting and it can't move. 

This image is the side that the ring gear is attached to I labeled the two arrows as R.  That is the bearing race. What I have labeled as AN is the adjusting nut and you can see the threads.  Next I'll add a picture of the opposite side.  Look at how much of the race is showing on the inside of that casting and how far in the threads are.

[timecapsule]

This is the opposite side and note how little of the race you see on the inside and how much of the threads on the adjust nut you see on the outside.

The issue I see here is if by mistake you were to tighten the nut too much, there is no way of getting that race back out.

[timecapsule]

Here is a shot from the end of the nut looking inside.  So the bearing is pressed on to the casting.  Inside that casting you can see the female half of the spline that the male spline on the axle fits into.  I have half a mind to believe that maybe the movement I have (.060th. ) could be the loose fitting between those splines. 

S is spline

B is bearing

C is casting

AN is adjusting nut

[timecapsule]

This shot shows the large gear that is part of the female spline I believe.  meshing with one of the 4 spider gears. If I hold the ring gear still and the axles are held firm, I can move the spider gear slightly.  To be honest it sort of wiggles almost like it's floating in there.  Not much but a little.

SG spider gear

DG differential gear

Edited September 6, 2022 by timecapsule (see edit history)

[timecapsule]

If you go back to the first picture of the assembly on the table with the axles held by the pipe wrenches and zoom in.  you can see there just isn't any place to adjust any back-lash ??

So the only adjustment that I can figure out that this assembly has is those large shims that go between the differential housing and the pinion housing.  But I'm led to believe they are only for adjusting where the pinion rides on the ring gear.  Correct?

[timecapsule]

In this picture I twisted a screwdriver between the housing and that differential gear and it did move sideways a little bit.  But I have no clue what that tells me.  Obviously there is no place to get any kind of a shim in there without disassembling the entire assembly.  I don't even know if it would be possible once you did that.  It would be great to have a spare parts car to tear it all down to see what's inside and what holds what together.

[Bloo]

44 minutes ago, timecapsule said:

Just as the illustration shows, and as I suspected.  Unless someone can correct me and I'm not seeing clearly, all the adjusting nuts are doing is preloading the bearings. 

Yes, that and they move the case from side to side. Am I missing something?

44 minutes ago, timecapsule said:

The nut is tightening on the outer lip of the nut on to the race.  the rest of the assembly is part of the casting and it can't move. 

I can't figure out what you mean exactly.

 

41 minutes ago, timecapsule said:

This is the opposite side and note how little of the race you see on the inside and how much of the threads on the adjust nut you see on the outside.

The issue I see here is if by mistake you were to tighten the nut too much, there is no way of getting that race back out.

I think you would remove the adjuster nuts and pry the case from side to side. Should be doable with the pinion and bearings assembly out. Do you mean if it fell completely in? On a modern rearend, the area where the outer race goes is split. The bolt(s) are at 90 degrees compared to yours, and half the bore the bearing rides in comes off like a main bearing cap from an engine. I don't know the answer there. My guess is it would come out with a little sideways prying, and maybe only  one would have to come out to get the case out.

37 minutes ago, timecapsule said:

I have half a mind to believe that maybe the movement I have (.060th. ) could be the loose fitting between those splines. 

That should not enter into it. On some cars, those gears wouldn't even be in there when you set the ring and pinion. We have got to start talking about the ring and pinion separately from the differential. I don't know if your differential needs work, but it is a completely separate affair from all this shims and backlash stuff for the ring and pinion.

 

On this ring and pinion, like most of them, you have exactly two adjustments. 1) side to side adjustment on the ring (by moving the case), sometimes called the backlash adjustment, and 2) in and out on the pinion, usually called pinion depth. Additionally you have bearing preload to worry about.

 

The adjuster nuts adjust the ring gear side to side and set the preload on the carrier bearings. The brass(?) shims shown in this picture set the pinion depth. The pinion bearing preload is adjusted with the smaller shims on the spline we discussed earlier in the thread.

 

 

Edited September 6, 2022 by Bloo (see edit history)

[timecapsule]

5 minutes ago, Bloo said:

Yes, that and they move the case from side to side. Am I missing something?

I can't figure out what you mean exactly.

 

I think you would remove the adjuster nuts and pry the case from side to side. Should be doable with the pinion and bearings assembly out. Do you mean if it fell completely in? On a modern rearend, the area where the outer race goes is split. The bolt(s) are at 90 degrees compared to yours, and half the bore the bearing rides in comes off like a main bearing cap from an engine. I don't know the answer there. My guess is it would come out with a little sideways prying, and maybe only  one would have to come out to get the case out.

That should not enter into it. On some cars, those gears wouldn't even be in there when you set the ring and pinion. We have got to start talking about the ring and pinion separately from the differential. I don't know if your differential needs work, but it is a completely separate affair from all this shims and backlash stuff for the ring and pinion.

 

On this ring and pinion, like most of them, you have exactly two adjustments. 1) side to side adjustment on the ring (by moving the case), sometimes called the backlash adjustment, and 2) in and out on the pinion, usually called pinion depth. Additionally you have bearing preload to worry about.

 

The adjuster nuts adjust the ring gear side to side and set the preload on the carrier bearings. The brass(?) shims shown in this picture set the pinion depth. The pinion bearing preload is adjusted with the smaller shims on the spline we discussed earlier in the thread.

 

 

If you study the illustration and the pictures.  You'll see that there just isn't any way to move that ring gear, sideways.  There are only two moving parts on that whole assembly.  The adjusting nut and the race that the adjusting nut pushes in.  The ring gear is bolted to the casting ( sorry I don't know the proper name)  It's not going anywhere.

[Bloo]

13 minutes ago, timecapsule said:

In this picture I twisted a screwdriver between the housing and that differential gear and it did move sideways a little bit.  But I have no clue what that tells me.  Obviously there is no place to get any kind of a shim in there without disassembling the entire assembly.  I don't even know if it would be possible once you did that.  It would be great to have a spare parts car to tear it all down to see what's inside and what holds what together.

It's probably normal. There is a spec on most cars for differential backlash (NOT ring and pinion, separate spec,separate parts). In quite a few old cars, there were bronze washers under the side gears or the spider gears or both. Sometimes they were selective fit. Maybe you could tighten things up a bit, maybe not. Later on, no more bronze. In that case, if the lash is too much you start replacing parts. The side gears run in a journal inside the case, and the spider gears USUALLY run on a removable shaft. Yours is a little different but I am not quite clear how they did it. I'm not going to say it doesn't matter, because slop matters everywhere, but compared to the voodoo of the ring and pinion, it doesn't matter.

 

Those side gear splines just float on the axle splines as far as I know. The axles have the end play they do because they grow with heat, and would probably cause rear wheel bearing failure if they didn't. The shims for that are out at the wheel bearings, on the outer races.

 

[timecapsule]

1 minute ago, timecapsule said:

If you study the illustration and the pictures.  You'll see that there just isn't any way to move that ring gear, sideways.  There are only two moving parts on that whole assembly.  The adjusting nut and the race that the adjusting nut pushes in.  The ring gear is bolted to the casting ( sorry I don't know the proper name)  It's not going anywhere.

Tomorrow I'll try to take more pictures that will show what I mean.

[Bloo]

1 minute ago, timecapsule said:

If you study the illustration and the pictures.  You'll see that there just isn't any way to move that ring gear, sideways.  There are only two moving parts on that whole assembly.  The adjusting nut and the race that the adjusting nut pushes in.  The ring gear is bolted to the casting ( sorry I don't know the proper name)  It's not going anywhere.

The casting the ring gear is bolted to is called a case. It has bearings pressed on it, and the outer races run in bores in the carrier. The adjusters push on the bearings. If you loosened the nut on one side one turn, and tightened the nut on the other side one turn, I believe you have moved the case sideways and the ring gear along with it. That is how most of these work.

[timecapsule]

Is it possible to add short 10 second videos on here?  That might help illustrate what I'm trying to say about everything being fixed in place other than the bearing race and the adjusting nut.

I've tried to imagine a complex machined part that somehow the adjusting nut is also pushing on that is a hidden part that runs past everything and pushes the differential gears closer to the spider gears but it just isn't possible. Plus you would think that you would see some telltale in the illustration.

[Bloo]

I am ONLY talking about the ring and pinion, not the differential. Those adjusters do not affect the differential.

 

They push the bearings side to side. The case, running on the bearings, moves from side to side. The ring, being bolted to the case moves side to side. This is one of your two available ring and pinion adjustments. The other is those brass colored shims in the picture above.

 

[timecapsule]

2 minutes ago, Bloo said:

The casting the ring gear is bolted to is called a case. It has bearings pressed on it, and the outer races run in bores in the carrier. The adjusters push on the bearings. If you loosened the nut on one side one turn, and tightened the nut on the other side one turn, I believe you have moved the case sideways and the ring gear along with it. That is how most of these work.

I did try to tighten one of the adjusting nuts once I loosened the other one, using a long small round steel punch.  But it didn't budge, so I may have to make up a special spanner like tool.  As I mentioned if nothing happens, than I've just put more preload on that bearing and I just don't see an easy way of bringing that race back out to where it was again, without going to a machine shop.  Right now the preload on both bearings seem to be set nicely.  

[timecapsule]

This may sound like a stupid question but does the pinion shaft/housing assembly need to be bolted in place with the proper preload on the pinion castle nut, in order to set the back-lash? 

[timecapsule]

14 minutes ago, timecapsule said:

I did try to tighten one of the adjusting nuts once I loosened the other one, using a long small round steel punch.  But it didn't budge, so I may have to make up a special spanner like tool.  As I mentioned if nothing happens, than I've just put more preload on that bearing and I just don't see an easy way of bringing that race back out to where it was again, without going to a machine shop.  Right now the preload on both bearings seem to be set nicely.  

If you look at one of the pictures above looking at the adjusting nut from the end where I have the blue arrows. In that picture you can't really see the race because it is behind the inner portion of the adjusting nut.  But you can see that the adjusting nut doesn't touch the bearing.  

I'm not trying to disagree with you and your knowledge.  I'm just saying that in this particular assembly that the adjusting nut only pushes on the race.  This is the first time doing this so I have zero knowledge,  I'm just trying to relay what I'm looking at.  No disrespect intended.

Edited September 6, 2022 by timecapsule (see edit history)

[Bloo]

You'll be moving those bearings back and forth a lot before this is over. I doubt they are tightly pressed. They might be, but they still have to move. It is the only way to make an adjustment.

 

32 minutes ago, timecapsule said:

This may sound like a stupid question but does the pinion shaft/housing assembly need to be bolted in place with the proper preload on the pinion castle nut, in order to set the back-lash? 

Yes, because the backlash you are setting is the backlash between the ring and pinion.

 

You'll have to get those pinion bearings set right first. The idea there is to shim it so that when it bottoms out hard on the shims with the nut tight, the preload on the bearings is right.

 

Then, in a perfect world, with a brand new ring and pinion, you would have a spec for pinion depth for the particular ring and pinion set you have, that the maker marked on it when they lapped it. Every set is a little different. You then put it in exactly the right depth, measured with some special tool, and then IN THEORY, all you should have to set is "backlash", in other words side to side on the ring, for best pattern.

 

In the real world, sometimes the carriers, etc. aren't machined quite perfect and you might have to move the pinion in or out a few thousandths to get a decent pattern. I might have mentioned an elephant in the bathtub earlier. This is that elephant. It is not as simple as adjusting backlash. We have a good guess where the pinion belongs in an in/out direction, but we don't really know. There is also wear because it is a used ring and pinion set. It may need more backlash. Getting a contact pattern that won't break teeth or cause howl is the goal. The possible combinations of pinion depth (in-out) and backlash (ring gear side-to-side) is almost endless.

 

As I mentioned before, If I were you, I would start out with all the brass colored shims in, and about .010" backlash, and see what the pattern is. Then try more or less backlash.

 

You need to create drag somehow. Paint about 4 ring teeth with gear marking compound, roll it around to they are going to contact the pinion, and create drag. The drag puts pressure on the teeth and makes a pattern in the paint. Maybe you jam a paint stick or something under the edge of the ring and turn the companion flange. Maybe you make the flange drag somehow and turn the ring. Whatever. Go both directions so you get the patterns for both power and coast (see @Oldtech's chart for an example).

 

Once you think you have a decent pattern, try it at about 4 places around the ring gear to make sure it's OK all the way around.

 

24 minutes ago, timecapsule said:

I'm just saying that in this particular assembly that the adjuster only pushes on the race.

On any axles I am familiar with, that is what they do. They push on the race. The race pushes on the rollers (or balls), The rollers (or balls) push on the inner race. The inner race is pressed onto the case, so when it gets pushed on, it pushes the case sideways. The ring gear has to come along because it is bolted (or riveted) to the case.

 

 

Edited September 6, 2022 by Bloo (see edit history)

[timecapsule]

I think something just clicked in my brain and it's making sense now (possibly) .  I keep worrying about if I move the adjusting nut too far in, that I won't be able to get it back out.  BUT, I see now ( possibly)  it will reset itself, so to say, to it's original position if I loosen the adjusting nut, and then tighten the adjusting nut on the other end, Right?  I keep referring to the "race"  but maybe I should be calling it a "sliding seat" for the bearing.  (lay terms of course)  I keep saying nothing can move.  But I wasn't thinking that the entire "inner guts" move back and forth by adjusting those nuts.   I'll check that possibility out tomorrow when I have it in front of me. 

[timecapsule]

2 minutes ago, Bloo said:

You'll be moving those bearings back and forth a lot before this is over. I doubt they are tightly pressed. They might be, but they still have to move. It is the only way to make an adjustment.

 

Yes, because the backlash you are setting is the backlash between the ring and pinion.

 

You'll have to get those pinion bearings set right first. The idea there is to shim it so that when it bottoms out hard on the shims with the nut tight, the preload on the bearings is right.

 

Then, in a perfect world, with a brand new ring and pinion, you would have a spec for pinion depth for the particular ring and pinion set you have, that the maker marked on it when they lapped it. Every set is a little different. You then put it in exactly the right depth, measured with some special tool, and then IN THEORY, all you should have to set is "backlash", in other words side to side on the ring, for best pattern.

 

In the real world, sometimes the carriers, etc. aren't machined quite perfect and you might have to move the pinion in or out a thousandth or two to get a decent pattern. I might have mentioned an elephant in the bathtub earlier. This is that elephant. It is not as simple as adjusting backlash. We have a good guess where the pinion belongs in an in/out direction, but we don't really know. There is also wear because it is a used ring and pinion set. It may need more backlash. Getting a contact pattern that won't break teeth or cause howl is the goal. The possible combinations of pinion depth (in-out) and backlash (ring gear side-to-side) is almost endless.

 

As I mentioned before, If I were you, I would start out with all the brass colored shims in, and about .010" backlash, and see what the pattern is. Then try more or less backlash.

 

You need to create drag somehow. Paint about 4 ring teeth with gear marking compound, roll it around to they are going to contact the pinion, and create drag. The drag puts pressure on the teeth and makes a pattern in the paint. Maybe you jam a paint stick or something under the edge of the ring and turn the companion flange. Maybe you make the flange drag somehow and turn the ring. Whatever. Go both directions so you get the patterns for both power and coast (see @Oldtech's chart for an example).

 

Once you think you have a decent pattern, try it at about 4 places around the ring gear to make sure it's OK all the way around.

 

On any axles I am familiar with, that is what they do. They push on the race. The race pushes on the rollers (or balls), The rollers (or balls) push on the inner race. The inner race is pressed onto the case, so when it gets pushed on, it pushes the case sideways. The ring gear has to come along because it is bolted (or riveted) to the case.

 

 

Thank you, that last paragraph you wrote sort of agrees with what I just posted.  Sorry for being so thick headed.

[Bloo]

It is indeed the outer race of the bearing those adjusting nuts push on. Yes, you'll probably have to push it back and forth and try a few settings.

[timecapsule]

1 minute ago, Bloo said:

It is indeed the outer race of the bearing those adjusting nuts push on. Yes, you'll probably have to push it back and forth and try a few settings.

I might as well wait until the new outer pinion race and bearing arrive.  ( Should be here maybe tomorrow or Wednesday) Then I'll get the old outer  pinion housing  race pulled and the new one pressed in and then start experimenting with adjustments.

[timecapsule]

Well I finally got the bearing, race and seal for my pinion assembly and so I rolled up my sleeve and got the pinion assembly together without any issues and gave it a preload of about 15 in. lbs.  Then I fastened it to the differential carrier with the original  6 shims.  3 are .010th. and the other 3 are .001 th.  I tried a number of variation of shims and different back-lashes and I just wasn't getting it right. 

Long story short the ring gear is running out of axis to the pinion shaft from .000th to .040th 180 degrees around the other side.  It took awhile to figure out what was going on. At one point I added  4 painted teeth at about 180 degrees to the first set I painted.  That's when I realized, the paint was wearing not to bad on the one set of painted teeth,,, but weren't even leaving a mark on the other painted teeth .  How is this possible?  Looks to me like everything is original. The nuts and bolts holding the ring gear in place have the tabbed locking washer-strips still firmly in place. It's tight to left side of the housing case as well. 

So now I'm wondering what my options are??

[Oldtech]

On 9/5/2022 at 9:40 PM, timecapsule said:

In this picture I twisted a screwdriver between the housing and that differential gear and it did move sideways a little bit.  But I have no clue what that tells me.  Obviously there is no place to get any kind of a shim in there without disassembling the entire assembly.  I don't even know if it would be possible once you did that.  It would be great to have a spare parts car to tear it all down to see what's inside and what holds what together.

 

 

That could be a gnarly problem. Looking back through the old pictures I see a possibility. There looks like there may be a gap in the 2 halves possibly caused by the differential pinion shaft either being too large or not in correctly.  If not, the 2 most likely spots are this join of the 2 halves of the gearcase or where the crown gear is attached.  You may have DPM (dreaded previous mechanic) issues. 

[timecapsule]

5 minutes ago, Oldtech said:

 

 

That could be a gnarly problem. Looking back through the old pictures I see a possibility. There looks like there may be a gap in the 2 halves possibly caused by the differential pinion shaft either being too large or not in correctly.  If not, the 2 most likely spots are this join of the 2 halves of the gearcase or where the crown gear is attached.  You may have DPM (dreaded previous mechanic) issues. 

Once I discovered this, I studied the assembly super close and to be honest I don't see anything that looks like it's not right.  I see the gap that you're referring to.  I'll have another look at that tomorrow, but it didn't seem to set off an alarm bell when I was trying to see what was wrong.  The other thing that's weird is that there isn't any wear on either the ring gear or the pinion gear. 

Edited September 15, 2022 by timecapsule (see edit history)

[Oldtech]

The picture in the quote. Look at the junction of the 2 halves just below the shaft.  I tried to mark one but it didn't show.  The picture may be deceiving though.

Edited September 15, 2022 by Oldtech (see edit history)

[timecapsule]

I

1 minute ago, Oldtech said:

The picture in the quote. Look at the junction of the 2 halves just below the shaft.  I tried to mark one but it didn't show.

Yes I see it now.

[timecapsule]

I mean if I went to the effort of trying to put various thicknesses of shim stock washers between the ring gear and the left case, With a bit of effort I could probably get most of it out. However then the ring gear would not be surface to surface held in place.  Would that cause issues?  Balance issues?  Strength issues?

 

 

[Oldtech]

I can't see how anything could be bent to throw it off .040 thous. I would be looking closely at the assembly. If it is really out that much I would try to get a machine shop to put it in a lathe and machine the surface that the crown gear goes on.  Can the 2 halves be put together 180 out? (I doubt it.) Does it turn smoothly in the bearings with a bit of preload?

[timecapsule]

The other option would be to set the back-lash at about .008th when the ring gear is closest to the pinion, and try to get the wear marks as close to optimal on those teeth, and then check it after a few hundred miles.  

[timecapsule]

Just now, Oldtech said:

I can't see how anything could be bent to throw it off .040 thous. I would be looking closely at the assembly. If it is really out that much I would try to get a machine shop to put it in a lathe and machine the surface that the crown gear goes on.  Can the 2 halves be put together 180 out? (I doubt it.) Does it turn smoothly in the bearings with a bit of preload?

It turns smoothly for about 180 degrees.  Depending on where I set the back-lash. 

[timecapsule]

The other strange thing, although I don't see the connection, is the adjusting nut on the right side ( opposite to the ring gear side) is a bear to move clockwise or counter clockwise.  Where as the left side, once I loosen it about one notch is basically loose and I can spin it counter clockwise with my fingers.  I've looked closely at the threads on the stiff one and they seem fine.