Yours is on the bench without axles, and axle bearings in the wheel area. I didn't see directions for this project on the bench.

<img src="http://www.peachparts.com/shopforum/attachment.php?attachmentid=151851&d=1553820544">

Exactly. This is the issue I am having, but it seems that setting my pinion bearings to the high end of the settings should be close enough to compensate for any drag from the side bearings. It will at least be a more correct setting than what they were at before.

There may be two issues with your evaluation.

The length of a torque wrench handle is altered because your socket isn't centered with the pinion shaft. That alters the moment arm. Imagine using two pry bars. One is 1 ft in length the other significantly longer. Which will generate the most leverage. Imagine sitting half way to the pivot point, center, of a seesaw with your seesaw buddy being the same weight sitting all the way back to his end of the board. The distance from the middle is significant

The second factor is that there's a significant load generated by wheel bearings, friction on axles when leveled resulting in them sitting on jacks. Imagine how much force is needed to rotate a tractor trailer wheel as opposed to the force required to move a 12in wheel on a lawnmower.

What you have here, I believe to be a situation that the FSM does not address. I have certainly exaggerated the previous examples but only to share my concerns of the concepts involved.

These factors may be is significant when the measurements are so delicate as a Ncm. That's the rotational friction unit which a dentist uses as instructions when he screws a tooth implant into your jaw.

Best of luck.

The socket not being centered on the axis of the pinion shaft will not be a problem as the torque wrench will be attached to a socket over the pinion nut when I check the friction. I also checked it with the wrench over the center before to see if there was any noticeable difference, and there was not. But, it's a valid point.

The factory service manual covers checking the friction with the differential installed in the vehicle and the axle shafts in place. This same situation would apply for a differential that's out of the vehicle and on a bench. Where I am having a problem is the friction I was reading was less than what it should have been even with the gear set out. The reading with the gear set in place should be the same or higher than with it out, not less. So, this indicates the friction in my differential was far too low, and therefore could not have been used as a point of reference.

Had I had the differential installed in the vehicle and used this method, I would have gotten a higher friction reading due to the axle shafts being attached and turning wheel bearings and possibly wheels, however it would have been false security as when I went to set the friction to that same amount, I would have also been setting it too low (without knowing it).

Setting the friction to the previous value essentially allows for replacing of the pinion seal without changing the friction value, however, if the previous friction value was too low due to wear, it will be too low also when going back together. So, the method of setting friction back to the previous setting is only a good method if the previous setting was correct, which is just an assumption or hope. Now, if the differential has lasted a long time at the "too low" setting, it may well last a long time further at the same setting, and removing the differential may not be practical in all situations, but I'm glad I decided to remove it so I can get it more correct than it was.

The factory service manual also covers setting the friction with the differential out of the vehicle, but with the gear set removed, so that is a different situation than what I have before me, however, I can safely say that I do not want the friction set any lower than what it specifies without the gear set in place, as it is on mine. I can't know how much friction the gear set and side bearings add, but if I set the friction to the high end of the range I should be close, and I've read from others that have removed their differentials and axle shafts, but have left their gear sets in place and have checked the friction before removing the nut, that their friction was typically in the middle to high end of what the manual states it should be for used bearings without the gear set in place. So, this means that either the gear set and bearings make a negligible difference in friction, or after normal bearing wear it puts them into that range. So, I should be safe to set mine into the high end of the range.

I definitely don't want the friction with the gearset installed to be lower than what it's supposed to be without the gearset in place. Now I'll be able to get it closer to the correct setting and hopefully that will help it to last a little longer than it would have had I put it back to the previous setting.

I put the nut on the pinion, without the yoke in place, to see how much friction there is with no preload on the pinion bearings. It was exactly the same value as when the nut was on, so there was no measurable preload on the bearings before, or the bearings may have actually had slop.

So, 25ncm is how much friction there is due to the side bearings, gears, and unloaded front bearings. I think I could add 25ncm to the recommended bearing settings and come out spot-on, but I might still refrain from exceeding the 100ncm limit just to be safe. Even if 25ncm is the correct value for the friction of the gears and side bearings and setting the friction at 100ncm actually puts the pinion bearing friction at 75ncm, that's still in the middle of the range, which is fine with me.

Also, were KOYO bearings used from the factory in Mercedes differentials in 1981? That's what are in mine, and I wonder if they are original.

I would not take the carrier and pinion out. I have not done so in a MB, but have replaced carrier bearings and pinion bearings in a jeep. Setting up gears is not for the faint of heart, and learning by trial and error means lots of checking, adjusting, rechecking, more adjusting, swearing...

Put the seal in, slap some locktite on the threads, tighten the pinion nut to get you back in spec and install it. I would assume the pinion preload is measured with the carrier installed but without the axles unless specified otherwise.

Without preload on those bearings, the gears will be noisy at best and broken at worst.

I'm trying to avoid removing the gears. The pinion nut gets staked in place, so no thread locker is needed. The factory service manual says to measure the preload with the pinion gear in place, but not with the carrier gear, so therein is the confusion because I sure don't want to remove the carrier gear so I can do it the way the book says to do it.

The book also says it can be done on the vehicle, in which case you are measuring the amount of force it takes to turn the pinion before you remove the nut, and then tighten up the nut so the friction is once again the same. The flaw with this method is if the setting was wrong before, it will be wrong again. In my case, my pinion bearings did not have any preload, so if I had followed this method they would have once again ended up with no preload.

In my situation, there was no amount of preload for me to use as a reference, because my bearings didn't have any preload, and I couldn't set the bearings according to the book's specifications because the carrier gear has to be removed in order for that to happen.

Ultimately, the carrier gear and side bearings don't create much more than 25ncm of drag, so even though the book says to remove them in order to set the preload on the pinion bearings, leaving them in place isn't likely to make any real difference, so if a person wants to play it safe, they can still set them at 100ncm which would put the amount of friction at just the pinion bearings at around 75ncm, which is right in the center of the 50-100ncm range.