Limited Slip Differential
A simple add-on to a regular differential, that locks it automatically whenever a slip situation occurs.
Those of you who have tried building and driving off-road vehicles with differentials, are probably familiar with the so-called slip situation. It happens when one of the wheels driven via the differential slips, or loses contact with the ground. The regular differential transfers then the entire torque to that wheel, leaving the other wheel stopped. As result, the entire axle stops, which may easily stop the vehicle completely.
Lego differentials can be locked with the driving rings, manually or by a motor, but I wanted something smaller and fully automated. The solution is simple – all it takes is to add another pair of semi-axles to the ones that go through the differential, and place a pair of cams that are not connected to each other between them. As long as both wheels maintain equal speeds, the cams remain separate. When the speeds differ, the cams will eventually lock against each other, thus disabling the differential and forcing both wheels to run at equal speeds.
Please note that due to the simplicity of the design, no adjustment of cams position is possible. Therefore, they may lock occasionally while e.g. the vehicle is turning, but that should not cause much trouble, because the difference between wheels’ speeds is insignificant in such case. A perfect situation for the mechanism would be to have the vehicle turn right just as much as left – the cams would then return to their initial position. Since such situation is unlikely to happen, this mechanism may, in fact, slightly degrade the vehicle’s performance by disabling the differential while turning. At this cost, you get a 100% guarantee of being saved from the slip situation.
Moreover, please note that the efficiency of the mechanism may differ when it’s used for a differential on the axle, from when it’s used for a central differential. I have not tested the latter case, but I suppose this solution would be useless for the central differential, because of the difference of the travel done by steered and unsteered axles while turning. Such difference tends to be constant, and thus the cams would have no chance to return to the initial position. The same would happen if the mechanism is used in a normal axle, but the vehicle keeps turning only left, or only right.
@qwertyuiop
Yes, of course, except that 40 tooth gears are rare and very large, they will decrease the ground clearance quite a lot.
If you put an 8 tooth gear on the differential and a 40 tooth gear on the cams then it drastically increases the time required for the cams to lock. I tested this on a small vehicle with 43.2 wheels, with a turning circle of about 1 metre, and it turned about 3/4 of a full circle before the cams locked. Similarly, you can change the gear ratio to lock the cams faster or slower by using other combinations of gears.
@Badachelli
when the axles are turning hte same direction the rubber band is pulled in one direction: lock does not move.
slight difference in speeds causes axles to move in opposite directions: lock swing slightly to one side.
major difference causes lock to swing enough to lock into the gears and limiting the difference in speed.
On the note that the differential will lock when turning: You could make the gears around the differential very small, and the gears running the lock very large. This would cause the lock to take a very long time to lock while turning, but lock in seconds once slip occurs. The rare times at which it will lock when turning will only take seconds to purposely stop a wheel to get the lock back to the original position.
@Sariel I saw on youtube (http://www.youtube.com/watch?v=R-TV8PYnGuE) this guy making an RC mod to the 8297 Off-Roader set, and he was putting in a “Limited slip” differential (stop the video a 1:10) and it seems that he was using rubber bands? I don’t know. I dint understand how what he is doing would work. Have you seen anything like this before?
@aj
That shouldn’t really depend on the type of differential used.
can make a new video using the new differential because im having slips with the new one
I think what this really is is an automatic locking differential, as after a point, no slip is possible. However, I liked the idea and decided to build a modified version, taking slightly from Nic’s idea of using the slipper gears, but without the slippers and their drawback of limiting the torque that is possible to transmit.
I used the older 4-stud wide differential, as that is all i have, and then ran two different mechanisms off of the half axle. On the one, I used three linear 16-tooth gears, and on the other, I used a belt drive. both sides are connected to the same axle at the far end, similar to the locking differential on this website (http://sariel.pl/2009/03/3-studs-wide-differential-lock/). The belt allows a slip to occur, so the two alxes don’t have to run at the same speed, but limits the ease of slipping, which can be adjusted by using different rubber bands. Also, because the power is still transmitted through the differential (by a worm gear in my case), the torque that is possible to transmit is not limited by the system. The resulting system is slightly larger than the original proposed on this page, but includes all the bracing required, and can probably be refined to something smaller. it also uses fairly common parts and should be easy to build. I will try to get some pics latter and post back with a link.
@Tony
Not really, I’m afraid.
is there any way for me to get the instuctions or blueprints on your car design so I can make one for my Classroom at UVU (Utah Valley University)
@legofan
http://www.brickshelf.com/
You told that there are other better limited slip differential on the internet… Could you give me a particular site/s where i could find them.
Thanks
actually we want clutch type limited slip differential
You are right, it will not lock completely. I am not sure how the real ones works but some slip may be still better than nothing. Anyway, I have to test this in real life to see whether it makes sense.
@Dornbi
You’re right about the racetrack, I had Truck Trial tracks in mind when designing this. A linear clutch won’t lock the differential completely, I’m afraid.
I think that the equal amount of turns left or right may be a problem, for example on a circular track or any closed racetrack. Wouldn’t it be possible to use the linear clutch presented in one of your earlier posts?
@Nic That is theoretically correct, but I have discarded the idea of using slipper gears in the drivetrain of a Truck Trial vehicle. Not only it is very likely to fail to move the stopped wheel (note that this wheel supports the entire axle in such a situation, which means half of vehicle’s weight in case of 4×4 cars), but it will keep consuming part of the torque applied to differential, just to keep slipping.
Instead of a full locking diff could you maybe replace the semi axle with a solid axle and change the solid gears for the 24T slipper gears. It would theoretically (as i haven’t tried it myself) allow the wheels to spin at different speeds through a corner due to friction between the wheels and the ground but should stay solid if one wheel is raised off of the ground (will more than likely only work for lighter models as friction in the slipper gear is fairly low). Anyway just a thought, would have tried myself but i don’t have any of the slipper gears lying around
You are perfectly right. It’s clear that this mechanism doesn’t work perfectly, but it certainly serves its basic purpose. In my country, some Truck Trial builders remove differentials completely, so this could be of some use to them, as occasional lock while turning is still better than having no differentials at all.
A very interesting idea, Sariel! At first I thought that the problem of remaining locked after a single use would render this pointless, but since you explained the idea of equal left and right turns, it makes more sense. Perhaps if the cams were geared to go slower, the effect could be more easily avoided (I’ve taken a screenshot and uploaded it, just click my name). While you would have to wait a little longer for the cams to engage, there would be more tolerance for steering. Of course, once the cams have engaged, it will take longer for them to draw fully apart, so it isn’t perfect.