- First Name
- Rick
- Joined
- Sep 9, 2020
- Threads
- 1
- Messages
- 351
- Reaction score
- 953
- Location
- Rochester, NY
- Vehicle(s)
- 2013 F150 STX 5.0L, 1999 Mustang Cobra
- Your Bronco Model
- Badlands
- Thread starter
- #1
This is actually the first thread I've started here. In case anyone is interested, I put together this look at workings of the M220 e-locking differential. I got my hands on a sample part through work for development purposes. I thought it would be interesting to show it in depth, going through a tear-down of the assembly.
So hereâs the differential as it stands complete, as you might remove it from the axle (bearings removed):
The wiring connection goes into the axle casing above and to the right of the differential, thereâs a pass-through connector that it plugs into from the inside.
At the âtopâ of the assembly (in our business, we consider the flange-end of the unit to be the bottom) sits the locking solenoid. This is an electromagnet and has a sliding armature inside it that will push on a dog clutch mechanism inside the diff casing when activated.
The solenoid itself is easy to remove, in fact, the Ford service manual for the axle shows that you remove the solenoid when you replace the differential. It will lift off of the top end once the journal bearing is removed; the bearing is basically the retainer for it. With the solenoid removed, there is a thrust washer and weâd call an âapply plateâ remaining underneath where it sat. You can see those below:
The apply plate snaps onto the differentialâs locking plate with little spring tabs. The apply plate looks like this:
Its just a .06â thick steel stamping, but it serves the purpose of connecting the solenoid (which is fixed from rotation inside the axle â otherwise the wiring would wrap up) and the lock plate inside the diff which, of course, spins at road speed. The plastic thrust washer that sits above it provides a low friction sliding surface to translate the motion of the solenoid armature to the apply plate.
So, if we take off those parts, weâre left with just the differential itself. Fundamentally, itâs a basic open differential design. However, it has 4 pinion gears (sometimes called âspiderâ gears) to connect the output gears to the casing. Typical open diffs have 2 pinion gears, so using 4 effectively doubles the strength of the gearing itself by spreading out the load across twice as many teeth at any given moment. So thatâs a nice feature.
It might be nice functionally, but doing that adds a bit of complication (read: cost) to the design. For starters, the differentialâs casing has to be made from 2 pieces for assembly purposes. Most open diff casings are 1-piece (and thus dirt cheap to make), and all the gears can be loaded through the holes in the sides.
In this design, the casing is split through the ring gear flange and relies on the ring gear bolts to hold it all together when assembled. This is the end cap piece after removal:
Removing that exposes the bevel gearing:
And the by lifting out the flange-end output gear, you can see the 4 pinions riding on their cross pins.
Having 4 pinions requires a more complicated cross pin â you can see it has one full-length pin, and two âhalfâ pins, which are guided in a cross-drilled hole through the full pin. So, you can see where the part and assembly costs are increased.
If we further remove the cross pins (BTW, the half pins are retained with a spirol pin thatâs driven into the casing from the flange end, you can see the holes they go in near the â1â and â3â marked on the casing), you can see into the housing with the remaining output gear in place. Removing these pins is a pain in the butt, which I am not detailing here. Here we see inside the casing with the pinions removed:
This gear is part of the lock device, but the lock plate that is operated by the solenoid is underneath it:
Also visible in this photo is the return spring, which holds the lock plate away from side gear when not in use. This is just a simple wave spring.
And then, finally, removing the lock plate from the assembly leaves an empty diff housing.
Note the five slots in the end of the housing, these are important in a couple of ways, which Iâll return to. The lock mechanism itself is about as dirt-simple as it could be. It uses what is called a âface clutchâ, which has sort of like spline teeth that run radially outward instead of being parallel along the sides of a shaft. These âdogâ teeth allow a lot of contact surface when engaged, with the benefit of only having slide a few millimeters to achieve a full engagement.
You could do the same thing with a sliding collar over a conventional spline, but it would have to translate over a greater length to have good strength. The picture to the left shows the face clutch in the relaxed position, to the right is engaged.
The actual teeth on the face clutch have a slight angle to them, so that they can drop into engagement easily when commanded. They also allow the spring to quickly pop them back out when the lock is turned off.
Of course, making it easier to pop out adds a concern about coming out when you donât want it to. This is where the slots I referred to before have a role.
The slots work in conjunction with the protrusions, or what I call âtowersâ on the back side of the lock plate. The primary function of these towers is to create a torque flow path from the casing to the lock plate, and then on to the output gear when the lock is used. Normally, torque in a differential flow from the ring gear to the casing, then through the cross pin to pinion gears. The pinion gears transfer it to the output gears, and on out to the axle shafts.
However, when you lock the unit, you create an alternate torque flow, where it flows from the case to the lock plate, and to one output gear. It then is transferred to the other output gear through the pinions. The towers engage in the slots in the casing to allow that transfer, like a kind of keyway.
The towers and slots have a secondary function, which prevents the lock plate from accidentally popping out of engagement with the lock output gear. Note the shape of the towers â they have a 15-degree taper to the sides, which is match by a taper in the slots.
This is a rather clever bit of engineering - by having a taper angle to them, when any torque is applied to the differential at all in the locked mode, the tapers act as a sort of wedge cam. The load on the taper forces the lock plate to thrust into the mating clutch face on the gear; it canât back out until the load is removed.
This also means that the solenoid doesnât have to try to prevent unintended disengagement. Itâs held in place mechanically.
Here is more detail of the components â to the right, gears laid out after removal from the housing. Below is all of the assemblyâs parts spread out.
At the bottom of the page, we have the solenoid up close. The brown ring on the underside of it is the armature I mentioned above. It slides inside of the solenoid when activated, moving a total of about 3mm from the relaxed to engaged positions. As I mentioned before, the solenoid itself is not a rotating part. It has a retaining bracket that is held in place by the bearing cap bolts, which in turn keys onto the pin you can see in the 7:00 position in the photo at the bottom right. The solenoid has a molded bushing on its inside diameter, which allows it to ride on the differential casing while the casing spins inside it. Note that the plastic bushing on this particular part was damaged in shipping (Ford's service part group could do a better job packaging)...
I am not going to get into details of how the computer controls work for this, as thatâs above my pay grade. But on a high level, the control module sends about 5 amps or so to the coil to activate the lock. It holds that current for a minute or so to make sure it has time to engage, then cuts the current down to maybe 2 amps to help hold the lock, but without drawing enough power to overheat the coil. When you (or the ECU) decide to disengage, it cuts the power, and the return spring pops the lock plate out of mesh just as soon as it isnât torque bound. Overall, this whole system is pretty simple from an engineering perspective - very effective and reliable. There is little to fail. For the lock to physically break, youâd have to either shear the dog teeth off or the towers off.
Iâve run through the math; those features both have about a 30% safety factor above the shear strength of the axle shaft itself. And that, BTW, assumes that only 40% of the dog teeth actually carry significant load, and that only 3 of the 5 towers are pulling their weight. Past that, the electromagnet canât really fail unless a wire is cut. The solenoid mechanism really must work if current is applied the to the coil, thereâs little choice from a physics perspective. The only other failure mode is then if the ECU that operates takes a dump. But that could be field fixed by rewiring the solenoid to a simple switch and relay.
Then consider that a design like this will essentially drop into the existing axle packaging. Aside from a wiring port, there are no extra pieces of hardware required, no plumbing to run, no servos or shift motors to mount to make it work. Thatâs why most truck manufacturers have adopted this or something similar. The locking differential in the F-150 is essentially the same design, though there are differences in how itâs packaged.
Although the M220 axle is made by Dana, I think that this e-locker is sourced from GKN. The tapered tower design is their patent feature. Ford also buys the F-150 lockers from GKN as well. Dana uses Eaton e-lockers in their aftermarket Jeep axle assemblies, so it could be that Ford directed them use the GKN design here. The Eaton design has a similar operating principle but relies on sliding pins into a lock plate instead of the face clutch as I recall. There are a lot of different variations on the theme out there, but hopefully this gives a good overview of how they work.
So hereâs the differential as it stands complete, as you might remove it from the axle (bearings removed):
The wiring connection goes into the axle casing above and to the right of the differential, thereâs a pass-through connector that it plugs into from the inside.
At the âtopâ of the assembly (in our business, we consider the flange-end of the unit to be the bottom) sits the locking solenoid. This is an electromagnet and has a sliding armature inside it that will push on a dog clutch mechanism inside the diff casing when activated.
The solenoid itself is easy to remove, in fact, the Ford service manual for the axle shows that you remove the solenoid when you replace the differential. It will lift off of the top end once the journal bearing is removed; the bearing is basically the retainer for it. With the solenoid removed, there is a thrust washer and weâd call an âapply plateâ remaining underneath where it sat. You can see those below:
The apply plate snaps onto the differentialâs locking plate with little spring tabs. The apply plate looks like this:
Its just a .06â thick steel stamping, but it serves the purpose of connecting the solenoid (which is fixed from rotation inside the axle â otherwise the wiring would wrap up) and the lock plate inside the diff which, of course, spins at road speed. The plastic thrust washer that sits above it provides a low friction sliding surface to translate the motion of the solenoid armature to the apply plate.
So, if we take off those parts, weâre left with just the differential itself. Fundamentally, itâs a basic open differential design. However, it has 4 pinion gears (sometimes called âspiderâ gears) to connect the output gears to the casing. Typical open diffs have 2 pinion gears, so using 4 effectively doubles the strength of the gearing itself by spreading out the load across twice as many teeth at any given moment. So thatâs a nice feature.
It might be nice functionally, but doing that adds a bit of complication (read: cost) to the design. For starters, the differentialâs casing has to be made from 2 pieces for assembly purposes. Most open diff casings are 1-piece (and thus dirt cheap to make), and all the gears can be loaded through the holes in the sides.
In this design, the casing is split through the ring gear flange and relies on the ring gear bolts to hold it all together when assembled. This is the end cap piece after removal:
Removing that exposes the bevel gearing:
And the by lifting out the flange-end output gear, you can see the 4 pinions riding on their cross pins.
Having 4 pinions requires a more complicated cross pin â you can see it has one full-length pin, and two âhalfâ pins, which are guided in a cross-drilled hole through the full pin. So, you can see where the part and assembly costs are increased.
If we further remove the cross pins (BTW, the half pins are retained with a spirol pin thatâs driven into the casing from the flange end, you can see the holes they go in near the â1â and â3â marked on the casing), you can see into the housing with the remaining output gear in place. Removing these pins is a pain in the butt, which I am not detailing here. Here we see inside the casing with the pinions removed:
This gear is part of the lock device, but the lock plate that is operated by the solenoid is underneath it:
Also visible in this photo is the return spring, which holds the lock plate away from side gear when not in use. This is just a simple wave spring.
And then, finally, removing the lock plate from the assembly leaves an empty diff housing.
Note the five slots in the end of the housing, these are important in a couple of ways, which Iâll return to. The lock mechanism itself is about as dirt-simple as it could be. It uses what is called a âface clutchâ, which has sort of like spline teeth that run radially outward instead of being parallel along the sides of a shaft. These âdogâ teeth allow a lot of contact surface when engaged, with the benefit of only having slide a few millimeters to achieve a full engagement.
You could do the same thing with a sliding collar over a conventional spline, but it would have to translate over a greater length to have good strength. The picture to the left shows the face clutch in the relaxed position, to the right is engaged.
The actual teeth on the face clutch have a slight angle to them, so that they can drop into engagement easily when commanded. They also allow the spring to quickly pop them back out when the lock is turned off.
Of course, making it easier to pop out adds a concern about coming out when you donât want it to. This is where the slots I referred to before have a role.
The slots work in conjunction with the protrusions, or what I call âtowersâ on the back side of the lock plate. The primary function of these towers is to create a torque flow path from the casing to the lock plate, and then on to the output gear when the lock is used. Normally, torque in a differential flow from the ring gear to the casing, then through the cross pin to pinion gears. The pinion gears transfer it to the output gears, and on out to the axle shafts.
However, when you lock the unit, you create an alternate torque flow, where it flows from the case to the lock plate, and to one output gear. It then is transferred to the other output gear through the pinions. The towers engage in the slots in the casing to allow that transfer, like a kind of keyway.
The towers and slots have a secondary function, which prevents the lock plate from accidentally popping out of engagement with the lock output gear. Note the shape of the towers â they have a 15-degree taper to the sides, which is match by a taper in the slots.
This is a rather clever bit of engineering - by having a taper angle to them, when any torque is applied to the differential at all in the locked mode, the tapers act as a sort of wedge cam. The load on the taper forces the lock plate to thrust into the mating clutch face on the gear; it canât back out until the load is removed.
This also means that the solenoid doesnât have to try to prevent unintended disengagement. Itâs held in place mechanically.
Here is more detail of the components â to the right, gears laid out after removal from the housing. Below is all of the assemblyâs parts spread out.
At the bottom of the page, we have the solenoid up close. The brown ring on the underside of it is the armature I mentioned above. It slides inside of the solenoid when activated, moving a total of about 3mm from the relaxed to engaged positions. As I mentioned before, the solenoid itself is not a rotating part. It has a retaining bracket that is held in place by the bearing cap bolts, which in turn keys onto the pin you can see in the 7:00 position in the photo at the bottom right. The solenoid has a molded bushing on its inside diameter, which allows it to ride on the differential casing while the casing spins inside it. Note that the plastic bushing on this particular part was damaged in shipping (Ford's service part group could do a better job packaging)...
I am not going to get into details of how the computer controls work for this, as thatâs above my pay grade. But on a high level, the control module sends about 5 amps or so to the coil to activate the lock. It holds that current for a minute or so to make sure it has time to engage, then cuts the current down to maybe 2 amps to help hold the lock, but without drawing enough power to overheat the coil. When you (or the ECU) decide to disengage, it cuts the power, and the return spring pops the lock plate out of mesh just as soon as it isnât torque bound. Overall, this whole system is pretty simple from an engineering perspective - very effective and reliable. There is little to fail. For the lock to physically break, youâd have to either shear the dog teeth off or the towers off.
Iâve run through the math; those features both have about a 30% safety factor above the shear strength of the axle shaft itself. And that, BTW, assumes that only 40% of the dog teeth actually carry significant load, and that only 3 of the 5 towers are pulling their weight. Past that, the electromagnet canât really fail unless a wire is cut. The solenoid mechanism really must work if current is applied the to the coil, thereâs little choice from a physics perspective. The only other failure mode is then if the ECU that operates takes a dump. But that could be field fixed by rewiring the solenoid to a simple switch and relay.
Then consider that a design like this will essentially drop into the existing axle packaging. Aside from a wiring port, there are no extra pieces of hardware required, no plumbing to run, no servos or shift motors to mount to make it work. Thatâs why most truck manufacturers have adopted this or something similar. The locking differential in the F-150 is essentially the same design, though there are differences in how itâs packaged.
Although the M220 axle is made by Dana, I think that this e-locker is sourced from GKN. The tapered tower design is their patent feature. Ford also buys the F-150 lockers from GKN as well. Dana uses Eaton e-lockers in their aftermarket Jeep axle assemblies, so it could be that Ford directed them use the GKN design here. The Eaton design has a similar operating principle but relies on sliding pins into a lock plate instead of the face clutch as I recall. There are a lot of different variations on the theme out there, but hopefully this gives a good overview of how they work.
Sponsored
Last edited:
