We were going to move on to the XM262 before circling back to the XM249 (the ultimate winner), but despite having devoted two posts to the Rodman Labs ingenious XM235/XM248 LMB, we’re still getting questions about it. Many of them relate to the feed mechanism, and why it’s better. Well, along with the patent for the gun itself, three of the Rodman engineers took out a patent specifically on the feed mechanism: 4,061,074 A, which notes it’s “also published as 3,999,461” — the main XM235/248 patent! Here’s 4,061,074A at Google Patents, but we’ll try to explain it in this post for you. Here’s the image that is on the first page of the patent, see if you can figure out what’s happening here. Muzzle’s to the left, belt entering from the bottom at left. Some key parts (these numbers are used in all the drawings):

  • 10, 20: lower and upper receiver tube/gas tube (there are two, upper and lower, and they’re identical, interchangeable parts. Each contains a gas piston and return spring).
  • 12: feed cam assembly. This contains a slot to drive it in rotary motion, and a row of seven saw – or gear-teeth on the nose to engage matching teeth on the sprocket (32) in ratchet fashion.
  • 22: bolt carrier
  • 24: bolt. This is much like an AR or AK bolt, with three locking lugs.
  • 28: feed sprocket. This is rotated clockwise (when viewed from behind the gun) by the feed drive cam.


This is a complex design and it has a lot of parts, but if you’ve ever detail stripped any rimless-round GPMG’s feed tray (M60, M240/MAG, MG42/MG3) you know there are a lot fewer in this kind of gun than in that, and even fewer of the parts are the sort of small fiddly parts that find the most out-of-reach place under the storage locker when you’re reassembling your M240. Indeed, the patent cites that as a big advantage:

Conventional self-powered machine gun feed mechanisms are typically of the reciprocating transverse variety. This type of mechanism incorporates a feed lever, generally located near the top of the weapon, which pivots in a rather large feed cover and tray member. As the bolt carrier reciprocates it moves the feed lever back and forth. The feed lever actuates a pawl which advances belted rounds of ammunition to the feed position. The rounds are held in this position by another pawl, which is also controlled by the feed lever. This is a complex mechanism consisting of many moving parts, springs, etc. For example, there are 75 parts in the feed mechanism of the M60 machine gun now in use. Obviously a mechanism of this type would be more susceptible to failure or breakage than one containing fewer parts.

Here’s Fig. 1, which is those same parts in an isometric view. This may make “how it works” a little more evident to you.


You can see that a hinged door (26) opens on the side, which the inventors also saw as a big advantage:

Another disadvantage of this prior art system is that the feed cover is pivotally mounted at its forward end and must be pivoted upward in order to load or clear the weapon. This is a shortcoming because, from the prone position, a gunner would have to expose too much of himself to accomplish the loading or clearing function, to his possible detriment.

Figure 2 appears to have been the same one we’ve already seen:


Figure 3 shows the feed cam, which comprises a driven part, the shaft (45) that is acted on by the bolt carrier, and the driving ratchet (48)’s teeth (50) that act upon the driven teeth (52) on the belt feed sprocket (28).


It’s not clear from this illustration, but the patent makes clear that the ratchet (48) can move in the axial, fore-and-aft direction and is springloaded to stay in contact with the drive teeth on the sprocket.

Figure 4 shows the feed mechanism from forward, looking aft:xm235_248_feed_fig4

The mechanism is simple, driven quite directly by the bolt carrier, and does not induce translational moments laterally, unlike a traditional feed tray mechanism. It also probably induces less significant roll moments about the longitudinal axis of the gun. This would contribute to increased accuracy, particularly in multishot bursts.

via Patent US4061074 – Ammunition feed mechanism – Google Patents.

Ah, but what about the bolt carrier? What’s with those two funky cam-slots visible in Fig. 2 of the above patent. We’ve scarcely mentioned it, is there anything special about it? Turns out, there is.

Which brings us to Patent 4,022,105A. Google Patents page.

Once again we’re clearly looking at the mechanism of our old friend, the XM235.


This time the muzzle is to the right. A key item here is the drive pin (20) which connects the bolt carrier to the solid gas pistons that ride inside the receiver tubes. The two pins that bear on the slots in the bolt carrier are #40, which rotates the bolt, and #44, which cycles the firing pin.


And it turns out the purpose of the bolt carrier’s multiple slots is rather simple. One does exactly what you would expect: it rotates the bolt to lock and unlock it in the barrel extension (which has a similar role here as the barrel extension on an M16, although this is a three-lug bolt). The other slot deald with the firing pin and sear. As the mechanism recoils and the bolt carrier slides back on the receiver tubes, the firing pin’s cocking pin rides in a slot which cocks it and locks it in place as the mechanism recoils. If the trigger has reset, the cocking pin is caught; otherwise, it is tripped and released by the slot as the weapon gets back into battery.


One particular feature of this bolt carrier is that it has a good bit of “excess” movement fore and aft, beyond the bare minimum needed to allow the system to function. The purpose of this is to allow the bolt carrier to absorb excess energy and burn time, slowing the rate of fire to the desired 500 rounds per minute. There’s a good bit of reciprocating weight, but it’s reciprocating right along the longitudinal axis of the firearm. These things taken together should produce a firearm with less unwanted dispersion of rounds.

This entry was posted in Machine Guns, The Past is Another Country, Weapons Education, Weapons Technology on by Hognose.

About Hognose

Former Special Forces 11B2S, later 18B, weapons man. (Also served in intelligence and operations jobs in SF).

5 thoughts on “The SAWS that Never WAS, part 3b: the feed of the XM248


Sweet thanks for posting this, I was wondering how the feed mechanism worked, i figured it worked something like the DSHK. This is definitely a lot simpler than that

Hognose Post author

The DShK is pretty ingenious in its own way. It would be interesting to compare the 1946 (I think) version (with that rotary feed you mean) with the much less common earlier one. It would also be interesting to compare the DShK with its much simpler replacement, the NPV, and look at what Ivan gained with the change. (A whole bunch of hours of machine tool time freed up, for one thing).


I’m pretty sure that the HK belt-feds like the HK21 also use a sprocket to drive the belt. It seems like a neat, sensible way to pull a belt, but I have little experience with belt-fed stuff.


These technical firearm posting series are awesome. I love them.

One suggestion though: you should create a section of the blog with links and descriptions of these. The existing archives, organized by date of posting are well suited to the majority of the postings as they relate to current events. These postings, being essentially reference material, would be great to have in a quickly accessible area that would highlight the other ones you have done like the toggle locking series.

Thanks for writing these.


Hognose Post author

That’s a great idea, Pete. The XM262 one is taking a while to get right.