Why Drum Magazines are a Bad Idea

I often see questions form people asking why drum magazines are not more widely used – the BAR and the MP40/I in particular. People often view the box vs drum magazine question simply in terms of capacity – where the drum is obviously superior. However, there are several other elements to the question, and drums don’t fare so well with most of them. Drum magazines are heavier, more expensive to make, and much more complicated to carry than simple box magazines. Drum magazines are a tempting idea, and they periodically pop up in military usage around the world. However, what I find telling is that virtually no military adopted two drums in a row. The US dropped the Thompson drums, the Finns went form a drum to a box magazine for the Suomi, the Soviets dropped the PPSh-41 drum in favor of a box mag on the PPS-43, and neither the Lewis nor DP was followed by another gun using their pan magazines.


  1. Russian infantrymen carrying PPSh’s might go into action with a loaded drum in the gun, and switch to stick mags when that was empty.

      • It’s not quite a speculation. Drum magazines for PPSh were fitted to individual gun, and drums above the basic 2\gun were intended as replacement for lost\damaged mags (and still had to be hand-fitted by the armorer). So if you needed more than 140 rounds you definitely had to use sticks.

        Multiple memoirs do mention loadouts like “2 drums and 2/4/6 horns (rozhkov)”.

        • The Suomi and PPSh drums I’ve played with (so a small and unrepresentative sample of drums in total)

          Were a pain to mount, even under perfect conditions

          There didn’t seem to be any flared funnel to help guide the drum attachment points into place on the guns.

          But then, that does seem to be a failing of most interchangeable mag guns.

      • No, not speculation. I’ve read it in accounts of WW2. I shouldn’t have said “might.” Should have said “sometimes did.”

    • I suspect that hey carried one mag in gun and that was it. Those moving with tanks did not last long enough to change mag. Never seen a spare one at their belt anyway.

  2. Were the drums dropped due to cost or reliability issues or simply because of weight and trasportability?

  3. There were remarkable glitches in the German manufacturing monolith of the 1930s and 1940s — Krupp never quite mastered the screw breech, and consequently their artillery needed fixed ammunition up to enormous calibers, like 240mm. Another weakness was spring tempering; they were bad at it. See the incredibly extended recoil spring on the MP/StG44 rifles. And thus no German drum magazines after World War I. The snail magazines for the Mondragon rifle and P08 were brief exceptions (and incidentally required two kinds of spring!)

    • Manganese shortages, not inability to heat-treat. Also,
      semi-auto breech system, and obturation and flash safety is better with a case; see also case head stub on modern MBT ammo vs. De Bange greased asbestos behind bagged charge. Even Gustav, cal 80cm, used a case for the main charge, and also 60cm, 40cm, 38cm, 30.5cm, 28cm etc etc. “Zere is nussing wrong vis OUR bluddy ships twoday, ja?”.

      • It wasn’t a flaw, it was a choice. Going back to the late 1860s, with the advent of the metallic cartridge, Krupp opted for the sliding block breech/rimmed metallic cartridge combination on grounds of simplicity of manufacture, plus durability and reliability in service.

        The rotating breechblock/bagged charge breech was largely a French invention, although the original “pipe cap” style rotating breech block was the work of Joseph Whitworth in Britain, used with a metallic cartridge. The main French contribution was de Bange’s “elastic” breech seal, the one we’re familiar with on things like 16″/50 naval guns using bagged charges from WW2. Although the perfected version relies on ductile metals rather than the material de Bange started with; suet. (!)

        Krupp (and later Rheinmetall) opted for the block breech and metallic cartridge because it gave just as good or better a breech seal than the de Bange type, but had the additional attraction of allowing a higher rate of fire when using fixed or semi-fixed ammunition.

        The Prussians sort of surprised everybody in 1870-71 when their metallic cartridge breechloading field guns, even without actual recoil-absorption systems, proved superior to anything the French had in the Franco-Prussian War. It was the drubbing the French Army, especially the artillery branch, took in that one that resulted in the French Model 1897 75mm field gun; the French were determined not to come out on the short end of a paradigm shift in artillery technology a second time.

        Ironically, the “secrets” of the “French 75”, its rapid-acting Nordenfelt rotating breechblock for metallic cartridge and its hydraulic recoil system, were both largely Germanic in origin. The breechblock was essentially an enlarged version of that on the Austrian Werndl-Holub single-shot rifle of 1867;


        And the recoil system was substantially based on braking systems for locomotives, and balancing systems for cranes, developed by the German Siemens company.

        It’s also rather ironic that in 1865, the U.S. Army’s artillery was declared the best and most destructive force on any battlefield on Earth. Yet by 1870, everything they had was obsolete.

        The American Civil War may have been the first war fought with breechloading rifles, at least in part. But it was also the last war fought with muzzle-loading rifles and smoothbore muskets, and the next-to-last fought with smoothbore muzzle-loading artillery on either side.

        The French went to war in 1870 with the same sort of artillery the Union and Confederacy had used on each other. Natural enough; the M1857 12-pounder “Napoleon” gun-howitzer was based on a French design, as shown by its name.

        They found out the hard way that military technology had moved on in the decade and a half since they’d first fielded it.



  4. Drum magazines are great when loaded albeit cumbersome , a big draw back is trying to reload a drum in combat situation ! Easily accomplished at the range or at home .

  5. The belt-carrying box magazines for the Stoner 63 system were perhaps the most reasonable approach to the: the complexity is in the gun (the belt feed mechanism) and box is simple. The 100 round version in the Stoner 63A were 100 rounds (instead of 150 or 250), I think, because 100 was the practical weight limit for a magazine hanging off the gun.


      • Wait… are pan magazine (like 100-rounder for BREN and default Lewis machine gun magazine) subset of drum magazines?

        • Technically, yes and no.

          “Yes” in that they tend to have the same basic spring-wound system to rotate the cartridges around to be presented to the bolt/feed system.

          “No” in that the presentation is fundamentally different. The pan lies above (or in a couple of rare cases, below) the barrel and bolt, and usually relies at least partly on gravity to deliver the round to same.

          Generally, pan magazines are more common with gas-operated weapons like the Lewis Gun or DP, as opposed to recoil-operated weapons like the MG34 and etc. No, I don’t know why.



          • Pan magazine seems natural choice if you need both to use rimmed cartridge and high capacity.

            In this case, pan magazine was proliferated due to inability to get belt-fed system to work reliably with 7,62×54 R cartridge, with many attempts in 1930s which ended in failure and only – what we known as RP-46 today – starting as early as 1939, but reaching maturity (ability to being adopted for service) in 1946.

  6. When the M1 Carbine was under development, the specifications were originally for 20 round and 50 round magazines. At that time (1940) there wasn’t a lot of combat experience with the Thompson–Marines used them in the Banana Wars but not many. By October 1941 the select-fire capability had been dropped and the Carbine magazine capacities were postulated at 10 and 20 rounds–obviously enough had been learned to ditch the 50 shot magazines (almost certainly a drum magazine of some sort).

    • Check out the Forgotten Weapons pages for the Light Rifle submissions that resulted in the M1 carbine. Especially John Garand’s carbine. Fifty round box magazine. No drum magazines that I know of were submitted. Those entries were what got my attention and I began following Ian and his work then.

    • “(…)50 shot magazines (almost certainly a drum magazine of some sort).(…)”
      Note that .30 Carbine cartridge is much lighter and has much smaller diameter than .45 Auto cartridge used in Thompson sub-machine gun, so if anyone would want drum-50 for .30 Carbine it might be of smaller diameter and lesser weight when fully loaded.

  7. Drum magazines may provide a lifesaving feature in mass shootings–their complexity requires a skilled operator.

    During the Aurora theater massacre, the drum-fed AR-15 quit working.

    The AK rifles used by the bank robbers in the North Hollywood Shootout malfunctioned excessively due to overheating–apparently the drum magazines worked for them until several hundred rounds had been fired through closed-bolt weapons with “thin” barrels. Exceeding the fire rate jammed those “jam-proof” AK rifles through overheating–the AK wasn’t designed as a light machine gun.

    I think that banning 100 round drum magazines will cost lives because when not properly loaded and adjusted those really big and expensive magazines fire four or five shots and then quit, often tying up the rifle with double-feeds and other malfunctions requiring tools and time and skill to correct.

    • Forget about it. Most people will look at gangster movies and then tell you how scary the drum magazines are in any gunfight. Political advocates/lunatics aren’t expected to study rocket science, after all!

  8. Having learned that from a practical standpoint that drum magazines are not ideal, they are just too cool to not have one for display. I defer to the concept that standard magazines were developed for a functional-reason. Thanks for your work. Appreciated, as always.

  9. There were a couple of other SMGs that used the Suomi drum. Both the Swedish Karl Gustaf M45B and the Yugoslav Model 1949 9mm SMGs had a magazine housing for their 35-round box magazines that could be removed, allowing the Suomi drum to be locked in place as on the PPSh-41. The Egyptian “Port Said” copy of the M45B also had this setup.

    Having dealt with both the 50-round L drum and 100-round C drum of the Thompson M1921 and M1928 in person, I would point out three other problems with drums.

    1. Weight; The Thompson 50-round drum loaded weighs 3 pounds, the 100 round weights 8 pounds. Added to the 12-pound weight of the Thompson empty, the result actually weighs more than a loaded BAR with bipod. The C drum turns the M1921/28 into a 20-pound boat anchor.

    2. Handling; The Thompson needs that vertical foregrip with either drum, as it’s almost impossible to “reach around” the drum to get a proper grip on the squared horizontal (military) forearm. Controllability is not enhanced by a drum on a milspec (early 1942 model) M1928A1.

    Note that the PPSh-41 has no forearm, not even the abbreviated one of the PPD-40. The proper address for the Shpagin is to hold it with your off hand around the wood behind the drum, which incidentally is how you’re supposed to hold an MP38 or MP40, hence the grooves on the lower receiver cover. Which nobody ever does on the Erma, naturally; they wrap their hands around the top of the magazine housing and generally end up with scorched fingertips.

    With a drum in place, there is pretty much just no way to get a proper grip on a horizontal forearm on any SMG or rifle with the drum hanging underneath.

    3. Noise; Drums rattle. The Thompson drum allows rounds to shift back and forth as you move, making a distinctive sound rather like marbles in a kid’s pouch.

    During the Phony War, BEF troops along the front line found that carrying a Thompson plus 50-round drum on a night patrol almost guaranteed that sloshing/rattling sound would give away their position. The usual response was a burst of MG34 fire from an alert German MG crew. Not optimal.

    Drum magazines began with Charles Tyler’s 1853 patent, but never really became common until the introduction of things like the Trommelmagazin 08 for the Parabellum pistol before WW1. By the end of same, you’d think everybody would have figured out how impractical they were, but John T. Thompson had other ideas. Most of the later drum magazines were actually copied from his designs. Yes, even the Suomi drum, that begat the PPD-40 drum, that begat the PPSh-41 drum. BTW, the PPD-40 and PPSh-41 drums are emphatically not interchangeable; different interfaces with the guns.

    The drum magazine is an attractive notion that can work. If it’s something like the Beta C “drum” which is really fundamentally a WW2 MG34 “saddle drum” turned upside down with an M16 box magazine in the middle to hook up to the rifle.

    But the “traditional” drum magazine is generally more trouble than it’s worth.



    • “(…)If it’s something like the Beta C “drum” which is really fundamentally a WW2 MG34 “saddle drum” turned upside down with an M16 box magazine in the middle to hook up to the rifle.(…)”
      Twin-drum for originally box-magazine-fed full-auto weapon? I know. This is like Doppeltrommel for MG 13, see 3rd image from top:
      is there any known feedback from their actual combat usage?

      “But the “traditional” drum magazine is generally more trouble than it’s worth.”
      Maybe, but they are recurring in Soviet/Russian design of intermediate cartridge firing machine guns, recently developed RPK-16 https://modernfirearms.net/en/machineguns/russia-machineguns/rpk-16-eng/ accepts drum-95 (though also box-30 and box-45)

      • Right, the C-magazine may be an exception. And as you properly mentioned it originates from MG34 saddle mag.

      • I was going to use the MG34 saddle magazine as a counter example, especially since it is still belt fed. If anything the saddle helps balance the weight distribution.

        I’ve fired a Suomi with a drum magazine. IMHO:
        #1 a PITA to load
        #2 A PITA to attach to the magwell
        #3 It really shifts the center of gravity and makes a heavy SMG even heavier.

        I can’t imagine reloading a drum while someone is shooting back at me. I also feel that drum magazines do not do well in mud and the general banging around that goes on in the daily life of the infantry. In the defense, I expect it is faster starting a new belt and in the offense/maneuver a stick magazine provides better agility.

        • The MG34 “saddle” double-drum magazine (patrronentrommel) holds loose rounds, not a belt. The later single “drum” side-mounted round thing (gurttrommel) isn’t a drum. It’s a round belt box holding a 50-round belt.

          The saddle mag was originally intended to be used in the LMG role, and loose belt in the MMG one. An issue was that the double-drum saddle mag and belt feed required a change of feed block for each to work.

          Not clear why that was judged a big problem (reliability, cost, logistics..?), but the saddle mag and associated feed block got dropped, with the round belt box being standardised for the LMG role, as continued with the MG42.

  10. I’ve played with a PPSh while embarking on a government funded vacay in Iraq and can’t imagine trying to reload a drum magazine in limited vis or any other circumstances other than broad daylight and complete and total calm.

  11. Good to see Ian ‘lashing’ on his own instead of describing someone else’s idea. I concur with his point.

    One additional concern he did not mention – with drum magazine of large capacity it happens that full magazine can be just as heavy as empty weapon. With this also changes dynamic behaviour, which will have effect on accuracy, not to mention initial mobility.

    • One thing I came across is the detail of drum magazine kinematics which is not often thought of. It is the fact that half of ammunition weight is assisting drive spring due to gravity; minus friction that is. In regular box magazine the drive spring has to muscle up entire column on its own. That leads to too much pressure at start and possibly not enough at the end.

      Also, another interesting detail: magazine springs are tested on set/ fatigue. Full magazine is let to ‘relax’ for some time (days to weeks) and then run if it will feed as expected. In no way magazines are easy thing to do well.

  12. In terms of production, the drums are complex and expensive to manufacture.
    From the point of view of the user, they are heavy, large and sticking out.
    My grandfather preferred to use PPSh (1943-47) with a box mag. He said literally “the drum is heavy and uncomfortable.”

  13. “…eload a drum magazine in limited vis or any other circumstances other than broad daylight and complete and total calm…”(C)

    And my grandfather also mentioned this. You can also add a “piece of clean cloth.” Since filling the disk with cartridges while squatting in the trench, when any garbage is pouring from above, and not collecting this garbage inside the drum, is practically unrealistic.
    At the same time, drop another half of the rounds…

  14. I agree that there are better options than drum magazines. For individual weapons, two box magazines and a coupler offer similar capacity with lower cost and greater reliability. A concrete example is the Magpul D60 for AR15s, 60 rounds for $120, compared to two 30 round PMAGs and a coupler for $60, plus you can always use the PMAGs individually to save weight and bulk.

  15. Just as a note, the US did have a 300 round Thompson drum which was used in some mounted applications. They’re stupid rare, and I wouldn’t want to lug 300 rounds of forty-five off of a mount on a jeep.

  16. I use the Magpul D-50 drums for my P308 SPR, my son uses Magpul D-60 drums for his Steyr AUG NATO 20″ barrel, both the D-50 AND D-60 have performed flawlessly in both of these platforms. In a SHTF situation, both of us would be comfortable with a drum inserted in the rifle for the beginning of a possible skirmish, backed up with stick mags for when the drum is depleted of ammo

  17. The only place where a drum magazine would make sense apart from GPMG or LMG usage is in a fixed-position mount. Examples of such applications include wing-mounted or engine-mounted auto-cannons (for fighter planes or attack planes), bomber turret guns, fortress positions, and, of course, flak mounts. Did I mess up?

    • The major user of drum magazines in fixed wing installations was the Hispano 20mm cannon. Which of course was simply the Oerlikon redesigned slightly as an aircraft gun.

      Antoine Gazda of Oerlikon-Buhle in Switzerland scammed pretty much everybody in the years leading up to WW2, by selling them the “exclusive” rights to the Hispano cannon and not mentioning that he was doing the same with their future enemies. He sold the same design to the British, the French, the Germans, the Poles, and the Japanese as aircraft guns, and then turned around and sold it to the Royal Navy and U.S. Navy as a light AA gun.

      All versions used a drum feed because the original version, intended as an anti-tank gun (!) on a two-wheel carriage, used either a 10-round top-mounted feed hopper like the later 40mm Bofors, or a 20-round drum.

      Most users stuck with the drum feed even as an aircraft wing gun. Only the RAF and U.S. Navy modified the design to use a belt feed.

      If you look under the wing of a Bf-109, just outboard of the wheel well, and see that little bulged door right next to the wing gun, that’s where the drum for the 20mm Hispano-type cannon is.

      It wasn’t until everybody on both sides started examining the enemy fighters they’d shot down, and said “Hey! This 20mm cannon is just like ours!” that they all realized how thoroughly Gazda had snookered them.



      • “(…)drum feed because the original version, intended as an anti-tank gun (!) on a two-wheel carriage, used either a 10-round top-mounted feed hopper like the later 40mm Bofors, or a 20-round drum.(…)”
        Do you have any photo or drawing of “anti-tank gun” version of Hispano-Suiza 20 mm caliber auto-cannon at wheel carriage? So far I know drum was chosen, because Hispano-Suiza 20 mm auto-cannon was supposed to be engine-mounted (V-engines), see 3rd image from top: http://airwar.ru/weapon/guns/hs7.html

        • See Artillery by Ian Hogg, pp. 84-85, for a large illustration of the Becker-Semag cannon. The caption reads,

          The forerunner of the Oerlikon gun, this was intended as a light 20mm infantry support weapon, but many nations tested it out as a potential anti-tank gun in the 1920s.



    • “(…)Did I mess up?(…)”
      Generally speaking drum-magazines are better suited for weapons which are supposed to be fired from support, rather than free hand – which might be bi-pod, tripod, vehicular mount and so on.

      Some drum-magazine-fed auto-cannon were used in early part of Second World War, but with development of aeroplanes to being faster and faster, evolution direction was towards belt-fed as magazine hanging below caused additional drag (drag is proportional to square of velocity with all other being equal, for more data see:
      https://physics.info/drag/ )

    • I read somewhere (“THE GUN THAT MADE THE TWENTIES ROAR”?) that the Thompson Type C magazine (100 round drum) was a proposal to mount a bank of Thompsons in an airplane for low-altitude strafing runs.


      Perhaps the 100 round drum wasn’t designed for portable firepower but instead was intended as a vehicle mounted weapon. That would explain why only a few hundred 100 round drums were made.

      Firing pistol cartridges from a low-flying plane flew in the face of the “independent air force” boys quest for freedom from “holding the infantry’s coat” and was dangerous to boot. Submachine guns have much less range than do rifles (or rifle-caliber machine guns) and lower penetration. If a plane got close enough to hit with a pistol, obviously the plane was in rifle range.

      Interesting what engineers can come up with.

  18. This is a very negative review of drum magazines.

    Would it be fair to say that Ian…

    [puts on sunglasses]

    …PANNED them?


  19. We tend to forget that a military that uses firearms spends most of its time carrying them, not shooting them. And in that use drum and pan magazines are generally horrible.

  20. They are not completely useless. In a number of situations, patrols, ambushes, and the like, they can very well replace belt-powered weapons.
    In any case, there is a steady demand for 50-round cartridge bags for machine guns.
    But they can not completely replace belt feed. Nothing can be done about their cost and dead weight.

  21. I’m going to throw something out here that may stick in some craws.

    The primary reason that drum/pan magazines don’t work very well is that they’re poorly designed. Most of them are self-powered, using springs, and mostly have really poor interfaces with their host weapons. This being the case, I think we may well be judging their performance based on these factors, rather than any class-inherent characteristics.

    Stop and analyze: The drum/pan form-factor, if we may term it that, is not well-suited to being internally powered by springs. Box magazines are very well-suited to springs, and do well because of that. The thing is, however, that we have to look at a lot of things. Box magazines suffer limitations due to the nature of the beast–Double-column means that you’re automatically going to be limited to certain lengths to reach specific capacities, or they’re just not going to be very effective due to their unwieldy length. If you need a certain amount of room under the weapon, you may find yourself interested in trading that long stick magazine for something else, like a nice, compact drum. It’s all a tradeoff.

    What I haven’t seen is anyone looking at this from a standpoint of either shifting the cartridge shape paradigm, or perhaps having the magazine powered by the weapon. Either option may offer us some realm for improvement in some applications.

    I don’t think it’s necessarily the form factor for the magazine, in other words: Drums, pans, and even helical magazines might be very workable in certain circumstances we haven’t identified or imagined as of yet.

    Consider a cartridge like the Dardick Tround, for example. Put that into combination with a helical magazine a la the Calico or the one under the North Korean ones, and you might find that a very effective combination.

    The reason I think we have such poor performance from the drum/pan style of magazine has rather more to do with the current paradigm surrounding our weapon ecosystem than anything else. Change a few features of that ecosystem, and we might find that the form factor works better than it does.

    • If you search for tround on youtube, you can actually find footage of a.50 tround machine gun that uses a helical mag.

      In fact if I remember Ian’s video on the Dardick, it was only supposed to really be a demonstration of the tround concept.

  22. There are no universal solutions. Each in its own niche.
    And why limit yourself to two columns? For a long time there are quite reliable and cheap (compared to drums) mags with 4 columns. Although they do not reach the capacity of the drums.
    Rather, the point is apparently that for the army mags of this capacity are not very needed. And those who believe that this is necessary, can buy in the store right at the base.
    Like in Vietnam. Basically, there were enough M16 20-rd mags for everyone. Those who lacked ordered a 30-rd wholesale in the mail, when they appeared.

  23. Thanks for mentioning 4 column mags.

    The very little that I’ve read about the Carl Gustav 4 column “coffin” mag, suggest that it probably benefitted from the vibration from recoil in order to get those four column to coalesce.

    I’ve a copy of the patent somewhere on an old hard drive

    It had an interesting arrangement of followers, to make sure that one follower stopped and made a smooth path for the other follower to continue up to the feed lips.

    I can’t remember whether the follower which stopped short had more spring pressure, to make sure that it got there first, or whether there was another interlock to stop those two columns getting cut off prematurely by the follower that continued to the feed lips.

    While the coffin mag does have a less inconvenient shape than the Suomi drum, it is still heavy and complicated.

    I don’t think that there is a good answer for the question of “what orientation to put a magazine in”

    Small capacity tube magazines for rimfire cartridges, do work well in the stock, eg, the little, browning .22RF semi autos. But as soon as a larger capacity is needed, along with mag changes, things get both ugly and uncomfortable to carry.

    • The Italian Spectre M4 SMG has a “four column” magazine. But it’s actually more like two double-column magazines side-by-side, with a specially-shaped set of feed lips to allow it to feed at the top like a conventional double-column rifle magazine, aka M16/STANAG type.

      Its main virtue is that it’s shorter top-to-bottom than most over-30-round-capacity SMG magazines. Simplicity of loading and maintenance are not among its attributes.



  24. The mags Suomi was forced to exit in one row, because he was interchangeable with the drum. With PPSh the same.
    There is already a mags for both M16 and AK with two-row access, and they are quite functional.
    Could be.
    If those who created them are mneeee …
    they didn’t get carried away by excessive simplification and did not throw away “extra” structural elements.

  25. You could use NiTi actuated drums, POW! POW!! POW!!! Nitinol contracts; rotates drum, heat generated by gun.

    Might work.

  26. Ditto, 2x 30rnd mags taped up ‘jungle style’ brilliant until you fire 60 shots and the prick doesnt fit back into a pouch ( dump pouches kinda work)

  27. Completely opposite; Yugoslavia never made 40 rnd mags, but did actually produce drums for its AK variant.

Leave a Reply

Your email address will not be published.