San Cristobal Model 2: A Lever-Delayed .30 Carbine

The San Cristobal armory was a surprisingly large-scale arms manufacturer set up in the Dominican Republic in the late 40s. They produced several different weapons, but the most common was the Model 2 Cristobal carbine. Designed by Pál Király, it was a lever-delayed blowback mechanism chambered for the US .30 carbine cartridge (7.62x33mm).

Tens of thousands were produced (exact numbers are unclear), and they were a popular weapon in central and south America. Unlike the US M1 Carbine, the Cristobal used a magazine mounted at slightly less than 90 degrees to the barrel. Most of the carbines produced were select-fire, with a front trigger for semiauto fire and a rear trigger for fullauto. A semiauto only variant with a single trigger was apparently also produced. All types are quite uncommon today.

Many thanks to the Royal Armouries for allowing me to film this tremendously rare artifact! The NFC collection there – perhaps the best military small arms collection in Western Europe – is available by appointment to researchers, and anyone can browse the various Armouries collections online.


  1. It always surprised me that the Dominican Republic had such a large small arms operation. Brazil, Argentina…sure. The DR?

    And also that these carbines are so rare.

    • Think of Europe, where small countries such as Belgium and Czechia are major players in the arms business. In the end it is always very specific regions harbouring an industry branch, and that may be within the borders of a small country.

    • The reason for this venture was part of F. Trujillo’s ambition also the fact that some remarkable designers/ technicians from Axis armament industries found a convenient hideout there. Remember what some Germans had to go thru if they wanted to stay in Europe (detention in France, later temporary job Spain).

      There was not just an arms factory in Dominica, but also foundry/ steel mill producing required materials; all run by foreigners.

    • “(…)Dominican Republic(…)” explain it as follows:
      Trujillo Molina observed actions of Papa Doc Duvalier in adjacent country and feared he might launch invasion on his country. So he decided army with manpower above 50000 is required as deterrent, thus he needs armament. He has good relations with then ruling dictator of Argentine, then ruling dictator of Nicaragua and then ruling dictator of Spain (see 2nd photo from top), but they were unable to provide required number of weapons. Armeria San Cristobal was already producing some copies of Beretta 1938 and there was Pal Kiraly which left Hungary due to unfriendly political climate. Rest of story is known.

  2. The stock is very much Beretta-like, indeed. The description Ian made, about this being a sort of a Beretta M38 derivative, scaled-up to .30 Carbine and given a lever-delayed system is really spot-on. The Trujillo regime managed to export a few of these guns to neighbouring countries, Cuba being one example.

  3. Forgot one thing: how many of these carbines did the Colombians order? One could expect some to pop-up from that source… I agree that the Cuban ones will not show up in the foreseable future but I wonder what happened to the lots exported to other nations.

    • As Mel said, that camming system is closer to roller-delay than any other delaying system in production, with the vertical slide acting as a roller, and the charging handle as the locking piece.
      But with a single “roller”, I’d say it resembles the Budischowsky flavor (also recently seen on the Stribog A3) more than the Vorgrimler/HK one.

  4. It is a bit of surprise to me that this carbine did not find greater user base in the world (well yes, the U.S. M1/2 carbine snubbed it a bit). It is an ideal combination of SMG and rifle; plenty sufficient up to at least 200m range.

    I am observing some changes in the mechanism in comparison to previous iteration. It looks more streamlined, however in my opinion it would be better if the delay lever could be located in middle of receiver.

    Samples may be seen in Cuban museum of revolution. When Ian eventually goes there he will find some real treasures. Amigos Yanquis are welcome 🙂

    • If the San Cristobal had been made in larger numbers and not been exclusively military-purposed, could it have competed with contemporary self-loading rifle designs on the private sector? Or would people laugh it away as a “poor man’s excuse of a gun” to be used as scrap material?

      • You might be correct. If it was presented as say hunting weapon, it could have sailed well. Of course, we see it with todays eyes, not of those in 1950s. The civilian markets grew exponentially.

        One reason I give this discussion so much attention is because I believe something like this, in modernised interpretation may have sales potential today.

        • Rather than ‘hunting’ the ‘security’ destination may fit even better. Assault rifles (although without select fire capability) in hands of police are not exactly what public likes to see.

    • “(…)ideal combination of SMG and rifle; plenty sufficient up to at least 200m range.(…)”
      According to manual in English available here (after photos of said weapon in usage):
      The present model Model 2 of the Automatic Carbine Cristobal is
      adapted for the Caliber .30 Winchester M.1. Carbine Cartridge.

      The useful range of this carbine is between 300 and 400
      yards. Of course, this carbine can be manufacture for other
      kinds of ammunition, the useful range of which might be
      over 500 yards.

      It can also be provided with a bipod to use it as a light
      machinegun or with a bayonet holder and a cleaning rod in
      order to use it in the same way as an infantry rifle for
      which it is particularly suitable due to special trigger
      for single fire.

      The trigger mechanism is of the double-trigger type. The
      first trigger serves for single fire. When pulling this
      trigger, an interrupter frees the sear from the trigger
      in the usual way, so that at each trigger pull one shot
      only can be fired. The second trigger keeps the sear in
      its low position as long as the trigger is kept in its
      rear position; automatic fire ensues. The rate of the
      automatic fire is somewhat less than 600 shots per minute,
      so that the automatic fire can well be kept under control
      by the operator.

      • What gain it would provide?
        1. keep load in centre of symmetry (always desirable).
        2. simplify machining and increase strength of mentioned lever.

        However, I realise this is hard to accomplish due to presence of magazine and that is why it is off to one side. IF the bolt/lever arrangement is reversed and adequate support surface is provided in top part of receiver it may be a realistic proposition.

        This is where I see a possibility for further development and chance for revival of this project.

  5. How do firearms designers determine the delay time necessary
    in conjunction with the mechanics in order for the gun to operate
    properly? I hope I’ve phrased this question correctly and that it makes sense. I find the seemingly minuscule adjustments necessary to be fascinating. Thanks in advance.

    • This is not a type of mechanism generally implemented these day, however factor of a “delay” or “dwell” of sort is always a major consideration to start with.

      Likely you start with pressure/time and pressure/travel curve for given cartridge; this gives you a picture where your constrains are. Later, during prototype(s) testing it becomes obvious if initial judgement was sound.

      Testing takes typically at least half of design effort. Revisions need to be tracked meticulously. It is useful aid to future optimisation. When design performs to satisfaction, tolerances requirement vs mfg. cost analysis is carried out. A final word belongs to process/ manufacturing engineers.

      Did I forget something? Yes, of course – marketing and sales. They will be asking the unpleasant questions all the time 🙂

    • They try until they find the right combination of (in this case) lenght of the lever, and weight of the two parts of the bolt. Or only weight of the bolt, or lenght of the short recoil of the barrel, etc.
      And obviously, once the designer thinks to have found the right combination, the tests has to be repeated in any kind of environment (weapon wet, dirty, greasy, unlubed, frozen…).
      There are, for example, formulas to determine the right weight of the bolt for a certain cartridge in a blowback action, but, if you check the weight of the bolts of real blowback weapons, they almost never correspond to the theoretical ones.

  6. Eventually it is a possibility to visit the “Danish Armament Collection”, which for much more than 100 years have bought and kept possible weapons for the Navy and Army, so that their collection now is better than even the factories own! And they get free samples of new weapons from several factories, because of this tradition.

      • Thanks Daweo… This means, the gun having the weight approaching nearly to the weight of a rifle using more powerfull cartridge than M1 Carbine’s… Then, using that delay action in a “Open Bolt” SMG seems completely useless… Pity… However, at firing, the stepped closing of two piece bolt through delay lever should lessen the effect of impact of closing bolt thus, not ruining the accuracy as much as other open bolt firing samples…

        • Good mention Strongarm.
          With open breech firing mechanism, the recoil energy is better absorbed (momentum forward subtracts from momentum back) in addition of cook-off prevention during prolonged firing. If I was to duplicate this mechanism, I would stick to original concept.

      • Again, good point. The gas operating gun has always weight advantage over alternative mechanism (such as in case we are looking at here) equivalent weapon – see older generation recoil operated machine guns for example. There is one negative part to it though – a lighter weapon produces greater felt recoil.

        In case of Cristobal carbine the designer has directed his effort to minimise the mention recoil (shock) by implementing forward momentum of relatively (when compared with gas operated gun) breech mechanism. I think he has done rather well in that regard.

        • “…implementing forward momentum of relatively (when compared with gas operated gun) breech mechanism.”
          should be:

          ..implementing forward momentum of relatively heavier (when compared with gas operated gun) breech mechanism.

    • This man (Pal Kiralyi) must have had trust to his mechanism. He was obviously right when applied to carbine (for less powerful than rifle round). Not so much in case of rifle (he did try .308 caliber rifle version) and rather doubtful when applied to machinegun (AAT52).

      • Thanks Denny… It seems Kiraly was one of the keen minded firearm designers of auto feeding systems at starting age.

  7. A few thoughts,

    If you are running the rough figures for a first prototype of a delayed blowback

    If you know what barrel length you are going to use, and the weight and muzzle velocity of the intended bullet out of that barrel,

    You only need to use Newton’s laws of motion

    (Dont ask me what number, I actually couldn’t answer a whole question in a British public exam because I didn’t know what number each law was – and four decades on, I still cant remember)

    For pure blowback

    Mass of bullet x muzzle velocity of bullet = mass of bolt x velocity of bolt at the point that the bullet exits the muzzle. Unless you are using a very heavy (in terms of actual weight) spring, you can ignore everything about the springs.

    Putting time into that, we get distances traveled

    A bolt that’s 20 times heavier than the bullet will travel 1/20th of the distance that the bullet does, in the same time

    What you want is something like less than 3mm travel of the case head, before the bullet exits the barrel.

    When you put “delays” in there, what you are doing, is accelerating part of the bolt mass.

    You calculate mass the front part of the bolt as though it was pure blowback, and multiply the weight of the back part by the ratio of your accelerator, add them together and treat them as simple blowback.

    Open bolt operation allows you the fire, to take about half of the recoil impulse before the cartridge actually fires.

    In terms of control ability after that, it helps if the bolt assembly doesn’t have to hit the rear end of the receiver tube. Springs and buffers are very useful here, as are long receiver tubes.

    De kiraly’s parts are probably needlessly complicated to machine, I’m guessing that you could probably cut a long prismatic piece, in the profile of the accelerator, and slice accelerators off it.

    And use a simple hole through the accelerator to put a pivot pin through, rather than machine the whole thing as one piece.

    Felt recoil is going to be approximately proportional to rate of fire

    A long tube for the bolt assembly to recoil down and a soft return spring (that still manages to stop the bolt before it hits the end cap with a bang) will minimize shocks

    But it won’t return the bolt to the firing position as quickly as a stiff spring and / or bouncing the bolt off the end cap of a short receiver.

    This is actually something that gas operation has an advantage in

    You can use the gas system to accelerate the bolt carrier, but as that internal assembly is accelerated backwards, it acts to accelerate the frame and stock of the gun forwards

    You don’t get a free lunch though. You eventually have to stop the rearward travel of the bolt carrier and bolt, then return them,

    But you can use it to stretch out the period of time that recoil is occurring (by slowing the cyclic rate of fire).and reduce felt recoil

    The only actual way to reduce recoil (rather than spreading it over a longer period of time) is to convert some of the muzzle blasts into thrust, through a good muzzle brake, but that will also stir up dust and win you both friends and influence, by deafening and concussing everyone around you.

    .30 carbine is interesting.

    It is actually surprisingly close in muzzle velocity and energy, to 7.62×39

    Unfortunately it was thought of as a pistol replacement, so it was loaded with pistol style round nosed bullets, rather than pointy spitzers.

    There’s far better comfort, trajectory (and possibly feed reliability) to be had with a .22 Johnson spitfire round, or even the. 256 Winchester, that Universal chambered its “ferret” carbine for.

    If American army procurement people had been thinking rifle instead of pistol (they weren’t). A 125 grain spitzer bullet with a light weight or steel filler in the nose, along with rifle operating pressures, would have made the carbine even closer in performance to the later SKS and AK.

      • Newtonian mechanics, “enchanting nonsense”?

        Please enlighten us about any manual that corrects/supercedes Newton

        Along with the hows and whys that might be so.

        I await with interest…

  8. And to juggle, too, please stop.
    The “enchanting nonsense” is not Newtonian mechanics.
    And only what was said in the post to which my comment applies.

    In short.
    When fired, the gases press the brass against the chamber walls and the bolt’s mirror.
    When the pressure exceeds the limit (depending on the case material and chamber design), the brass “sticks” to the chamber. The case is deprived of the opportunity to shift back.
    In this case, the pressure reaches a maximum.
    If the pressure curve is relatively steep, the bottom of the case continues to move relative to its walls, and since there is a mirror gap, it has room to move and it “tries” to break away. To this gap is added the displacement of the bolt mirror due to the elastic deformation of the locking unit itself.
    If the total amount of this offset is greater than the brass can stretch without breaking…
    a break occurs.

    Therefore, as the value of the maximum permissible displacement of brass when fired, take the value of the plastic elongation of the case before destruction.
    Which usually corresponds to the mirror gap.
    And Your attempts, among other things, to focus on the thickness of the bottom of the brass as the amount of allowable displacement, date back to the last millennium, when black powder did not usually cause such a rapid and significant increase in pressure.
    For a charge made of black powder or another with a sufficiently smooth pressure curve, the magnitude of the mirror gap can be neglected. And we focus on the magnitude of the elastic deformation of the bolt, which is usually clearly less than the thickness of the body bottom.
    In this case, the assumptions You proposed can be taken for rough calculations.
    But this is not our case.

    If still not, try to beat this.

    Happy and Merry New Year!

Leave a Reply

Your email address will not be published.