The Gemini ‘egg’

I am here at the safe opening weekend and can tell you some monster safes are being opened as I type. As always the weekend is a great success, with safe techs from all over the globe attending.

Gemini 'egg', set up key and opening tool

One person who is a safetech and brought some tools to sell (like the electronic stethoscopes people here are found of) at the little marketplace here is Gerald from Berlin. And Gerald also is the inventor of the Gemini set up key. With this tool we finally managed to open the Gemini shield in the Dutch lockpick competition, and I now had the opportunity to ask him some more details about it.

One thing he did show me is a fifteen year old prototype (more images here and here). And since I know blackbag readers are curious I shot some images of the latest generation tool and the internals for you. Here is an overview of the tool, and here you see the pins used to keep the inner and outer pins in place. The inner pin can be changed in height by turning the knobs at the back of the tool.

The depth of the outer pins are changed using a screwdriver and the inner pin goes straight trough it (and is also threaded). I hope the images speak for themselves.

Geralds stories are interesting. And a little mysterious too. The tool was available on the free market for one year (for around 500 euro), and later just more or less disappeared from the popular lock tools catalog. Maybe not enough demand?!?

Anyway, got to go now … Rob just fired up his (by now famous) BBQ ….

32 Responses to “The Gemini ‘egg’”

  1. Arcturus says:

    Neat! So would you say the working principle is similar to one of those tubular picks that allow locking of individual pins? Or is the working principle closer to that of impressioning?

    If I understand these photos correctly, it appears the egg must be removed from the lock before the outer pin’s tip heights can be adjusted with a screwdriver. Correct? If so, is setting the outer pins simply a matter of trial and error rather than feel? And because the inner pins are held/controlled by threaded rods and not free-moving friction rods like traditional tubular picks, does that make it more difficult to sense the shear line? Also, is there any recommended picking order? All outer pins first, then inner? One stack at a time? Bound pins first?

    In any case, from these photos, it looks like you could probably use this tool to impression the lock by pasting foil to the flat ends of the inner pins. That might speed things up.

    BTW, has anyone tried bumping these shields before? Wish I had a Geminy to experiment with, but I’m thinking if the pin stacks aren’t too long for it, and the springs aren’t too powerful, perhaps you could create a bump key with a modified egg tool or a set of tryout bumpkeys using a hunk of metal with screws or bolts installed at the proper heights?

    Just thinking out loud.

  2. What size are those threads? 1 mm pitch? And couldn’t you quite easily use the same kind of knobs for the outer pins also? Would ease decoding probably.

  3. Martin says:

    I have built a tool like this some months ago (looks much worse but might be more useful). To be honest I don’t think it is anything special.
    Isn’t this the first thing that comes to mind when you see a geminy?
    I was always thinking the “geminy is unpickable” bullshit is due to the fact that Mr. Drumm is a nice bloke and has nice friends. 😛

  4. Barry says:

    Martin: The first time I ever saw a Gemini (on a railroad ticket machine) I knew for sure it could be picked easily. But I learned it is an (almost) impossible lock to manually pick. And for a lock that has all pins on the outer side that is quite an achievement.

    I have seen Han put a reward of $1000 US (If I remember correct) for anyone to pick the lock at the Aloa locksmith trade show, and nobody succeeded.

    And in six months time none of the members managed to pick it manually, and I know some people really tried.

    But with a tool is seems possible ….

  5. mh says:

    Martin: Been there, done that, same thoughts 🙂

    And yes, I also think it’s nearly impossible to pick manually.
    The versions I have seen are however very easy to decode, once you have a decent variable makeup key.


  6. Tom says:

    It indeed looks simple to pick. Until you actually try. Then you find out that you can feel some correct positions and set some pins, but when you start playing with other pins it is unavoidable that pins that were already set pop back. You will need to keep them into their set position, and for doing so without a dedicated tool, you simply don’t have enough hands… (and every pin (10 or on some Geminis even 14 of them) needs to be moved down at least 1mm minimum, as can be read in Han’s excellent article)

  7. Henk says:

    To pick this lock you have the set the outerpins first,there is no feel at all when you try innerpins first,there is an order in setting the pins,you can feel when a pin has reached the shearline,once that pin has been set you can move on to the next,once you,ve set all the outerpins,you can try the innerpins one after another,you will feel which innerpin has to be pushed in first,after that it opens within seconds.
    When using the egg the feel is,imho,different.
    Arcturus,it is not by trial and error you can set the pins,you can feel what pin to set first.
    Next week there will be a TOOOL gathering in Eindhoven,I hope to share my results overthere,not using the egg…..:-)

  8. Arcturus says:

    Thank you Henk. That’s helpful information. Do you think the outer pins are susceptible to expansion heating with a soldering iron? Perhaps that might facilitate picking. Also, what do you think about the bumping prospect? Feasible?

  9. Henk says:

    I don,t think heating with a soldering iron will do any good,besides that you don,t need heat to make things easier.
    I,ve been thinking about bumping as well,haven,t thought it trough good enough though,but possible?Testing will tell,talking about a problem often helps to find a solution,but practice does make perfect;-)

  10. Lockpicker says:

    It’ a good lock, but far from “evil”.

    The way I would attack this lock is by impressioning. It have no protection against impressining as the “outer” pins are non modyfied. Similar picking protection is engaged in the Swedish most used lock, ASSA 700, but better due to the groves in the core.

    With a tool like tubular pick it would be self impressed in a pair of minutes. Two tools and two operations to make the tool less expensive.

  11. Till says:

    Hello, i am very interested how you will make it possible to cut the inner and outer cuts seperatly and how you read the marks lockpicker.
    Do you have blanks for it? Did you already ipress a key?

    I have tryed for hours and days to pck this lock and never could set more then 2 Pin pairs ( Inner and Outer ) at a Time.

    I dont think a Bumpkey will work because you cant tension it very controled.

    Of course there can is always a better way to build for every thing but This is working, this is exsisting since some years, and it is improved that it opens the lock in good time.
    Witch is the most important for me.
    nice greetings

  12. mh says:

    This refers to the only setups I’ve seen: 2 “normal” outer pins and 3 “spool” outer pins.
    The “normal” pins obviously bind first. Pick them, then the shield will move a little bit and the spools catch. Push the spool pins in as far as possible. [If you compare their depth to the key pins, you should find a simple correlation…] -> set the variable setup key accordingly. Pick the inner pins (easy). Open.


  13. Arcturus says:


    An impressioning attack on this lock would use a tool very similar in principle to impressioning with tubular picks. Thus, no need for key blanks or reading of marks. In other words, the probe tips of such a tool would be set to the shallowest cuts to begin with, gradually moving to the shear line as you “twist push”. The probes stop at shear automatically. At least that’s the theory.

    It doesn’t appear that the egg supports this technique because the probe tips are threaded and impressioning requires adjustable friction locked probes that are free to move during the “twist push”.

    Re: bumping, I suppose the only way to know for sure if it’s possible or not is to try it. I plan on ordering a Geminy as soon as I can scrape together the cash for it. Until then, it’s fun to speculate. ;-P

    Are the pins really that sensitive to applied tension? You make it sound as if the mere weight of the cover plate is enough to create binding tension. Also, what is the strength of the pin springs? Firm? Medium? Light? If the springs are too firm, bumping might be difficult or even impossible.

  14. Henk says:

    The enemy of the shield is gravity,100%!!Tensioning the shield is possible.

  15. Henk says:

    I think those who own a shield have tried to pick it,and I remember all to well the excitement when I pushed,like mh said,the two normal outerpins first and the lock opened a bit.But soon I found out that this was not the way to open the shield,first the hard way,by trying to set the spoolpins,then the normal pins,because the spoolpins like to block the opening,once you,ve set them there is no way in blocking the shield by normal pins.Also it depends where the spoolpins are located,for example,when on positon 1 and 4 are normal pins and on position 2 is a spoolpin,this pin will block as soon as one picks position 1 and 4,you have to use too much force to push the spoolpin in,this will pull the shield back in its first position.
    When using the egg I,m pretty sure it does not matter what outerpin you take first,no use of gravity,;-)

  16. Walter says:

    I’ve build a bump-key a long time ago. I used a plastic (PVC) pipe that fits the geminy snugly. In it: a wooden stick with plastic screw sticking out on the right positions. The tube can be used to apply tension.

    I’ve never managed to bump open a lock. The bump-key is now in the personall collection of Klaus Drumm.

  17. Henk says:

    Neat design Walter,pity it didn,t work!I like it simple:-)

  18. till says:

    Of course you can use gravity as tension but i will be same as tensioning any other lock witch a weight using garvity.
    The problem for me is that i need the tensioner presure to get feedback from the lock.
    So i use a tensiner and put it between schield and the moving part.
    I could not get the tension very controled like this.

    What you are talking about is for me reverspicking not imping the Lock.
    So i understud you wrong.
    I think if you have a Gem. in your hand you will see better what is possible on it and what not.

  19. Lockpicker says:

    Does anyone have a link to a page where I could find the messuarments of the pins location, diameter and so on. I’m planning to make a impressioning tool, at least for the outer pins, inner pins seems easy to pick anyway. I’m will send this to someone who has a gemini lock and let it be tryed.

  20. henk says:

    Lockpicker:Read Han Fey,s article on Gemini,it will give enough info

    Till:Using gravity on Gemini will give you enough feedback!

  21. Arcturus says:

    Glad to hear you tried bumping. Did you use plastic to prevent aesthetic damage to the lock? Have you tried a soft metal like brass instead? I’m thinking a brass bumpkey could be mounted to a mandrel attached to some kind of pneumatic or electric impact tool.

    I think the following issues complicate bumping this lock:

    1) Ensuring that every pin on the bumpkey remains in contact with the lock pins until the very moment striking force is applied.
    2) All pegs on the bumpkey must strike the lock pins in perfect unison.
    3) Striking force must be sufficient to drive the top pins (including spools) well past the shear line, yet not so hard that the bottom pins follow after them.
    4) After striking, gravity tension might not be fast enough to block rebounding top pins.

  22. Henk says:


    1.When you use a bumpkey on a gemini contact is ensured,the pins are flat so that should not be a problem,all you have to ensure is that the “key” is placed firm,like the way Walter used by a piece of pvc.
    2.I think when one makes a bumpkey for a gemini,all pegs are alike,this way it should not be a problem to strike in perfect unison when they are helt in place like in 1.
    3.That,s the headache point I,m afraid:-0,trial and error…..
    4.There is a weight solution for that problem.

    Hope this answers some of your questions.

  23. Lockpicker says:


    Han Fey’s article DONT give such info. I could estimate diameters and distanses between the pins by using the Kaba key as reference, but thats not god enough for me…

  24. mh says:

    Till: What I described above is picking 2 pins, then decoding 3 pins in about 10 seconds.

  25. Henk says:

    Lockpicker:Send me an email,I,ll give you the measurements,pic,s.
    Send it

  26. Till says:

    @mh make the toool i will be happy to have a try on it.

    @Henk I thought you dont have a Gemini Shield.
    How do you know about the tension?
    Have you already tryed to pick it?
    Did you pick any lock with using only gravity?
    Greetings Till

  27. “I thought you dont have a Gemini Shield.”

    Well, Henk should have one, except his lock is in my hands at the moment, as I have had a little financial issues in posting it back :/

  28. Henk says:

    At the moment I have a shield and yes Jaakkoo has a shield of mine, also true,Jaakko,no worries I,ll see it back before it gets rusty:-)

    Till:I know about the tension because I tried.And if I only needed gravity to open a lock it would be too easy;-),But gravity helps a lot in this case.

  29. mh says:

    Till: Just take any variable setup key (my version is not for sale) and have a close look at the shape of the spool pins, you will see what I mean.

  30. AK says:

    So, how would you improve the security against picking of the Geminy?
    Certain changes to pin shapes / designs are one obvious way.
    I can also imagine adding a mechanism similar to a side bar with finger pins. It’s a bit difficult to describe what I’m thinking of in writing, but you can probably picture a number of possible setups.

    However, I’ve always seen it primarily as a protection against physical attacks and such and not really a pick-proof lock so there’s a question of how relevant it would be.

  31. ImSchatten360 says:

    The Drumm Geminy is a really cool shield. I like how it is build, how it looks and what it is for.
    But it is really not so hard to pick.
    But for protecting a cylinder against vandalism it is a great choice.

  32. ImSchatten360 says:

    The Drumm Geminy is a real cool shield. I like how it is build, how it looks and how it serves its purpose.
    It is a great protection against vandalism, but it is not so hard to pick.