Korean arcade parts discussion


So good. Can use a high shore grommet and adjust the activation window window of the cardinal directions to ~0 and get godly execution.

Edit: BTW, did I say the golden base and actuator system is amazing, if so reiterating. So sexy and finely machined in comparison to the shotty molds of other k-levers.

For anyone interested in steps to replicate this mount, here is the method I used:

Tools used:
Drill press
Table saw
Various drill bits
Thread tap
Coping saw
Carpet tape
Forstner, spade bit or hole saw

Almost any solid ~3/8" material will work, I used some hdpe (seaboard) ordered from amazon
12mm M4 screws
Matching nylon washers
Metal washers 6mm i believe

I have a switch plate from a myounshing air bag lever. I cut the dimensions of my 3/8" material to the same dimensions as the switch plate. I then used carpet tape to secure the switch plate to my material. This allowed me to use the switch plate as a guide to accurately drill holes for mounting it to various K-levers. You can use various bits to drill the center hole to allow room for the shaft/actuator to move freely, forstner, spade or hole saw for example. I don’t recall the size needed, maybe 25mm or 1".

Luckily the base of the Magenta is also the same size as the switch plate. So again, I carpet taped it too my material and used it as a guide to drill the pilot holes which are used to mount the magenta pcb. I ended up drilling the pilot holes out to 3.5mm and using a 4mm tap to thread them. It may be better to use a 2.5mm drill bit and a 3mm tap to allow some play room for centering the Magenta in case holes go off center at all.

You also have to drill out recess areas on your material to allow room for the screws and bolts used to secure the various pieces of the k-lever together to nest. Now you can screw the base material you created to most any K-lever using their standard screws. You can use nylon washers to properly space your magenta PCB, screwing it directly into your material.

Regarding the actuator, I ended up modifying a standard myoungshing actuator. I had to shorten it a bit, I don’t know the specific amount off hand. I drilled a 10mm recess in the bottom of the actuator for the brass piece of the magenta to nest, the bottom of this recess stops at the top of the magenta shaft. You can then use this as a guide to know where to cut down the actuator to. You can cut it to the top of the brass piece that way it sits flush. I used a vise and a coping saw to make the cut, but you could also mill it down with a sander or other various ways.

The magenta shaft is a bit shorter than a standard k-shaft. I used about 1/4" of metal washers to raise the bat top to the standard height. There wasn’t much thread left on the shaft, but enough to get a secure fit. I haven’t had any issues with it. I wish the shaft was a bit smaller in diameter, I believe it is 9mm, but that is just my preference.

If mounting to a Golden base the stock screws used with the top plate stick out to far at the bottom, so you will need to use the screws from a myoungshin lever or equivalent.

Initially I saw the potential in using the magenta, but noticed a decline in my execution. I went back to it a few times and found that having the directions activate closer to the center allowed for much better execution with a Korean style lever. This activation distance will vary depending on the tensioner used. I only play Tekken, so I can’t speak to execution in other games. This setup suited Tekken movement and Mishima play for me.

The best thing about it is being able to watch your input motions live while in practice mode and make conscious efforts to correct your bad inputs rather than blindly repeat motions in practice mode. This allows you to develop proper muscle memory much faster. Similar to practicing things slowly on an instrument and building up speed.

I wouldn’t say this is 100% better than switches; I can still execute well with switches, but it is very neat to mess around with and for now I’m going to stick with it.


As someone who has had both I don’t think that’s the case

The fanta has a really heavy rubber grommet and if you grew up playing it it makes sense that you’d prefer it to the Crown stuff which tends to be a lot lighter. The downside of being lighter is that it won’t return to neutral as fast but coming from like a Sanwa JLF, even the crown will return to neutral faster

It’s mostly preference I think


Modified one of my golden shafts to work with the magenta over the weekend, it was easy to work with since it’s brass. I probably wouldn’t have been able to manage with a steel shaft.

Used a steel button head screw to secure the magnet seems to work fine. I like the phillips head better than the t6 torx used by default, t6 torx bits are so small they tend to round over easy in my experience. you have to make sure the screw is secure though, if it comes loose and makes contact with the PCB components it could damage the magenta, I found that out the hard way. I need to re-solder a component on one of my megentas due to not fully tightening the screw while testing a few things.


So… Good job in voiding any warranty on the magenta as well as throwing any sort of accuracy out the window. The head on that screw is to big and will not position the magnet at the proper distance from the hall sensors on the PCB.


Nah, it works fine. It’s only slightly taller than the stock screw, just a bit bulkier on the edges. Not really worried about warranties, but to each his own.


Has anyone on here ordered from IST mall? I made my order on the 30th and they’ve sent me the invoice but no shipping confirmation (I got K-packet)


I’ve ordered from them a few times, though I get EMS shipping. Usually they provide a tracking number within a day or two in my experience. I can’t speak for K-packet though.


K-Packet is pretty slow but it should come. I think it took about 10-11 days for my order to come to the mail.


Are Green Fanta Sticks still available? Haven’t seen Etokki restock in a while.


Sanjuks stick V5 with short hinge lever switches :

This type of hinge lever length is what has been and still is in use for seimitsu ls-5x/ls-6x series.

Sanjuks , is it really the first k lever with short lever switches?
Crown 306 2nd gen used 305 lower body with same spacing etc (specs are the same) but instead of the hinge lever-less (basic pin plunger) switches and square actuator seen in 305 (which was a partial knock off of Happ/IL sticks) , the actuator in 306 2nd gen is round and the switches are short leaf type or so it seems at first glance.
…but not really. The switches are not hinge lever length nor short lever length. It’s a custom length in between.
Without even measuring (I did measure to check though), hinge lever pivot point is slightly behind where COM connector sticks out, but the protruding length of the leaf does not extend as far as “normal” hinge lever.
So yes there were already (custom sized) shorter hinge levered switches in k -levers, but no commercial nor custom joystick with “official” catalog standard short hinge lever.

Custom hinge levers are unusual but not unseen, recently seimitsu had omron make angled hinge levers (stock length though) to replace the panasonic switches.

notes :
SANJUK stick has rotated locating/securing pins 90° for crown type grommet. Crown stock locating pin array is such that they are pointing to CARDINAL directions.
Sanjuk has these locating nubs placed in 4 CORNERS, probably following the general “custom” trend in Korea to have looser feel in diagonals (modders use many different techniques : holes in grommet groove to loosen tension(fujin) , fanta style grommet&crown grommet with corner holes of same dimension as locating cardinal holes/notches (ASI-Wazwuz), Filed down (very coarsely) underside of fanta collar plate only in the corners to reduce clamping pression and permit a form of slippage (Green Lever) .

Short lever hinge switch version of Sanjuk has also an actuator whose switch contact side is BEVELED, unusual in k-levers but classic for other types (japanese etc such as JLF).
Imho it’s a good idea since the deadzone is small in neutral (zero with oversized actuators which contact the levers in neutral), but it gives some SPACE to avoid having too large a window for diagonals for those who feel uncomfortable with this (very personal) and thus reduce unwanted diagonal inputs (recurrent problem with oversized actuators and or small switch array with zero deadzone).
As a comparison, usual k-lever actuators are cylindrical and non beveled or VERY SLIGHTLY so, even if one part has larger diameter it is not beveled (not speaking of the edges, buit the SIDES) and the sides are parallel to the other section of smaller diam.
There are no k-lever commercial actuators that have a bevel angle as important as sotck JLF top portion actuator.
Kowal did in fact a korean type actuator mod for his JLF oversized actuator, but I am not a fan of this mod since in a JLF, at max throw (thus max return force) the sharp corner of the pin presses on the Kowal actuator’s side which is non-angled. Straight is more suited to levered switches, but a bevel can make things easier in some circumstances.
Stock JLF bevel angle is calculated so the actuator side is flat with pin plunger surface at a critical point (close to actuation point)

The bottom opening guide on all his models (see here for closer look)

is in fact a squircle

compare to usual truncated circle shape used in fanta & crown

squircle is similar in function to rounded square as the corner window input angle is higher than if it were a true circle, unless you have a squircle chose shape is closer to a circle in that case it performs closer to a circle guide (the latter having the smallest possible window for corners, same as octo gate when in pure 45° diagonal)
Sunjuks squircle opening is rather circular in shape though(similar but slightly less round than Crown sdl301 squircle gate)

Sunjuks and Kaze both claim snap microswitches have hinge levers than “bend” too much. It is not true, the forces from Gersung A2 or A3 are insufficient to , in opposition to the force the player is putting in, bend the metal plates. I’ve never seen switches in ls-32, jlw ls40 or others with worn “bent” hinge levers unless the arcade tech bent them on purpose himself to compensate for various reasons( deazone, uneven centering etc, bending with pliers is NOT RECOMMENDED, btw Hibal also does not recommend this in his switch mod tutorial)
Sunjuks’ “solution” is different from that of KAze, as the latter proposes “signature switches with stiffer hinge lever alloy” (unproven at this point and even then useless), while Sunjuks opts for using short lever switches a la seimitsu ls-56.
Standard hinge levers DO move during use, shifting side to side a little bit and also twisting off axis a bit also in joystick use (as the actuator tilts) , due to tolerance issues which are part of the design, as the hinge levers pivot without lubricant nor bearings, the space is implemented to permit a slight loose fit so the wear is minimized (the leaf rubs on the plastic case) and the lever can move very fast up and down .
Surprise surprise, even short levers shift side to side with the EXACT same tolerance (but do not twist as much), since the casings have invariable dimensions concerning the pivoting “openings” for the metal hinges and there is a whole variety of different length and shape metal hinge lever types available.
most of you have noticed a certain “wobble” even with the highest quality switches, it is natural and intentional.

Real benefit of short hinge levers for switches is that the TIP of the lever is located barely past the pin plunger contact point, the only place besides this small area where the actuator will slide is a bit behind the pin plunger position, where the actuating force is HIGHER than above pin plunger point, but again on a small zone. you maximize actuation force then compared to longer hinge levers whose actuation zone is closer to the tip and never far back beyond pin plunger .
Reminder: pivot point for hinge lever is behind pin plunger , so if you go BETWEEN pivot point and pin plunger, the necessary force is SUPERIOR to the pin plunger leverless actuation max force Metal blade pivot point is 8.3+2.8mm=11.1mm away from edge furthest from pin plunger (with Gersung) or simply 8.3mm away from center of pin plunger.
So in a way YES short lever switches do show higher stiffness in use but not because longer ones have “soft” blades, because of the law of the lever you all learned in school .


More about Sanjuks (with standard levered switches)

Here he uses collar with 15% reduction in collar internal diam and shaft 4mm lower (compared to Crown 303) , squircle bottom guide(see post above), and ZIPPY switches.
He does mention that squircle is more precise than pure circle to get a wider corner window input range (although his is very close to circle and will present a rocking motion if trying to lock in corner).

Speaking of the switches, Zippy =/= zippyY with 2 “'y” which makes seimitsu clones.
Zippy, a taiwanese brand, makes the famed LOW actuation force switches with quite deep actuation point sold in various arcade shops as a “silent” mod and or “ultra low force” actuation microswitches . Their catalog offers as low as 5-9gram force options if you want, depending on actuator chosen (lever type).
Quality is acceptable but not optimal as Sanjuks claims. I’ve used them extensively, although he probably uses standard k-lever actuation force zippy switches ( A2-A3 gersung forces ) whose actuation point is thus less deep (but more so than brand names such as omron and panasonic).
He also claims Zippy is fat better than Deco, which is debatable. I only tired old Deco type, not recent prodution levers with Deco switches(no clue if the “new” ones are NOS or new production also). Most used series from Zippy is the VM (Omron V series anyone!!! ) , their mechanism is cloned, prices are quite low, actuation range is vast ranging from ultra low to realiv(good point there). PAS used to sell some higher actuation force zippys even at one time, in the 125gf range, but it is unavailable now from their store.
VM series zippys range in hinge lever(rear position pivot) from 9 to 220gram-force, and 20-400gram force in short lever type (as used in ls-56 and sanjuks) which has same values as pin plunger.
Reminder : Force is measured not by joystick actuation standards, but by pressing directly above pin plunger for said model and short lever, and for “hinge lever” ( medium) and “long hinghe lever”, forces are measure when pressing exactly at a certain point near the tip of the metal blade (just a reminder). In joystick configuration, the actuator hits on a zone specified by the design whcih can sometimes cover a large area on the hinge lever, it is not always the zone of testing as per catalog specs so the leverage will depend on where you are situated relative to the pin plunger.
back to catalog values, Zippy’s highest actuation force for pin plunger is 400gf, Gersung has 500gf,
Happ uses “e-switches” which are similar to zippy switches. Acceptable but far from Cherry quality (original supplier)

Sanjuks mentions how tolerance in silicone grommets from CROWN is loose in terms of Shore hardness ( up to 20% VARIATION which is enormous, search forum for exact tolerance I posted it before many times),
It doesn’t mean that color codes corresponding to increasing shore hardness values are a bad idea, quite the contrary it means you NEED more consistency in the production to avoid having :
20A <= “official crown 25A” <= 30A or worse once you up the scale ,
28A <= “official Crown 35A” <= 42A
…or even
36A <= official 45A <= 54A
This sloppy tolerance was for 307 and 309 grommets as specified before on ISTmall (warning notice is now absent from product page ) , infiltration silicone type was said to be better accodrding to posts on Ruliweb during release, but hardness variation % was not specified, Sanjuks says Crown Infiltration grommets are still inconsistent in hardness.

Sanjuks also confirms what I told to some forum members here via PM that the tendancy is to get closer to a JLF shaft in custom models.
He compares JLF shafts which he says are lathed, to poorly chromed stock k-lever shafts (Crown here) which reading between the lines he supposes may still have unequal diameters along the rod and could possibly be slightly curved (he never sid it, remember we’re reading between the lines) . It’s true older fanta and crown models sometimes had curved shafts , even off center grommets & bottom guides which were still sold instead of being put aside during quality control .

On this post, https://cafe.naver.com/sanjuks?iframe_url=/ArticleRead.nhn%3Fclubid=27718294%26menuid=7%26boardtype=I%26page=1%26specialmenutype=%26articleid=670%26referrerAllArticles=false
Sanjuks V5 in proto stage is said to be using different top collar aperture diameters with CORRESPONDING contact angles (with shaft). Cool to mention it, but in fact this should be mandatory for making custom gates, unlike some custom offerings which are centered around the aperture general shape and size solely, without considering the slating position the shaft or actuator will be in at max throw.
This being said, korean grommets, especially the softer ones, see their pivot point shift a bit during use, much more than with pivot bearings in japanese designs where the bearing slides and is guided, so the correct angling will only minimize slightly the problem, it is not a “final cure” to shaft vs collar wear (angle changes when pivot point shifts, simple as that).

Now it would be nice if those who did purchase one of the sanjuks lever could post their impressions here.


Remember the OTTO DIY stick mentioned by Hibachifinal on his blog?

He tested the JLF body partial repro with “V5” japanese type sliding pivot ( OTTODIY socket is groove-less) and “V2” korean grommet modules as well, also with jlf pcb switch option and loose switches (which hibachi says is “hayabusa” switch option, but already existed before hayabusa release as a mod for JLF).

OTTO DIY store:

The type used by Hibachi is one of or the latest grommet type with a ring covering the grommet edge .

Initial grommet housing had no ring, just a machined bushing to fit the shaft and a machined housing (housing HEAVILY inspired by JLW pivot housing, but without the 4 screws which are replaced in the OTTO DIY design by 4 locating notches)
The designer counted on the stock JLF steel mounting plate to hold the grommet’s edge, except that the steel plate’s opening is too wide to do that correctly… so he needed to revise this aspect.

Later design upgrade (similar to what Hibachi purchased) showed an additional top plastic ring with matching 4 locating notches, to be placed right above the plastic grommet module, slightly recessed now.
Notice how the extra securing RING has a much smaller internal diameter than the JLF steel mounting plate !
This way the grommet is secured.
Is the zone with exposed silicone large enough so the bushing doesn’t contact the ring at max throw? only those who tested could tell.
notice also the bottom opening in the revised version is much larger than the bottom opening in the first iteration.

Notice also the whole grommet design is very close to what is seen in earlier grommet sticks such as Suzo 500 rubber version where the grommet is a flat sided cylinder with no notches whatsoever from preventing it to rotate inside housing during use (which could wear the grommet) . Fanta and Crown grommets do have these notches, and for a reason. Some were not bothered with a non secured rubber ring using the suzo 500 (and other similar designed grommets) though and complained only about centering issues (namely an important deadzone issue)

Shaft pullout in OTTO DIY levers is prevented due to classic JLF/Hayabusa shaft assembly , with a supposedly weaker than stock spring when using grommet V2 module, and OTTO actuator (which like stock jlf part has a cavity to house the spring). The whole is supposed to work with stock JLF shaft .

notice actuator is straight (non beveled ) type like the Kowal one (I do not agree with this actuator design as explaiend above) , and oversized parts are available.

Two other things: the different “pivot” modules can fit the JLF repro but also a " + " shaped type body which is mounted instead with standard HINGE LEVER switches (also the official name for its fully assembled versions are accordingly V2+ and V5+, thanks to hibachifinal for clearing that out) , the latter cross shaped body is DEVOID of clip-on tabs and is assembled together with its matching restrictor plate and mounting plate ONLY by fastening with screws (think ls40 etc), although the “+” features additional locating/securing tabs as seen in hayabusa body except here they are located a bit further towards the outer edge, right next to the mounting screw, the locating tab type is in return coming from JLF anyways despite the custom tab spacing. LS-40, LS-56 etc levers which assemble solely with screws do not have bodies with locating tabs that go through the gate, instead the gates themselves have locating pins. for example with ls-56- the locating pins fit exactly inside the 4 empty remaining microswitch mounting holes).
Hibachifinal told me the “+” switch spacing is the same as ls32/ ls40 /JLW.

It’s a little bit off topic but the sliding pivot type bearing module was available a certain time ago with a custom polished stainless steel housing , now the only available housing is made of plastic, perhaps steel will be avilable soon again on special order you need to check that out soon. EDIT: as of now november 2018, early pivot modules were made of POM, but now OTTODIY switched to teflon based polymer while some parts are still made out of nylon, also the pivot type seems to have transitioned from ball in conical housing type bearing (similar to what is found in HORI hayabusa stick ) to the traditional JLF and Seimitsu ball-in-socket (socket matching the spherical contact area of the pivot bearing), due to constant requests from certain local gamers.
Notice the absence of any grooves on both machined pivot bearing and housings (both plastic or metal) , so it will be tough to evacuate dirt particles or hold some grease (that’s what grooves are made for) , also grooves (as well in jlf socket and hayabusa pivot bearing-latter is grooved like agetec bearing) minimize surface contact to help getting better speed and lower friction.
Perhaps in a third version this will be available?
There are frequent revisions and upgrades apparently, including material selection, so keep an eye on available stocks.


No, they’re out of production for the time being. You could get a Taeyoung Fanta and mod it with the Golden Fanto Mod by Arcade Stick Indonesia to get a Green Fanta clone. Or you could get an original Green Fanta from eBay or the Korean arcade parts Discord, but you’d pay a lot more compared to a Golden Mod.


Hey ! im about to buy a myoungshin fanta, but im wondering if i choose the normal one or the airbag one. any suggestions ?


Can you even easily get your hands on an airbag version?


well, i just bought an airbag fanta from a handmade arcade stick builder, he said that the airbag version its better, so i decided to try it out, i posted the question between choosing the normal one or the airbag one beacuse i dont see any airbags, so i was wondering that the airbag maybe its not that great hahah


“Airbag” refers to a name, it doesn’t use actual airbags inside the joystick or anything like that.



looks good! even though its Unpure. My Aryan Korean Blood boils at that shit @butteroj lol!


Korean Lever Breakdown Episode 3: 303 Series

Finally got the 3rd Episode up. I hope to get the next episodes out quicker. I took a different approach to this one, try not too get too aroused looking at my pretty face. I know I look like an idiot trying to read from the monitor as I talk. Truth is I got a memory like a goldfish, So either I do this, or end up doing it multiple times lol! anyways Its good to see some life still in this place, and more ppl coming in!