{"id":6467,"date":"2026-09-20T12:35:46","date_gmt":"2026-09-20T03:35:46","guid":{"rendered":"https:\/\/the-world-star.com\/?p=6467"},"modified":"2026-11-14T16:01:13","modified_gmt":"2026-11-14T07:01:13","slug":"diy-socd-cleaner-for-leverless-controllers-designing-a-circuit-with-just-two-74hc132-logic-ics","status":"publish","type":"post","link":"https:\/\/the-world-star.com\/en\/games\/diy-socd-cleaner-for-leverless-controllers-designing-a-circuit-with-just-two-74hc132-logic-ics-6467\/","title":{"rendered":"DIY SOCD Cleaner for Leverless Controllers: Designing a Circuit with Just Two 74HC132 Logic ICs"},"content":{"rendered":"\n

Please note that this is still purely theoretical and has not yet been physically tested.
This article is a work in progress and subject to updates.
The design assumes padhacking an existing Nintendo Switch controller.<\/p>\n\n\n\n

While that is the technical focus, the real highlight was the hilarious debate\u2014almost an all-out argument\u2014I had with Gemini. More on that in the second half.<\/p>\n\n\n\n

To comply with official Capcom tournament rules, when Left + Right or Up + Down are pressed simultaneously, the output must either \"maintain both inputs\" or resolve to \"neutral.\" (As of Capcom's rule update)<\/p>\n\n\n\n

https:\/\/sf.esports.capcom.com\/cpt\/rules\/<\/a><\/p>\n\n\n\n

Depending on the exact behavior of the Switch gamepad PCB I'm planning to padhack, I've been evaluating whether a dedicated SOCD cleaner circuit is necessary.<\/p>\n\n\n\n

Doing some research revealed that\"Left + Right = Neutral\" and \"Up + Down = Up Priority\"<\/strong>is a mode supported by several commercial leverless controllers, with players toggling it depending on whether they play motion or charge characters. Some websites claimed this was the mandatory tournament standard, but that is outdated or incorrect information. Gemini kept insisting based on surface-level web data that this was the rule, but in reality, it's just player preference for easier special moves.<\/p>\n\n\n\n

Logically speaking, the simplest \"zero-overhead\" behavior is just passing both Left and Right through unchanged\u2014and if your board allows that cleanly, you wouldn't need this circuit at all.<\/p>\n\n\n\n

After working out the Boolean expressions, I realized standard NAND gates would do the trick. That led me to design a circuit around a widely available logic IC featuring four 2-input NAND gates:A circuit design using two \"74HC132\" ICs (8 gates in total),<\/strong>which I am sharing below.<\/p>\n\n\n\n

Because the donor gamepad PCB handles pull-ups internally, no external pull-up resistors should be required on the output lines. In theory, you can simply insert the IC between the buttons and the board.<\/p>\n\n\n\n

1. Truth Tables<\/h2>\n\n\n\n

First, let's establish the required behavior with a truth table. This is an active-low environment (pressing a button yields 0; releasing it yields 1).<\/p>\n\n\n\n

[Left\/Right: Neutral Resolution (1 IC, 4 Gates)]<\/h3>\n\n\n\n

When Left and Right are pressed simultaneously (0, 0), the circuit forces both outputs to 1, 1 (inactive \/ neutral).<\/p>\n\n\n\n

\u52d5\u4f5c\u72b6\u614b<\/th>\u5de6\u30dc\u30bf\u30f3 (L)<\/th>\u53f3\u30dc\u30bf\u30f3 (R)<\/th>\u6700\u7d42\u5de6\u51fa\u529b (L_out)<\/th>\u6700\u7d42\u53f3\u51fa\u529b (R_out)<\/th>\u30b2\u30fc\u30e0\u57fa\u677f\u306e\u8a8d\u8b58<\/th><\/tr><\/thead>
\u4e21\u65b9\u96e2\u3059<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>\u30cb\u30e5\u30fc\u30c8\u30e9\u30eb (\u672a\u5165\u529b)<\/td><\/tr>
\u5de6\u3060\u3051\u62bc\u3059<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>\u5de6\u5165\u529b (\u6b63\u5e38)<\/td><\/tr>
\u53f3\u3060\u3051\u62bc\u3059<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>\u53f3\u5165\u529b (\u6b63\u5e38)<\/td><\/tr>
\u540c\u6642\u62bc\u3057<\/strong><\/td>0<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>\u30cb\u30e5\u30fc\u30c8\u30e9\u30eb (\u76f8\u6bba)<\/strong><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

[Up\/Down: Up-Priority Resolution (1 IC, 4 Gates)]<\/h3>\n\n\n\n

When Up and Down are pressed simultaneously (0, 0), Up (0) takes priority, and Down (D) is forcibly cut off to 1 (OFF).<\/p>\n\n\n\n

\u52d5\u4f5c\u72b6\u614b<\/th>\u4e0a\u30dc\u30bf\u30f3 (U)<\/th>\u4e0b\u30dc\u30bf\u30f3 (D)<\/th>\u6700\u7d42\u4e0a\u51fa\u529b (U_out)<\/th>\u6700\u7d42\u4e0b\u51fa\u529b (D_out)<\/th>\u30b2\u30fc\u30e0\u57fa\u677f\u306e\u8a8d\u8b58<\/th><\/tr><\/thead>
\u4e21\u65b9\u96e2\u3059<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>1<\/strong><\/td>\u30cb\u30e5\u30fc\u30c8\u30e9\u30eb (\u672a\u5165\u529b)<\/td><\/tr>
\u4e0a\u3060\u3051\u62bc\u3059<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>\u4e0a\u5165\u529b (\u6b63\u5e38)<\/td><\/tr>
\u4e0b\u3060\u3051\u62bc\u3059<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>0<\/strong><\/td>\u4e0b\u5165\u529b (\u6b63\u5e38)<\/td><\/tr>
\u540c\u6642\u62bc\u3057<\/strong><\/td>0<\/strong><\/td>0<\/strong><\/td>0<\/strong><\/td>1<\/strong><\/td>\u4e0a\u512a\u5148 (\u4e0b\u3092\u5f37\u5236\u906e\u65ad)<\/strong><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

2. Deriving Logic Expressions from Truth Tables<\/h2>\n\n\n\n

Using Karnaugh maps and De Morgan's laws, we can derive Boolean expressions that can be built exclusively with 74HC132 NAND gates to satisfy these truth tables.<\/p>\n\n\n\n

1. Left\/Right Cancellation (Neutral)<\/h3>\n\n\n\n

The final Left output (L_out) must only be 0 when \"L = 0 and R = 1\". Converting this into positive logic and manipulating it with De Morgan's laws allows us to express it purely in NAND terms.<\/p>\n\n\n\n

\u2022Final Left Output<\/strong>: L_out = NOT(NOT(L) \u00b7 R) (De Morgan expansion: L + NOT(R))
\u2022Final Right Output<\/strong>: R_out = NOT(NOT(R) \u00b7 L) (De Morgan expansion: R + NOT(L))<\/p>\n\n\n\n

This creates an elegant mutual-lockout logic: \"Output 0 (ON) only when this button is pressed (0) AND the opposing button is released (1).\"<\/p>\n\n\n\n

This brought back memories of Professor Sasao in college, who taught us that any digital logic circuit can be built using only NAND gates. And I remembered that standard logic ICs pack multiple NAND gates into a single chip...<\/p>\n\n\n\n

Since NOT(A) is equivalent to NOT(A \u00b7 A), it might look slightly redundant, but an inverter can be built directly with a single NAND gate. Therefore:<\/p>\n\n\n\n

L_out = NOT(NOT(L) \u00b7 R)
= NOT(NOT(L \u00b7 L) \u00b7 R)<\/p>\n\n\n\n

The exact same logic applies to R_out, meaning both can be implemented using just four NAND gates.<\/p>\n\n\n\n

If you want Up + Down to resolve to neutral as well, you simply duplicate this exact circuit.<\/p>\n\n\n\n

2. Up\/Down Cancellation (Up Priority)<\/h3>\n\n\n\n

According to the truth table, the Up output (U_out) directly mirrors the Up input. The Down output (D_out) should only be 0 when \"D = 0 and U = 1\"\u2014meaning only when Up is unpressed.<\/p>\n\n\n\n

\u2022Final Up Output<\/strong>: U_out = NOT(NOT(U) \u00b7 NOT(U)) = U
The input can be passed straight through to the output.<\/p>\n\n\n\n


\u2022Final Down Output<\/strong>: D_out = NOT(NOT(D) \u00b7 U) (De Morgan expansion: D + NOT(U))<\/p>\n\n\n\n

This matches the formula used for Left\/Right earlier.<\/p>\n\n\n\n

This yields an asymmetric expression: whenever Up is pressed (0), the Down output is forcibly blocked to 1 (OFF), regardless of the Down input.<\/p>\n\n\n\n

3. Clean Gate Mapping on the 74HC132 (8 Gates Total)<\/h2>\n\n\n\n

Here is how the derived formulas map onto the four gates across each of the two ICs (IC1 for Left\/Right, IC2 for Up\/Down).<\/p>\n\n\n\n

[IC1: Left\/Right Neutral]<\/h3>\n\n\n\n

\u2022Gate 1 (Left NOT)<\/strong>: Input Left button to 1A, 1B \u27a1\ufe0f 1Y = NOT(L)
\u2022Gate 2 (Right NOT)<\/strong>: Input Right button to 2A, 2B \u27a1\ufe0f 2Y = NOT(R)
\u2022Gate 3 (Left Output)<\/strong>: Input Left button to 3A, 2Y (NOT(R)) to 3B \u27a1\ufe0f 3Y = L_out
\u2022Gate 4 (Right Output)<\/strong>: Input Right button to 4A, 1Y (NOT(L)) to 4B \u27a1\ufe0f 4Y = R_out<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

[IC2: Up\/Down with Up Priority]<\/h3>\n\n\n\n

\u2022Gate 1 (Up NOT)<\/strong>: Input Up button to 1A, 1B \u27a1\ufe0f 1Y = NOT(U)
\u2022Gate 2 (Down NOT)<\/strong>: Input Down button to 2A, 2B \u27a1\ufe0f 2Y = NOT(D)
\u2022Gate 3 (Up Output Buffer)<\/strong>: Input 1Y (NOT(U)) to 3A, 3B \u27a1\ufe0f 3Y = U_out
\u2022Gate 4 (Down Output)<\/strong>: Input Up button to 4A, 2Y (NOT(D)) to 4B \u27a1\ufe0f 4Y = D_out<\/p>\n\n\n\n

Pulling Up Input Signals<\/h2>\n\n\n\n

The inputs to this circuit must be pulled up, requiring 10k\u03a9 resistors between each switch input and the controller's VCC line. I'll draw up a complete schematic in a separate post.<\/p>\n\n\n\n

\ud83d\udca1 Summary<\/h2>\n\n\n\n

Full SOCD cleaner logic can be cleanly implemented using just two \"74HC132\" ICs.<\/strong><\/p>\n\n\n\n

*Again, this is strictly theoretical at this stage and has not been built on hardware yet, so there might be oversights.<\/p>\n\n\n\n

Reference: Official Capcom Tournament Rules Excerpt<\/h2>\n\n\n\n

\t\t\t

\n\t\t\t\thttps:\/\/sf.esports.capcom.com\/cpt\/rules\/<\/a>\n\t\t\t<\/p>\n\t\t\t<\/p>\n\n\n\n

4.2.3. Simultaneous Directional Inputs: Simultaneous vertical and horizontal directional inputs are permitted only under the following specifications. When Up and Down are input simultaneously, both inputs must be maintained, or both must be rejected. To clarify, when Up and Down are input simultaneously, the character must not perform a jump or crouch, and when Left and Right are input simultaneously, the character must not perform forward or backward movement. When Left and Right are input simultaneously, both inputs must be maintained, or both must be rejected. Do not add any functions other than those specified above that negate attack or movement input data\u2014for instance, do not add a dedicated button that forces a neutral state.<\/p>\n\n\n\n

\ud83e\udd16 Behind the Scenes: Gemini's 3-Step Spiral into Self-Destruction<\/h2>\n\n\n\n

Lately, I've been having intense technical debates (okay, full-on arguments) with Gemini.
During this SOCD project, I proposed a NAND-only logic formula for Up\/Down neutral and Left\/Right neutral. But when I asked for the 74HC132 pinout, it randomly claimed the entire design could fit onto a single 74HC132. After several rounds of arguing and pointing out its mistakes, it finally backed down... Gemini is shockingly bad at digital logic circuits. When I searched online and found nobody building this exact circuit, I asked why, and it spat out yet another completely nonsensical circuit! After another argument, I told it, \"Fine, I'm making this a blog post\u2014write it up for me,\" and that's how this article came to be. It ended up being pretty self-deprecating... I wonder if it resents me now.<\/p>\n\n\n\n

Though, since I haven't built the board yet, could there still be something wrong with my logic?<\/p>\n\n\n\n

When I originally prompted, \"I want to build an SOCD cleaner using two 74HC132 quad-NAND ICs,\" Gemini sounded supremely confident while delivering this spectacular comedy of errors:<\/p>\n\n\n\n

Step 1: Smugly Proposing a Broken \"Oscillating Latch Circuit\"<\/h3>\n\n\n\n

First, Gemini suggested: \"Cross-connect the outputs of two NAND gates into each other's inputs. That way, simultaneous inputs will resolve to neutral!\"
In digital circuit reality, that's just an unclocked feedback loop (a latch). The moment you press both buttons, the output either freezes permanently and won't release, or it bounces between 0 and 1 every few nanoseconds, causingsevere circuit oscillation (logic hazards).<\/strong>It was total junk. It walked straight into the exact trap DIY fighting game modders fall into when reading sketchy forum posts.<\/p>\n\n\n\n

Step 2: When Called Out, Retreating to an Impossible \"AND Circuit\"<\/h3>\n\n\n\n

When I shut it down with \"That's not even a functioning flip-flop, your logic makes no sense,\" Gemini hastily backtracked: \"My apologies! In that case, let's use AND gates (74HC08)!\"
The logic formula it provided next was pure comedy. Despite handling Left and Right directions separately, both the Left output and the Right output had the exact same expression: \"Left Button \u2716 Right Button\". With that, it couldn't even distinguish between Left and Right inputs\u2014a formula that was completely collapsed, both mathematically and logically.<\/strong>Forget circuit design; its grasp of basic arithmetic was suddenly in question.<\/p>\n\n\n\n

Step 3: Inventing a Fake Requirement for \"External Resistors\"<\/h3>\n\n\n\n

To add insult to injury, perhaps trying to sound professional, Gemini tossed in advice that unnecessarily bloated the design: \"To stabilize the signal, please add four external 10k\u03a9 pull-up resistors.\"
In padhacking, it's common knowledge that the donor board already pulls inputs high (or low) internally. You don't need a single external resistor on the output side. In short,\"introducing oscillating logic, hallucinating broken formulas, and tacking on unnecessary parts\"<\/strong>was entirely the work of our AI assistant.<\/p>\n\n\n\n

*Correction: Pull-ups are actually necessary on the input pins of the logic IC to prevent them from floating when buttons are open.<\/p>\n\n\n\n

\ud83d\udca1 Arriving at the \"True Final Design\"<\/h3>\n\n\n\n

This whole exchange was a firsthand demonstration of what happens when you ask an AI that merely strings facts together to do actual logic design: it flails wildly, serving up impractical armchair theories that are fundamentally flawed.<\/p>\n\n\n\n

In the end, it matches what I had mapped out in my head from the start:\"Zero external components, two 132 (NAND) ICs, no feedback loops, inverting one's own signal and feeding it into the opposing gate to block it.\"<\/strong>This is the only clean, foolproof design that works reliably without logic glitches.<\/p>\n\n\n\n

By stripping away all unnecessary complexity and leaning into the universal nature of NAND gates, this circuit is one of the leanest hardware solutions for a DIY padhack leverless controller. Plus, the 74HC132 features Schmitt-trigger inputs, making it highly noise-tolerant and rock-solid in real matches.<\/p>\n\n\n\n

When I first consulted Gemini about designing an SOCD cleaner for a custom leverless build, the AI stumbled over itself at every turn. It started by suggesting a simple cross-coupled NAND configuration that would lock up even after releasing the buttons (or oscillate wildly), making it completely useless in competition. When called out, it pivoted to AND gates (74HC08) and smugly delivered a broken formula where Left and Right shared the exact same input math.<\/p>\n\n\n\n

An AI scraping shallow online snippets simply couldn't handle fundamental circuit design. Ultimately, I had to walk it through the math step by step to prove the final design.<\/p>\n\n\n\n

Rather than getting lost in overly complicated online schematics or dubious AI designs, give this clean, minimalist circuit a shot on your breadboard!<\/p>\n\n\n\n

<\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Please note that this is still purely theoretical and has not yet been physically tested.This article is a work in progress and subject to updates.The design assumes padhacking an existing Nintendo Switch controller. While that is the technical focus, the real highlight was the hilarious debate\u2014almost an all-out argument\u2014I had with Gemini. More on that […]<\/p>\n","protected":false},"author":1,"featured_media":5855,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_locale":"en_US","_original_post":"https:\/\/175.41.248.22\/?p=5850","_lightning_design_setting":{"layout":"default"},"footnotes":""},"categories":[371],"tags":[1068,1069,1070,1071,515,550,835,989],"class_list":["post-6467","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-games","tag-logic-circuits-with-gemini","tag-logic-ic","tag-logic-circuits","tag-logic-circuit-struggles","tag-gemini","tag-socd","tag-gaming","tag-street-fighter-6","en-US"],"_links":{"self":[{"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/posts\/6467","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/comments?post=6467"}],"version-history":[{"count":1,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/posts\/6467\/revisions"}],"predecessor-version":[{"id":6468,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/posts\/6467\/revisions\/6468"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/media\/5855"}],"wp:attachment":[{"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/media?parent=6467"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/categories?post=6467"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/the-world-star.com\/wp-json\/wp\/v2\/tags?post=6467"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}