Chips & Components

Everything that isn't a breadboard strip lives in the Parts palette on the left: 74xx logic chips, CPUs, 65xx interface chips, memory chips, switches, LEDs, displays, resistors, oscillators, and power/clock bricks. This page covers finding a part, seating it on a board, and the (surprisingly varied) ways different parts rotate and flip once they're down.

The parts palette

The palette opens with every section collapsed, grouped by function:

• BOARDS — the breadboard kits and loose strips, pinned at the top (see The Desk & Breadboards).
• CHIPS — every 74xx logic family, folder-grouped, ending with the Interface group (the 65xx PIA/VIA) and the PROCESSOR group (the W65C02 and Z80A CPUs).
• COMPONENTS — Switches, Resistors, LEDs, Displays, Oscillators, and Power, in that shelf order.
• Memory — the ROM/RAM chips, pulled out of CHIPS into a top-level group of their own.
• ANNOTATIONS — labels and notes, pinned at the bottom (see Probing & Net Names).

The tray's right edge is a drag handle: pull it out to give long part names more room, and the width you leave it at is remembered.

Type in the Filter parts… box to search by id, title, or description — matching, it forces every group open so results aren't hidden behind a collapsed folder. Click any entry to arm placement — a ghost of the part follows your pointer until you click it down on a board, or press Esc to cancel.

Placing a DIP chip

A DIP chip always straddles the trench: half its pins seat in row e, the other half in row f, running the standard counterclockwise DIP numbering — pin 1 at the anchor hole in row e, pins continuing left to right along e, then wrapping back right to left along f. The notch end of the chip (or the dot beside pin 1) marks pin 1 and always faces left. Move the ghost over a pin-board and it snaps to the nearest legal seat; it turns red if the seat is already occupied or falls off the edge of the board.

Because a chip's footprint always occupies rows e/f no matter how it's turned, placing one is a matter of picking the column — there's no click-to-rotate step while placing a chip the way there is for a rail or a resistor; instead, rotation happens afterward (see below).

Placing a discrete

Most discretes — slide switches, push buttons, toggle buttons, LEDs (in their default horizontal form), single-digit 7/8-segment displays, and LED bars (bar8) — are linear: they seat along a run of adjacent holes in any single grid row (any of a–j), not just rows e/f. Drop one anywhere its footprint fits and every free hole underneath it is available.

A few parts don't fit that linear model:

• bar8iso — the isolated 8-segment LED bar — is packaged as a 16-pin DIP, so it straddles the trench exactly like a chip: anodes A1–A8 in row e, cathodes K1–K8 in row f.
• DIP switch banks (sw-dip1, sw-dip2, sw-dip4, sw-dip8) are likewise DIP-packaged (2/4/8/16 pins for 1/2/4/8 switch positions), straddling the trench the same way: each position's two facing pins — one in row e, one in row f — are its own independent SPST switch.
• Oscillator cans (osc-full, osc-half) are rigid four-cornered shapes rather than a line of pins — a full can is 7 holes by 4, a half can 4 holes square, with legs only at the four corners. A can can seat anywhere on the grid, including straddling the trench, since its shape (not a row) determines its footprint.
• Character LCD modules (lcd16x2, lcd20x4) are linear — a 16-way header along 16 adjacent holes in one row — but the module itself is much bigger than that row. See below.

Character LCD modules

The Displays group holds two HD44780 character-LCD modules: a 16×2 (the standard 1602A) and a 20×4 (the 2004A). They're drawn to their real sizes — 80 × 36 mm and 98 × 60 mm of PCB — and their screens are live: run the circuit, drive the module, and the characters appear on the glass.

Both plug in through a 16-way header along one row, and the pin assignment is identical between them, so what you learn wiring one applies to the other. The header runs along the module's top edge, which means the body hangs below the row it plugs into — so seat one on a bottom row (a) and it clears the board it's plugged into rather than covering it. The placement ghost shows you the whole module, so you can see this before you click.

Driving one is the ordinary HD44780 parallel bus: put a command or character code on DB0–DB7, set RS (0 = instruction, 1 = data) and R/W (0 = write), and pulse E — the byte latches on E's falling edge. Wire VDD/VSS to a 5 V rail. V0 (contrast) and A/K (backlight) are present on the pinout but cosmetic here. During a read the module drives DB0–DB7 itself, so tri-state anything else sharing that bus.

Both modules show the same controller datasheet in their pin-assignments window — it's one document, because RS/R/W/E, the bus and the address maps are the controller's and are identical across the two sizes.

Rotating & flipping

Rotation behavior is not one rule for every part — it depends on what kind of part it is. This is the part worth reading carefully.

Chips (R, mid-drag or while selected). A DIP chip's footprint maps onto itself when flipped — same two rows, same columns — so flipping only reverses which physical pin sits where; the chip never has to move. Select a placed chip and press R to flip it 180° in place, or press R while mid-drag to flip it before you drop it. Either way its pin-assignments window updates to show the new numbering.

bar8iso and DIP switch banks (R while selected). The isolated LED bar and every DIP switch bank are DIP-packaged, so they flip exactly like a chip: R turns the part 180° in place, the same holes, only the pin numbering per position reverses (a switch bank's own position states don't move — position 1 is still position 1, just wired to the opposite pins now). Neither has the turn-in-hand behavior of a plain LED — treat them as chips for rotation purposes.

Resistor and LED (R, both while placing and once placed). These two-lead parts start in a horizontal footprint form (pin 1 and pin 2 a fixed span apart along one row). Press R while the ghost is armed to turn it into a vertical two-free-ends form instead: pin 1 stays at the anchor hole, and pin 2 becomes a free lead that can land on any other free hole — including a hole on a different strip entirely, such as reaching up to a power rail. Each R press while placing steps the ghost a further quarter turn, cycling through all four compass directions before repeating. Once the part is placed, select it and press R to rotate it 90° at a time — pin 1 stays put and pin 2's lead swings around it, hunting for the next free hole to land in.

Oscillator cans (R, but the exact step differs by state). A can spins around its own centre, not around one pin, and the step size changes depending on whether you're still placing it:

• While placing, R steps the ghost a full 90° quarter-turn each press, so you can hunt through every orientation for one that fits.
• Once seated and selected, R behaves differently for the two sizes: the square osc-half can still steps 90° at a time, but the rectangular osc-full can jumps straight from 0° to 180° (and 90° to 270°) — because a non-square footprint only has two genuinely distinct orientations once it's down (rotating it a further 90° would just retrace the same two footprints it already swept through while placing).

LED polarity (F, while placing only). Independent of rotation, press F while an LED's placement ghost is armed to flip which lead is the anode and which is the cathode, before you click it down.

Moving a seated part — with or without its wiring

Dragging a placed part re-seats the part and leaves every wire where it was. That is often what you want while a circuit is still bare, but once it's wired it quietly changes the circuit: the pins land on different column-halves, the wires stay in the holes you laid them in, and what was pin 1's input is now pin 3's.

So, exactly as with pulling a mated strip out of a board group, hold a modifier while you start the drag:

• Option-drag — moves the part and everything plugged into it. Every wire end sitting in a column-half one of the part's pins occupies comes along, keeping its own row and its offset from the part, so the circuit after the move is the circuit before it. The far end of each wire stays put.

A resistor or LED plugged into the part comes too — by the leg that is connected to it. A leg in one of the part's column-halves is attached to it exactly as a jumper in the next hole along is, so it travels; the other leg stays exactly where it is and the part simply bends around it, the way a resistor's legs bend on a real bench. Plugged in at both legs, it travels whole instead. (A chip's pins can't bend, so a chip next door is never dragged along — it isn't a lead, it's a body.)

Only what is actually connected to the part comes — a wire or a leg in a column the part merely spans without having a pin there (a push button reaches two holes three columns apart, not the one between them) is left alone, because it was never connected to it. And nothing chases further than one step: what travels lands in the same column-half its pin lands in, so a rider never leaves anything of its own stranded behind it.

To see what would come, hold Option. With a part selected, holding Option rings every wire end and every leg that would travel with it, and releasing Option puts the rings away — so you can check before you commit to the drag rather than discover it during one. Something connected at both ends gets two rings, so a resistor that will travel whole reads differently from one that will bend; something connected at one end gets one ring, on the end that moves. A part with nothing attached simply shows nothing.

The drop is all or nothing. If any of those has nowhere to go — its hole is taken by something that isn't moving, it would run off the end of the strip, or the bend would squeeze a resistor's legs closer than its body is long — the part and everything it would have carried turn red, and releasing puts everything back. Half a move would silently cut the connections it left behind, which is the very thing the gesture exists to avoid.

A resistor or LED is dragged by its two ends rather than by a footprint, and that makes one distinction: dragging its body moves both legs together, so Option carries its wiring exactly as for any other part; dragging one leg is a re-bend, and carries nothing — that leg can land in any hole, at any angle, on any strip, so there is nothing for a wire to follow. (An LED lying flat is only one hole wide, so there is no body to grab between its legs; stand it up with R, or move it as part of a selection.)

Two details worth knowing:

• Whether the wiring comes is decided when you press, not when you let go. Once the part is in hand, the set is fixed.
• Moving across the trench takes the wiring over with it, keeping the arrangement. Rows a–e and f–j are separate nodes, so a wire that stayed in its row would be left in the half its pin had just left — it travels the same number of holes the part did instead, so a wire two holes from the part is still two holes from it, on the same side. Drag far enough that the wiring would run off the edge of the board and the drop is refused; a row nearer the trench fits.
• Moving to another board works the same way — the wiring lands in the matching holes over there, and anything the wires reach back to stays where it is. While the part is over the gap between two boards there is nowhere for it to land, so the wiring sits where it really is and turns red until the part is over holes again.

Shift is not this modifier: Shift-drag always rubber-bands a selection, including when the press lands on top of a part.

Moving several parts at once

A selection drags as one unit. Marquee a group of parts (Shift-drag), or add them one at a time with Cmd/Ctrl-click, then drag any one of them: every selected part travels by the same amount, keeping the arrangement exactly as you built it. Power and clock bricks count as parts here, and come too.

Everything above applies unchanged, just to the whole group:

• Option takes the wiring with it — every wire, and every resistor or LED leg, riding any member. A wire between two selected parts travels at both ends; one that leaves the group keeps its far end where it is; a resistor plugged into a member bends after it exactly as above.
• Holding Option rings every end and leg that would travel, across all of them.
• The drop is all or nothing. If one part, or one wire end, or one leg has nowhere to land, every member and everything riding it turns red and releasing puts the lot back — a group that dropped some of its members would be a rearrangement you never asked for.
• The whole move is one undo step.

A board in the selection is the one case that declines: strips have their own drag, which carries everything seated on them under rules a part re-seat knows nothing about. Pressing a part then does nothing and leaves the selection alone — Cmd/Ctrl-click the board back out and drag the parts, or grab the board itself, which drags its own snapped group with everything seated on it and replaces the selection the way any plain click does. And a plain click inside a selection narrows it to the part you clicked, exactly as clicking a part outside one does.

One thing a selection does not carry: a wire you selected that isn't plugged into any of the parts moving. Wires travel because they ride a pin, not because they are highlighted — so a wire selected on its own (to delete or recolor it) stays exactly where it is while the parts move around it.

Occupancy — one hole, one lead

Every hole on a breadboard — and every terminal on a power/clock brick — holds at most one lead, whether that's a chip pin or a wire end. Placing a part checks every one of its derived pin positions against every other part and every wire already on the desk; if any pin would land on an occupied hole, or off the edge of a board entirely, the ghost turns red and the drop is refused. This is the same rule a wire's endpoints follow (see Wiring, Nets & Buses) — chips, discretes, and wires all compete for the same holes, with no separate bookkeeping for any of them.

A rotated resistor or LED's free lead is the one case where a pin can legally resolve to nothing — if you later move or delete the strip under that lead, the part stays exactly where it is and that leg simply floats, unconnected, just as it would on a real bench.

The pin-assignments window

Right-click any placed part — chip, discrete, or brick — and choose Pin Assignment, at the top of its context menu, to open its floating pin-assignments window: a diagram of every pin/terminal and, for most chips, a cropped datasheet excerpt below it. A real chip's diagram stays fixed at its canonical layout no matter how you've flipped it on the desk (it matches the physical part, not the placement); a DIP-packaged discrete — bar8iso or any DIP switch bank — is the exception: its diagram reflects its current R flip, since it has no real notch of its own. See The Chip Library for the full detail on what the window shows and how it sources its datasheet crops.

See also: The Desk & Breadboards for how boards and strips work, and Wiring, Nets & Buses for connecting components together once they're placed.