Why six sizes
Six boards from 0.96 to 3.5 inches, and three of them have exactly the same number of pixels. Choosing between those three is choosing how big a pixel is, not how many you get, and only the largest board adds any.
Six, because the right size is a thing you find out
A display is the one component you cannot read your way to. Pixels tell you nothing about whether a number is legible from the sofa, and a diagonal in millimetres tells you nothing about whether your enclosure has room. The only way to know is to put it where it will live and look.
That is what six sizes are for. Each is a board with its panel already seated, sold alone or together in the 6-pack. Prototype on whichever is nearest and decide by eye: the boards carry the same circuit, so the swap costs you one constant in the sketch.
Three of them are the same picture
Step the control through 2.0, 2.4 and 2.8 inches and the resolution does not move: 240 × 320 on all three. The library gives all three the same controller family, so the choice between them is a choice about pixel size.
The 2.0 inch has the smallest pixels of the six. Text at the same font size looks printed rather than drawn, and a chart has room for gridlines. The 2.8 inch has the largest, which is what you want when it is read from across a room and not from a desk.
Only the 3.5 inch adds pixels: 320 × 480 is twice the picture of the 2.8 inch. It also takes twice as long to redraw, which is a real number and gets its own article.
And one of them is not like the others
The 0.96 inch is odd in more ways than its size. It is 80 × 160, a strip rather than a screen, and the one board whose backlight is switched on the high side of the lamp, so its backlight pin turns the lamp on by going low. The library handles that; a driver you write yourself will not, unless you know. The backlight article draws both circuits.
It is the right board for one reading in a space where nothing bigger fits: a temperature, a battery bar, a state word.
What none of them do
There is no touch on any board and no microSD socket. Both turn up on displays that look like these. If you need touch you need a different panel; nothing here can be made to have it.
When it does not work
The 2.4 inch. It is large enough to read across a desk, its 240 by 320 is the resolution most example code assumes, and a full redraw is short enough that clearing the screen every loop does not look broken yet. Moving up or down later costs one constant in the sketch and no rewiring.
No. It is the same 240 by 320 on bigger glass, so it is coarser up close and easier to read from further away. The three middle boards are one picture at three physical sizes. If you want sharp, the 2.0 inch has the smallest pixels of the six.
Same glass, described the other way round. The library treats the 0.96 inch as a portrait panel 80 pixels wide and 160 tall at rotation 0. The board mounts it turned a quarter, so the strip you see is landscape, and setRotation turns the picture to match. Neither number is wrong.
The vendor drawing for that glass is labelled 1.77 inch, 128 by 160. The product and the library constant (LB_TFT_18) use the round number. The active area on that drawing is 28.0 by 35.0 mm, which is what this book uses for its pixel density.
None of the six. There is no touch controller on any board, no connector for one, and no microSD socket either. They are display-only boards, which is why the whole interface is eight signals.
What the 6-pack holds, and what the boards themselves already are.
What is in the box →Edit this page — content/books/lcd/why-six-sizes.mdx
Questions about this product
See what other owners have asked, and read their solutions.
This page covers several products. Choose yours to see the right discussions.
Discuss this article
Ask about this page. The answer stays here, on the page it belongs to, for whoever hits the same wall next.