Solder the headers first
The headers come loose in the bag. A jumper pressed into a bare hole is a contact whose resistance changes when the wire moves, and every symptom it causes looks exactly like a bug in your sketch.
The one assembly step
Open the box and the modules have no connectors on them. The header strips are loose, and both edges of every board are bare plated holes. Four positions are left empty on the board for them: the design files call them H1 and H7 on the IN edge and U3 and U14 on the OUT edge. Soldering them is not a refinement; it is the difference between a display that works and a display that works sometimes.
The temptation is obvious, because the holes take a jumper wire and a wire pushed into one does make contact. It makes a contact of unknown, changing resistance, which is not the same thing.
Which board this is. This page counts the newer board, with six signal holes and a PWM row. The earlier version had five signal pins, no PWM and no dimming; everything below about solder applies to it the same way.
Run it with the wire pushed in. Eight bits go past the joint one at a time and the joint only has to fail on one of them; the ones that get through are fine and the ones that do not read as zero. What comes out is a byte that is not the byte you sent, which lights a digit that is not the digit you meant, and looks for all the world like a mistake in your segment table.
Male or female, and on which edge
The box has male pin strip to cut down and female sockets, and the footprints on the board decide which goes where. The two positions on the IN edge, H1 and H7, are footprints for male pins. The two on the OUT edge, U3 and U14, are footprints for female sockets. That is what the design files say, and it fits the use: pins on the edge your jumper wires meet, sockets on the edge the next module comes to.
What the files do not say is how two modules are meant to be held against each other, so this page does not either. If you are building a chain, solder one module's OUT edge and the next module's IN edge, and see the chain pages for what has to join.
How long it takes
Allow a few minutes a module: sixteen joints if you fit both edges, eight if you do not. Do it before the 12 V goes anywhere near the bench: soldering a board that is powered is how you find out which of your jumper wires was in the wrong row.
When it does not work
That is the signature of a pressed-in connection. A soldered joint does not care how the board is held; a wire wedged in a hole is a contact whose pressure changes with every nudge, so the fault follows the movement rather than the code.
On the IN edge, yes: all eight are connections you use, the six signal holes and the two power holes. The OUT edge matters only when a second module is joined to it. With one module, nothing in the circuit needs a header on the OUT edge, and you can fit it when the second one arrives.
This is the one step with no way round: the board ships with bare through-holes and no connector fitted to any of them. Soldering a through-hole joint is a short skill, and the workshop pages below cover it.
On the back, the face with the white printing and the holes you wire, with the plastic against the board so the pins stand clear of it. Use the same orientation on every module, so that two neighbours line up pin for pin.
What the 74HC595 does with the bits, and why there are two clocks instead of one.
Shift, then latch →Edit this page — content/books/seg4/solder-the-headers.mdx
Questions about this product
See what other owners have asked, and read their solutions.
4″ 7-Segment LED Display with 74HC595
Loading 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.