Quick answer: a Shaker only processes Wet Sand placed on top. If it looks broken, first confirm the input is Wet Sand, then clear the Gold drop space below, remove any non-Gold solid trapped in the machine body, and give Residue a downhill side exit instead of letting it bury the top.
The machine itself is simple. Each step has a checkpoint so you can confirm the line works before adding the next stage.
Dry Sand does nothing in a Shaker. Mix it with Water first — one Sand pixel plus one Water pixel becomes two Wet Sand — so Water also doubles your material before it ever reaches the machine.
Checkpoint: your pile darkens into Wet Sand where the water touches it.
One Shaker clears its own Residue painfully slowly. Run several as steps going down from the input: Wet Sand is the heavier pixel, so it sinks through the Residue layer and keeps the whole staircase self-clearing.
Checkpoint: each Shaker sits one level lower than the one feeding it.
Drop Wet Sand onto the highest Shaker. Every pixel always becomes 1 Residue, with a 1-in-4 chance of also dropping 1 Gold straight through. Expect roughly one Gold per four Sand.
Checkpoint: Gold falls out underneath while Residue creeps off the top edge.
The Residue spilling off the top is fuel, not trash. Burn it into Burnt Residue and drop it on a Kinetic Press; that turns waste into Gold and Seeds instead of a growing backlog.
Checkpoint: Residue keeps moving and does not keep blocking the staircase.
Start with the visible symptom. The smallest useful fix is usually one tile, one material path, or one blocked output, not a full rebuild.
| Symptom | Likely cause | Smallest useful fix |
|---|---|---|
| No Gold appears | The input is dry Sand, the Gold drop space is blocked, or the sample is too short for a 25% chance output. | Confirm Wet Sand reaches the top, clear the space below, then watch a longer flow before changing the layout. |
| Wet Sand sits on top | The line is overfed or the staircase does not slope away from the input. | Reduce input, add a lower Shaker step, and keep the top layer thin enough to move. |
| Residue buries the Shaker | The byproduct exit is blocked or too flat. | Open a downhill side path for Residue before feeding more Wet Sand. |
| Material is stuck inside the machine | A non-Gold solid entered the Shaker body. | Remove it manually, then filter or reroute mixed solids before they reach the Shaker. |
These are the three ways a Shaker line most often stops producing, and what to undo for each.
Only Gold passes through the Shaker body. If something else got dropped in, dig it out by hand and seal that entry — the machine keeps trapping material until you do.
A single Shaker cannot push Residue fast enough on its own. Add more staircase steps so density and gravity carry the layer downward instead of relying on the machine to move it.
Gold is one-in-four, so short stretches with no payout are expected. If the supply is still flowing, the machine is fine — keep feeding and the odds even out over a few hundred pixels.
The Shaker processes Wet Sand, drops Gold through the machine body, and pushes Residue out across the top. The line stays easier to control when the route keeps sloping away from the input instead of trapping loose material around the machine.
Verify how one Shaker behaves first, then scale only after terrain, supply stability, and Gold and Residue handling stay readable across the whole line.
A Shaker processes Wet Sand placed on top, always creates Residue, and has a 25% chance to produce Gold for each processed Wet Sand pixel.
The common causes are dry Sand input, blocked Gold drop space, trapped non-Gold solids, or Residue with no clear exit path.
Once the Shaker line runs, improve reliability before scaling size.
Use a tighter source-based layout when the basic staircase is understandable.
Route the guaranteed byproduct before it becomes the next production wall.
Tick off Water, Wet Sand, Gold, Residue and transport checks before expansion.
This guide is source-checked against official mechanics, but it is not labeled as a Sandustry Tips in-game test.
Source note: This recommendation is source-based, not an independently benchmarked build.
Official source