Grinders that dose by time deliver a different weight as conditions change.
Why time is unreliable
Hopper level. A full hopper feeds faster than a nearly empty one. The dose drifts down as the hopper empties.
Bean density. Light roasts are denser than dark. Switching coffees changes grams per second.
Grind setting. A finer setting passes less coffee per second. Every adjustment changes the dose.
Humidity and bean age, both of which change how beans feed.
Retention, which varies with static and with how much was left last time.
The fix
Weigh the beans before grinding — single dosing. Exact, and it removes every one of those variables.
Or weigh the grounds after, and adjust. Less elegant, equally effective.
Either way, a scale solves it.
Grind-by-weight grinders
Some grinders weigh the output continuously and stop at the target. Genuinely accurate, and expensive.
Improving a timed grinder
Keep the hopper at a consistent level. Re-check the timing after every grind adjustment. Use RDT to stabilise retention.
Those help. They do not match weighing.
The context
For filter, a gram either way is barely detectable. For espresso, a gram in an 18 g basket is a 5% ratio change and it is tasteable.
That is why espresso people weigh and filter people often do not.
Last reviewed 8 August 2026