Hopper suction problems

The pellet feeding system (vacuum + hopper + corrugated tube) is the single most common source of customer issues. Problems split into three flavors: weak suction, machine shutdown during the vacuum cycle, and crunching sounds from the hopper.

Symptoms

• Pellets don’t rise from the material source to the hopper, or rise too slowly
• The vacuum is audibly running but no material is transported
• The machine shuts down (Klipper shutdown) during the vacuum cycle
• The machine beeps repeatedly — 3 seconds on, 17 seconds off. That is the feeding alarm: four rounds of suction have run without material reaching the probe
• A “crunching” / scraping sound comes from the hopper when pellets are inside
• Hopper level slowly drops and the vacuum can’t keep up → under-extrusion

Diagnostic flow

Step 0 — If you feed from a D1, check the guillotine is open

The guillotine above the pellet gate is kept closed while drying and whenever the printer is off. Left closed, it produces exactly the symptoms of a feeding failure: the vacuum runs its cycles, no material arrives, and the buzzer sounds. Check it before dismantling anything — see Pellet gate and guillotine.

Step 1 — Check the seal between the corrugated tube and the hopper

This is the #1 cause — air leaking at the joint means the vacuum can’t build enough suction.

1. Power off the machine and the vacuum
2. Inspect the joint where the corrugated tube clamps onto the hopper (the upper end)
3. If the clamp is loose or the rubber gasket is missing/worn → tighten or replace
4. A snug gasket and a tight clamp fix most suction complaints. Check this before anything else

Step 2 — Check the line is built correctly

The feed line is three sections, not one tube — see Corrugated tube:

1. White corrugated from the material source
2. Smooth rigid grey pipe for the long static run
3. White corrugated to the extruder, so it can flex with the gantry

A machine with both corrugated tube and rigid pipe is correct — that is the design, not a bodge.

• Standard tube diameter: 32 mm internal
• Place the material source close to the machine

The routing rule

Pellets must never be able to travel downward by gravity anywhere along the line. Wherever they can, they pile up until the tube plugs. Keep the whole run rising toward the hopper — no dips, no descending sections. This matters more than the total length.

Step 3 — Check the material, but only for one thing

Ordinary pellet variation — uniform, flat, small, irregular — is not a feeding problem. Do not chase pellet shape.

The exception is recycled material containing strings, threads or microplastic: that clogs the filter. If you are running that kind of material, clean the filter and sieve the pellets before loading.

Step 4 — Check the socket and the vacuum’s voltage rating

The vacuum belongs on the FEEDING socket — the middle one of the three on the power box. That is the socket the printer switches on and off; a vacuum plugged anywhere else will never be controlled by the machine. See Power outlets.

All three sockets on the power box are 220V

A 110 V vacuum plugged into any of them burns out. This catches customers in 110 V regions, who receive a 110 V-rated vacuum: it must be fed from a 110 V supply, or the machine’s own supply must be stepped down. Do not plug it into the power box.

Vacuum triggering too frequently (every ~1 minute)

Normal is one cycle every 10–20 minutes, depending on how much material the print is consuming. A high-flow print cycles at the fast end of that range, a slow one at the slow end.

Cycling far more often than that — every minute or two — means the hopper is emptying faster than the vacuum refills it: each cycle is delivering less material than the extruder consumes. Most often a suction leak.

Check in this order:

1. The filter is installed between E17 and E04. Without it the vacuum pulls the pellets straight up the VACUUM line instead of leaving them in the hopper
2. The black rubber gasket is installed right-side up (not inverted) — an inverted gasket allows air bypass and cuts suction
3. The line is built as described above, rising all the way with no dips

If it still doesn’t work

Case A — Machine shuts down during vacuum cycle

Symptom: machine cuts off (display black, fans still on) while sucking pellets.

This is a mains capacity problem, not a feeding problem. Add up what the machine draws when everything is heating at once:

Load	Power	Current at 230 V
Bed heater	~2000 W	~8.7 A
Three heater bands (3 × 300 W)	~900 W	~3.9 A
Motors, electronics, fans	~100 W	~0.4 A
Printer subtotal	~3 kW	~13 A
Vacuum	~1 kW	~4.3 A
Peak total	~4 kW	~17 A

A standard 16 A circuit carries 3.68 kW. The peak is above that — so on a single circuit the machine is already at the limit before the vacuum starts, and the vacuum’s startup inrush (a universal motor briefly draws several times its running current) is what pushes the voltage down far enough for the Raspberry Pi to see undervoltage and Klipper to shut down.

That is why the power box is fed by two main switches which can go to two separate mains lines — one for the printer and vacuum, one for the bed. See Power supply 24V.

Fixes, in order of effectiveness:

• Split the load across two circuits using the two mains inputs — this addresses the cause
• Move the machine to a dedicated circuit with no other loads on it
• Install a UPS if the mains supply itself is unstable
• If it persists → see Klipper MCU communication lost
• Temporary workaround: turn off the vacuum and refill the hopper by hand

Case B — “Crunching” sound from the hopper

Symptom: a scraping / crunching noise when pellets are inside the hopper.

• Normal with non-uniform or recycled pellets — the system is just compressing pellets against the walls
• Abnormal if the noise comes from the gearbox or motor (you’ll feel vibration there)

Test:

1. Print briefly without material (empty extrude) — if the noise disappears, it’s pellets being compressed → keep printing
2. If the noise persists with no pellets → mechanical problem, see Belt and pulley mechanical issues

Case C — Vacuum doesn’t run at all

Verify in this order:

1. The vacuum’s own switch is ON. This is the usual cause. The vacuum has to be left switched on at the appliance — the printer then controls it by switching the socket it is plugged into. A vacuum switched off at its own button will never run, no matter what the printer commands
2. Vacuum plugged into the FEEDING socket (middle) on the power box — see Power outlets
3. FEEDER toggle state on the display — see below
4. The SSR FEED screws on connector [2] of the power box board. Two of its four screws carry the feeding relay’s signal, and one that has not bitten on its wire leaves the Feeding socket permanently dead with nothing else looking wrong — see the two power box connectors
5. If it still does nothing, the vacuum is faulty → replace

The FEEDER toggle is off after power-on, and that is normal

On a freshly powered printer the FEEDER toggle is off. It is normally left off: the START_PRINT macro switches it on when a print begins. So “the toggle is off” is not a fault on an idle machine — only worth investigating if it stays off once a print is running.

Documented workarounds

Hopper extender for manual refill

A customer-made part that replaces E17 so the hopper stays open and can be filled by hand. There is no official geometry — the customer models their own. Useful while the vacuum feeding system is being diagnosed, but it means feeding the machine manually for the whole print. See Hopper extender.

Newer hopper design (post-Apr 2026)

A redesigned hopper that mitigates feeding issues with difficult pellets (recycled, flat, fines-heavy). Rolled out to customers with recurring vacuum problems.

How to prevent

• Route the line so it always rises toward the hopper, with no dips or descending sections
• Sieve recycled material that contains strings or microplastic before loading
• Keep the filter and gasket in place whenever the hopper goes back together

Cleaning the feeding filters is part of routine machine maintenance — see Hopper.

Notes

• This is the most frequently reported problem on the G1
• Common combination: a badly routed line and a loose clamp
• The corrugated-tube-to-hopper joint is repeatedly identified as the leak point

Related notes

• Power outlets (power box)
• SmoothFlow
• Klipper MCU communication lost
• Belt and pulley mechanical issues
• Hopper
• Vacuum pump