Filling gear that pays for itself: trays, jigs, depositors and piston fillers

Pouring jigs, trays, heated depositors and piston fillers compared on units per hour, fill weight variation and cost, with the nozzle and jacket settings to use.

There are four rungs on the filling ladder and only one of them pays for itself on labour arithmetic. This page gives the throughput and fill weight variation of each, the settings that make a depositor work (jacket temperature, nozzle bore, suck-back), the flatness a tray has to hold, and the payback sums that tell you which rung you are actually on. Equipment prices are indicative UK trade, 2026, excluding VAT.

• Freehand: 60-90/hour
• Depositor: 600+/hour
• Jacket: pour plus 8-12 C
• Nozzle bore: 4-8 mm
• Tray flat to 0.5 mm

The four rungs, and what each one actually is

A jig holds containers upright and nothing else. A strip of stiff card with holes punched to the tube diameter, taped over a shallow box, is a jig, and so is a drilled hardwood block. You still pour each unit individually from a jug or a pipette.

A tray is a jig with a scrape plate: a block that holds a whole set of tubes at a fixed height with a flat plate across their rims, so you deliberately overfill the whole block, let the excess set on the plate, and scrape it off in one pass. The plate is a reference surface, which is why a tray produces level tops that freehand pouring cannot match. Technique is on filling lip balm tubes.

A depositor is a heated reservoir with a valve and a nozzle, operated by hand or foot. The melt is held at temperature in the hopper instead of cooling in a jug, and each pull of the lever delivers a set shot. Most craft-scale units are gravity fed with a water or oil jacket; the better ones are piston fed from a heated hopper.

A piston filler is a depositor with a metered piston, a positive shut-off nozzle and usually a foot switch or a small indexing conveyor. It is the first rung where fill weight stops depending on the operator.

The jump that matters is not jig to tray, which is a technique change, but tray to depositor, which is a temperature change. Everything below a depositor pours from a vessel that is cooling the whole time. That is the constraint the rest of this page keeps returning to.

Throughput, and why the fill is rarely the bottleneck

The numbers below are sustained rates for one operator filling a familiar product, not sprint rates over ten units. They include repositioning containers and topping up the dip, which is real work on tubes and tins alike.

Now put that beside the rest of the operation. Capping and wiping rims runs at roughly 160 units an hour by hand, and hand labelling a small tin at about 60. A depositor filling 700 units an hour into a line that labels 60 an hour has not made you faster, it has built a queue. The full time and cost breakdown of a batch, including the discovery that labelling takes longer than filling, is on costing and pricing.

The genuine argument for speed is not labour, it is temperature. A 250 unit batch filled freehand at 75 units an hour takes over three hours, and a jug loses roughly a degree a minute. Long before you finish, the melt has fallen out of the 65 to 72 C tube window or the 60 to 68 C tin window set out on pour temperatures, and the last third of the batch shows lumps, flow lines and unfilled bases. Reheating between trays is only a partial answer, since every reheat costs volatile top notes and antioxidant.

Fill weight variation, and what it costs

Hand filling lands units within about plus or minus 0.15 g of target. A metered piston holds plus or minus 0.05 g. On a 4.5 g lip tube those are 3.3% and 1.1% of the fill; on a 13 g tin, 1.2% and 0.4%.

Variation costs you in three places. The first is giveaway: a wide spread makes people target higher than they need to, and 0.3 g of unnecessary overfill on a 13 g tin at 6.9 p/g is about 2p a unit, or £200 across 10,000 units. The second is the short unit, which does not break any rule on its own but does look mean next to its neighbour on a market stall. The third is the overfilled tin whose lid smears when it is screwed down, which is a rejected unit rather than a light one.

There is a mechanism behind depositor drift that is worth understanding, because it links two settings that look unrelated. A volumetric depositor meters volume, not mass, and molten balm expands with temperature at roughly 0.0007 per degree, the usual figure for liquid triglycerides. Run the hopper 10 C hotter than the temperature you set the shot at and each shot carries about 0.7% less mass, which is 0.09 g on a 13 g fill. The jacket set point is therefore a fill weight control, not just a flow control. A piston filler with a fixed stroke has exactly the same sensitivity; only a checkweigher-fed feedback system removes it.

Trays: flatness to 0.5 mm is the whole specification

A scrape plate only works if it sits flush on every rim in the block. Hold it to within 0.5 mm across the plate. Beyond that, a gap opens between the plate and the tube rims, and molten balm at pour temperature has a viscosity in the tens of millipascal seconds, which is thin enough to run sideways through a 0.5 mm channel without hesitating.

What you get is a batch that has quietly redistributed itself: some tubes overfilled from their neighbours, some short, balm on the outside of the barrels where it will foul the twist mechanism, and a scrape that no longer produces a level top because the plate was never the top surface. It reads as a filling fault and is a flatness fault. The related leak paths in the tube itself are on lip balm tubes leaking.

Aluminium plates stay flat. Acrylic and HDPE creep above about 60 C and take a permanent set if they are left loaded while warm, and the usual killer is a dishwasher running at 65 to 70 C. Wash plates by hand, store them flat rather than leaning against a wall, and never stack anything on a warm one. Tube height also varies between suppliers by a millimetre or so, so test a tray against ten of your actual tubes before committing to a 500 tube order; container tolerances and formats are covered on the packaging guide.

An aluminium tray is also a heat sink. In a cool workshop, stand the loaded tray on a warming plate at 30 to 35 C before filling. Warming the container is always better than raising the pour temperature, because it removes the cold wall without adding superheat to the bulk.

Jacketed depositors: set the jacket 8 to 12 C above the pour

Melt loses heat on its way out: through the hopper wall, through the valve body, and most of all in the unheated last few centimetres of nozzle. Set the jacket 8 to 12 C above the temperature you want the melt to arrive at, then verify by catching a shot in a small beaker and reading it with an immersed probe. The hopper thermostat is not the number that matters; the nozzle exit is.

Use the low end of that offset for a short, insulated or heat-traced nozzle and the high end for a long unheated one. Then check the result against the ceiling. A tin pour at 64 C plus 10 C puts the hopper at 74 C, comfortably inside the 70 to 80 C melt window. A tube pour at 72 C plus 12 C puts it at 84 C, which is against the 85 C ceiling where beeswax darkens and tocopherol degrades.

Two operational habits go with a heated hopper. Discard the first two or three shots after any pause longer than about five minutes, because the slug sitting in the nozzle is colder than the hopper and will fill short and lumpy. And stir or recirculate the hopper if the formula contains pigment or a suspended solid, since anything denser than the melt settles towards the outlet over an hour, which is the mechanism behind pigment sinking to the bottom.

Nozzle bore: 4 to 8 mm, and the drip against stringing trade-off

Match the bore to the container, then to the melt. A 4 to 5 mm nozzle suits an oval lip tube, whose internal bore is only about 11 by 7 mm. A 6 mm nozzle suits round 15 mm tubes and small slider tins. An 8 mm nozzle suits 30 to 60 ml tins and jars, where the shot is large and the target is wide.

The trade-off runs in both directions from that range.

• Too small (below about 4 mm). Velocity rises, the stream necks down at shut-off, and the wax at the exposed tip is already cooling, so the shot leaves a thread behind it. That string lands on the tube rim or across the next unit and has to be picked off. High velocity also drives the stream into the container instead of laying it in, folding air into the fill, which is one of the causes covered on air bubbles in balm.
• Too large (above about 8 mm). Shut-off is clean but the residual column in the nozzle drips between shots, so you get spots on rims and on the tray plate. Shot definition suffers too, because the melt starts falling before the valve has fully opened and stops after it has closed.

Three fixes are worth having. Suck-back, sometimes called decompression, pulls the piston back half a millimetre to two millimetres at the end of the stroke and lifts the meniscus clear of the tip, which cures most dripping. A positive shut-off nozzle with a needle valve does the same job mechanically. And keeping the nozzle 3 to 5 mm above the rim means any thread that does form breaks against the fill rather than the container edge. If specks rather than threads are the problem, the cause is usually upstream in the melt, and that is on undissolved wax specks.

The cost ladder, and what actually pays back

Read that table honestly. The tray is the only rung that pays for itself on labour. Eight hundred units is four months at 200 a month. The depositor needs 22,000 units to repay its purchase price out of filling time alone, which at 200 units a month is nine years, and the piston filler needs a third of a million units, which no craft business will reach.

That is not an argument against buying a depositor. It is an argument against justifying one on labour. The real case is the temperature case from earlier: a depositor removes the ceiling on batch size, because the melt no longer cools while you work. If your batch is limited to what you can pour in 45 minutes, a depositor does not make you faster per unit so much as it makes a 500 unit batch possible at all, and the per-unit cost of a 500 unit batch is far below that of five 100 unit batches. The economics of batch size sit on batch scaling and the wider production picture on scaling up production. The rest of the workshop kit, from scales to melting vessels, is on equipment.

Checkweighing, and the average quantity rules

Weigh every tenth unit, plus the first and last of every batch, and write the figures in the batch record rather than in your head. Ten per cent sampling is enough to catch a drifting filler within a few units, and the record is what lets you defend a declared weight later. Recording practice is on batch records.

Two practical points ruin more checkweighing than any equipment fault. First, weigh net, not gross. Tin and tube tare weights vary by 0.2 to 0.4 g between units, which is four to eight times a piston filler's accuracy, so a gross weight tells you almost nothing. Either tare an empty container from the same delivery on the scale before each check, or weigh twenty empties and work from the mean tare. Second, use a scale with 0.01 g resolution for lip-sized units. A 0.1 g scale reading a 4.5 g fill is quantising at 2.2% of the fill, which is worse than the process you are trying to measure. Scale selection and calibration are on weighing and calibration.

In Great Britain, packages between 5 g and 25 kg sold by weight fall under the average quantity system in the Weights and Measures (Packaged Goods) Regulations 2006. Three packers' rules apply: the batch average must be at least the nominal quantity, no more than 2.5% of packages may fall below nominal by more than the tolerable negative error, and no package at all may fall below nominal by more than twice that error. For a nominal 5 to 50 g the tolerable negative error is 9% of nominal.

Note the last row. A 4.5 g lip balm sits below the 5 g floor, so it is outside the average quantity system entirely and cannot carry the "e" mark. That is not permission to underfill: a declared weight has to be true, and cosmetic labelling requires the nominal content at the time of packaging under Article 19 of the UK and EU Cosmetics Regulations, with an exemption for packages under 5 g or 5 ml and for free samples. In the United States the equivalent duty is the net quantity of contents declaration under 21 CFR 701.13. Which regime applies to you depends on where you place the product on the market, and this site cannot make that determination for you. Market-by-market requirements start at labelling cosmetics in the UK.

Cleaning down a filler

Wax and water are the two things that ruin filling equipment, in that order. Clean hot, in sequence, and never start with water.

1. Drain hot. Return what is left in the hopper to the melt pot while everything is still at working temperature.
2. Wax flush. Run 200 to 500 ml of the cheapest compatible liquid oil through the hopper, valve and nozzle at working temperature, and keep going until it runs clear. This is the step that matters: it dissolves and carries out the wax that would otherwise set solid in the valve seat.
3. Hot oil rinse. A second, clean charge of the same oil at temperature, to displace the pigment and wax loaded first flush. Repeat for tinted products until the oil comes out uncoloured, because iron oxides cling to a piston bore and will streak the next batch.
4. Drain and wipe. Wipe accessible surfaces while they are still warm, when the film is a liquid rather than a scale.
5. Isopropyl finish. Once cool, wipe down and swab through the nozzle with isopropyl alcohol on a lint-free cloth to remove the oil film. Check the seal material first: nitrile and PTFE tolerate isopropyl, some elastomers swell in it or in hot oils. Material compatibility is on packaging compatibility.
6. Dry completely. Leave the parts open until every trace of alcohol has evaporated before the next batch.

The last step is the one people rush, and it matters more in anhydrous work than anywhere else. Residual moisture in a filler is water introduced into a product that has no preservative and no water activity to spare, which is the argument set out on moisture control. Cold water anywhere in this sequence sets the wax instantly and blocks the valve, and a dishwasher will warp a plastic scrape plate and dull a piston bore. General hygiene practice for an unpreserved product is on workshop hygiene.

What to buy, and when

The decision rule is a batch size, not a revenue figure. Work out how many units you can fill before the melt leaves its pour window, which for most people is 45 minutes to an hour from the jug. If your target batch fits inside that, a tray is all you need and a depositor will sit idle looking expensive. If your target batch does not fit, no amount of technique will fix it, and a heated hopper is the answer rather than a faster pair of hands.

In practice that puts the crossover somewhere between 200 and 300 units a batch for tubes, a little higher for tins because the pour is quicker per unit. Below it, buy a tray, a warming plate and a decent scale. Above it, buy a jacketed depositor and expect capping and labelling to become the new bottleneck within a month. The piston filler is a different kind of purchase altogether: it buys fill weight consistency and freedom from operator fatigue, and it only makes sense alongside a conveyor and a second person.

Two honest limits. Equipment prices here are 2026 UK indicative bands and vary widely by supplier, condition and whether a unit is food grade; get quotations before planning around them. And the fill spread figures are working bands from checkweighing craft-scale equipment, not manufacturer specifications, so measure your own before you rely on a number. The batch-to-batch variation that filling equipment does and does not solve is set out on batch to batch inconsistency, and the technique that gets the most out of whatever gear you own is on filling lip balm tubes.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.