Skip to content

Why a 12.7 mm TIJ Cartridge Prints Clean on a Carton but Breaks Up on a Bottle

B

Brent

· 11 min read

Why a 12.7 mm TIJ Cartridge Prints Clean on a Carton but Breaks Up on a Bottle

Table of Contents

A plant proves a cartridge on flat outer cartons, the codes are crisp, and the purchase is confirmed. Three weeks later the same cartridge goes onto a round bottle or a shrink-wrapped component, and the code comes out uneven: solid in the middle, thin and broken at one end, sometimes legible and sometimes not.

The instinct is to blame the ink, or the cartridge, or the supplier. The cartridge is usually the same one that worked last week. What changed is the shape of the surface underneath it.

The short answer: a printhead fires droplets across a flat band, and a flat band only meets a curved surface along a line. As the surface curves away, the outside edges of the band sit further from the nozzle face than the centre does, so the dots at the edges land late, land wide, and drift sideways. The effect grows with the square of the band width and shrinks as the part gets larger. That is why the same 12.7 mm TIJ solvent ink cartridge is fine on a box and marginal on a bottle, and why the fix is usually geometry and technique before it is chemistry.

A 12.7 mm Band on a Curve Is Not 12.7 mm of Contact

The number 12.7 mm describes the height of the nozzle row, and nothing else. On a flat carton, the whole row sits at one distance from the print surface, so every nozzle fires across the same gap.

Put the same row against a cylinder and the geometry changes. The nozzle face is flat and the surface is round, so they touch along a single line at best. Move away from that line in either direction and the surface falls away from the head. At the two edges of the band, the gap is larger than it is in the middle.

This is why the failure is so often described as “the code is fine but the edges are weak.” The edges are not weaker because the nozzles are worse. They are weaker because those nozzles are printing from further away, at a shallower angle to the surface.

There is a second consequence that gets missed. Because the head is flat and the part is round, the code is not simply printed with a variable gap. It is also slightly compressed in the direction of curvature, which shows up as characters that look crowded at the ends rather than evenly spaced.

The Gap at the Edge of the Band Is Geometry, Not Ink

The extra gap can be estimated before anyone prints anything. For a band of width w wrapped around a part of diameter D, the surface falls away from a tangent nozzle face by roughly:

edge gap ≈ w² ÷ (4 × D)

This is the sagitta of a circular arc, not a measurement of any particular printer. It assumes a circular cross-section, a nozzle face tangent at the centre of the band, and no tilt in the hand or fixture. Real parts are rarely perfect circles, so treat the output as an order of magnitude rather than a specification.

Part diameterEdge gap, 12.7 mm bandEdge gap, 25.4 mm band
30 mmabout 1.3 mmabout 5.4 mm
60 mmabout 0.7 mmabout 2.7 mm
100 mmabout 0.4 mmabout 1.6 mm
200 mmabout 0.2 mmabout 0.8 mm

Two things fall out of that table immediately. The first is that the penalty is small on a large-diameter part and serious on a small one. The second is that the penalty quadruples when the band doubles, which is why a full-inch head can be excellent on cartons and awkward on a 30 mm cap.

What the Dots Actually Do Across an Arc

A printer is designed around a nominal standoff, the distance at which droplets land where the controller expects. A curved surface violates that assumption progressively across the band, and the errors are not symmetrical.

Droplets that travel further arrive slightly later than intended. On a stationary part that matters little, but on a moving part the timing error becomes a placement error, which is why the same code can look acceptable on a stationary cap and drift on a moving one.

Droplets that travel further also spread more, so dot diameter grows at the edges. Where dots overlap in the centre, they may stop overlapping at the ends, and a solid character becomes a thin, broken one.

The angle matters too. At the edges of the band the surface faces partly away from the nozzle row, so the droplet approaches at an angle rather than head-on. That shows up as edges that are ragged or feathered instead of crisp, and it is often misread as a nozzle problem.

Three Failure Signatures That Get Blamed on the Ink

Coding failures on curved work are usually reported as “the ink does not stick” or “the cartridge is bad.” Three distinct signatures account for most of them, and only one is a chemistry problem.

What you seeWhat it usually meansWhat to change first
Crisp in the middle, thin or missing at one or both endsEdge gap across the arcOrientation, band width, or fixture support
Ragged or feathered edges, centre acceptableShallow approach angle at the band edgesRotate the head so the row follows the curve, not across it
Characters crowded at the ends of the codeSurface curvature compressing the print directionMove the code to a flatter face, or reduce content
Fine on the first part, degrading through a shiftOperator-dependent distance and angleA rest, a guide, or a stop rather than more training
Fails only on chilled or damp partsCondensation between nozzle and surfaceWipe or dry the area, and check the material

The table is worth reading as a diagnostic order rather than a list. Position and geometry first, because they are free to change. Chemistry last, because it is the most expensive thing to change and the least often at fault.

Holding the Coder: Consistency Beats Technique

In portable TIJ printing, the operator is part of the machine. A handheld coder does not hold its own standoff, so the distance and angle are re-decided on every part, dozens or hundreds of times per shift.

The useful goal is not a better technique. It is a setup in which technique matters less. Three practical moves do most of the work.

Support the hand. Resting the coder or the wrist on a guide keeps the standoff from drifting, and it removes fatigue as a variable. The design of the grip matters here for the same reason: a coder that has to be held forcefully is a coder whose distance changes as the shift goes on, which is covered in the notes on handheld inkjet handle design and operator fatigue.

Fix the part, not the hand. If the product can sit in a simple nest or against a stop, the operator’s job becomes placing the part rather than aiming the head. This is usually a cheaper improvement than any cartridge change.

Slow the pass down on small diameters. On a 30 mm cap, the arc is short and the angle changes fast, so a fast stroke covers a large angular change in a fraction of a second. A slower, steadier pass gives the electronics a more consistent geometry to work with.

When the code comes out blurred rather than broken, the cause is more often parameters and motion than the surface, and the parameter and motion table for handheld inkjets is the better starting point than this article.

Why a Narrower Band Wins on Curved Work

The width decision is usually framed as a cost question: a wider head uses more ink, and the two articles on half-inch versus full-inch cartridges and the cost of the wrong cartridge width cover that argument properly. Those articles are about ink economics. This section is about curvature, and the conclusion points the same way for a different reason.

Because the edge gap scales with the square of the band width, a narrower band is disproportionately better on curved work. Halving the width quarters the edge gap. On a small-diameter part, that can be the difference between a code that reads and one that does not, regardless of how good the ink is.

This is a real constraint on the common advice to buy the widest head you might need. If the parts are round and small, a 12.7 mm head printing two stacked passes can outperform a single 25.4 mm pass, because each pass wraps a shorter arc. The cost advantage and the geometry advantage happen to align here, which is not always the case.

The Ink Still Has to Suit the Surface

Geometry explains break-up. It does not explain a code that prints cleanly and then rubs off, or one that never adheres at all. Those are chemistry and surface questions, and they have their own answers.

Most round packaging is non-porous, which is why solvent formulations dominate the category. If you are choosing a solvent ink cartridge for plastic packaging, the decision is about the specific polymer, its surface treatment, and whether the surface is clean, rather than about the word “solvent” on the box. The water-based versus solvent comparison sets out which substrates need which chemistry, and why ink fails to adhere to plastic film deals with the case where the code prints and then comes off.

One confounder deserves its own warning on bottles. A chilled PET bottle carries condensation, and condensation sits between the nozzle and the surface exactly where the arc gap is already working against you. A code that fails only on cold product is usually a moisture problem before it is a distance problem, which is what the cold PET condensation case deals with.

If you need to test TIJ solvent ink on non-porous surfaces, test the arc and the moisture together, because in production they occur together.

Test Across the Arc, Not at One Spot

The usual test procedure asks for representative samples and a recorded result. On curved work the sampling method matters more than the record, because a single print in the wrong place will pass while the production job fails.

Print at the position the code will actually occupy, on the face that will actually be coded. A cap tested on its flat top tells you nothing about the same cap tested on its side.

Print several parts in a row rather than one, because the operator-dependent variable is the one you are trying to measure. Consistency between parts is the result you care about, not the best single print.

Examine the ends of the code deliberately. The centre of the band is the part most likely to look acceptable, so a quick glance at the middle will confirm a pass that the edges do not support.

Then turn the part. If the code only reads from one orientation, or only on one side of the cap, you have found a geometry limit rather than a random defect, and no amount of re-testing will remove it.

An Illustrative Case

The following is an illustrative scenario assembled from a common pattern, not a report of measured results on a specific site.

A cosmetics filler needs an expiry date on a 35 mm cap. The line already runs a 12.7 mm solvent cartridge on flat cartons with no trouble, so the same setup is used on the caps by hand, with the code placed across the diameter.

The first run is inconsistent. Roughly a third of the caps read acceptably, a third are weak at one end, and a third are rejected. The cartridge is swapped, the ink batch is changed, and the results do not improve, because neither was the variable.

Rotating the cap so the code runs along the side rather than across the crown changes the geometry instead of the consumable. The band now wraps a longer radius in the print direction and a shorter one across it, and the weak end largely disappears.

What fixed the job was a decision about where the code sits on the part, made before any cartridge decision. That is the ordinary outcome on curved work: the geometry is the design variable, and the consumable gets blamed for it.

Conclusion

A cartridge that works on a carton and fails on a bottle is not a contradiction. A flat band meets a curved surface along a line, and everything away from that line prints from a worse position. The edge gap grows with the square of the band width and shrinks as the part gets larger, which makes the diameter of the component and the width of the head into design decisions rather than catalogue choices.

The practical order is short. Check where the code sits on the part. Check whether a narrower band or a rotated code would wrap a shorter arc. Support the hand or fix the part so standoff stops varying. Only then look at the ink, and look at it with a test that prints across the arc rather than at one convenient spot.

At FirstColor we supply 12.7 mm and 25.4 mm TIJ cartridges in solvent and aqueous formulations, and we are asked about curved-work failures often enough to recognise the pattern quickly. Send us the part diameter, the material, the code content and a photo of the code as it actually comes out, and we will tell you plainly whether the limit is geometry, surface or ink. You can reach us here.

FAQ

Why does my code print fine on cartons but break up on bottles?

Because a flat printhead band meets a curved surface along a line rather than across its full height. The edges of the band sit further from the surface than the centre, so the outer dots travel further and land wider and at an angle. Narrow parts make this worse, and the effect grows with the square of the band width.

Does a narrower cartridge really help on curved surfaces?

It helps more than proportionally. The edge gap is roughly proportional to the square of the band width, so halving the width reduces the gap to about a quarter. On small-diameter parts this can matter more than any difference in ink formulation.

Should I rotate the code to follow the curve?

Often yes. A code that wraps around the part is usually more consistent than one printed across the crown, because the band follows a longer radius in the print direction. The trade-off is where the code ends up on the finished package, so it is a design decision as much as a printing one.

Is TIJ cartridge compatibility with handheld printers a datasheet question?

Partly, and less than most people expect. The datasheet tells you whether the cartridge fits and is recognised. It cannot tell you whether an operator can hold a consistent standoff on a 30 mm cap at line pace, which is what determines the outcome on curved work.

Can I fix this with a different ink?

Only if the failure is adhesion rather than break-up. If the edges are thin or missing while the centre is solid, the cause is distance and angle, and changing ink will not close a gap. If the code prints cleanly and then rubs off, that is chemistry and surface, and a different formulation may well be the answer.