TIJ Cartridges for Production Lines: How Many Models Does Your Plant Actually Need?

B

Brent

· 12 min read

TIJ Cartridges for Production Lines: How Many Models Does Your Plant Actually Need?

Table of Contents

Walk into the consumables store of a plant that runs four or five coding machines and count the cartridge part numbers on the shelf. Most sites we visit are holding five, six, or seven. Ask why that particular set, and the answer is rarely a decision. It is an accumulation. A machine arrived with its own cartridge, a material changed years ago and a second chemistry was added, someone found a cheaper equivalent for one line and then a third supplier was qualified for a second. Nobody ever stood back and asked how many part numbers the plant actually needs.

That is a different question from the one most buyers are trained to ask. Selecting industrial TIJ ink cartridges for a single product is a product-level decision. Deciding which handful of part numbers your whole site will carry is a plant-level decision, and it has its own logic, because every part number on the shelf is a claim on production line uptime and not only on storage space.

The short answer: the number of TIJ cartridge models you stock should be derived from the printers you run, not from a supplier catalogue, and the right number is almost always smaller than what accumulates by accident — but larger than one. Two variables decide whether two machines can genuinely share a cartridge: print width, and the electrical interface generation the printer expects. Ink chemistry is not one of them, because it follows the substrate rather than the machine, and that is the one thing you must never converge just to free up a shelf slot.

The Question Changes When You Own More Than One Coder

A single handheld coder and a single product material is the simplest possible case, and it is the case most buying guides are written for. You match one cartridge to one substrate, order it, and move on. The decision is complete the moment the sample prints cleanly.

Add a second machine and the decision stops being complete. The cartridge you choose for line two does not only have to work on line two. It has to work alongside whatever is already on the shelf, with the operators who already know those part numbers, and with the storage and reordering routine that already exists. A cartridge that performs perfectly on its own line can still be the wrong choice for the plant.

This is why the buyer’s checklist and the plant-level question are not competitors. Our five procurement indicators for a single TIJ cartridge settles whether one cartridge suits one job. This article starts after that question is answered, and asks how many of those answers you want to end up maintaining. One is about the part. The other is about the set.

The commercial argument for a smaller set is not aesthetic tidiness. Every extra part number has to be forecast, ordered, stored, insured against running out, and taught to every operator on every shift. Those costs are real and they scale with the number of models, not with the volume you buy.

Audit the Printers Before You Look at Any Cartridge List

Convergence conversations fail when they start from the cartridge catalogue. They work when they start from an asset list. Before proposing any part number, record what is actually installed.

For each coding machine on site, capture:

  • Make and full model number, not the nickname used on the shop floor
  • Print width the machine is configured for
  • Whether it is handheld, inline, or an integral part of a packaging machine
  • The trigger method it uses in daily production
  • Which part number is loaded in it right now
  • Which materials that machine codes, and roughly what share of its running time each material takes

The last two lines are the ones people skip, and they are the ones that decide the outcome. A machine that spends ninety percent of its life on corrugated carton and ten percent on a coated film is not the same purchasing problem as one that splits its time evenly. The first can carry a single model with a documented exception; the second usually cannot.

Two further details belong in the audit because they are invisible until they cause a problem. The first is who sets parameters on each machine, since a machine that several shifts adjust independently will drift away from whatever settings the trial produced. The second is whether any of the machines share a printed template or a quality specification, because two lines producing the same customer-facing code have a reason to be aligned that two unrelated lines do not.

If the fleet includes machines of different ages or from different equipment generations, the audit is also the right moment to think about how the plant will be equipped over the next few years. Our five-step framework for choosing a coding machine walks the sequence from the code you need printed through to the equipment decision. Read the audit as the input to that sequence, not as a separate exercise.

Two Variables Decide Whether Two Machines Can Share a Cartridge

Once the asset list exists, the natural next move is to look for pairs of machines that could run the same part number. Two variables decide whether that is possible. Everything else is a consequence of them.

The first is print width. A cartridge is built to a fixed band — the common formats on TIJ platforms are a half-inch band and a one-inch band — and a printer configured for one of those bands is not going to accept the other. If two machines are configured for different widths, they cannot share a cartridge no matter how similar their work looks. This is also the variable most often confused with capability: a wider band is not automatically better, and we have covered what actually changes when you move between half-inch and full-inch cartridges and how a width mismatch quietly costs money over a year in separate articles, so this one treats width purely as a constraint on sharing.

The second variable is the electrical interface and the generation of the cartridge platform the printer expects. This is where convergence plans most often fail after looking fine on paper. A cartridge can be physically identical in shape, seat correctly, and still not be recognised reliably by a particular printer, because the contact arrangement and the cartridge configuration the machine is looking for sit a generation apart. Our comparison of the TIJ 2.0 and 2.5 cartridge generations covers the shell, electrical, and firmware layers separately, and it is the right reference before you assume two machines from different vintages are interchangeable. Treat industrial coding cartridge compatibility as a two-layer question — what the shell does and what the electronics do — because a plan that only checks the shell will look correct right up to the moment the printer declines to recognise the cartridge.

What to compareIf the two machines agreeIf they differ
Print widthA shared part number can be trialledDifferent cartridge formats, no sharing
Interface generation and configurationThe same configuration can be specified for bothThe cartridge may fit and still not be recognised reliably
Ink chemistry the substrate requiresOne part number can cover both machinesTwo part numbers, even on identical hardware
Duty cycle and fill sizeOne fill size is workable on bothDifferent fill sizes, or a shorter change interval on one
Operating environmentShared cleaning and storage routine holdsSeparate handling rules, and a reason to keep the models apart

The useful discipline is to walk the table in order and stop at the first row that disagrees. Convergence is only worth discussing for a pair that survives all five.

What Convergence Buys You, and What It Costs

A smaller model set pays for itself in four places, and the honest case for it is made by naming all four rather than by promising a saving.

The first is forecasting. Fewer part numbers means fewer demand signals to estimate, and estimates based on more consumption per line item are less sensitive to one noisy month. The second is the error rate at the point of use, which is a bigger number than most sites realise: with seven part numbers on a shelf, an operator reaching for the wrong one is a question of when, not whether. The third is the cost of carrying the set, since every model held in stock ties up cash whether or not it moves this quarter, and we have walked the general logic of that trade-off in our spare parts stocking strategy article — the arithmetic there is written for repair parts but the reasoning transfers unchanged. The fourth is commercial weight, because volume concentrated into three part numbers is a better conversation than the same volume spread across seven, and our guide to reading a cartridge quotation explains what to compare once you get there.

The costs are just as real, and they are the reason convergence should never be pushed to the extreme.

Concentrating onto fewer models concentrates risk. If a plant runs its whole coding operation on one part number held from one supplier, a batch problem or a shortage stops every line at once. That is the argument for keeping a second qualified model for the highest-consequence work, and it is also why consolidation and single-sourcing are separate decisions that should not be taken in the same meeting.

Re-qualifying costs time. A cartridge that is already proven on material A is not proven on material B just because a second machine has the same print width. Every material the new model will touch needs its own trial, run on production equipment at production speed, with the result recorded.

And convergence has an operational cost that never appears in a purchasing document: retraining. Operators who have loaded the same part number for three years have a working knowledge of how it behaves. Changing the shelf changes what they know, and the change has to be managed deliberately rather than announced.

The Red Line: Ink Chemistry Stays Per Substrate

There is one thing on the list that must not be converged for the sake of tidiness, and it is the easiest mistake to make because the motive is so reasonable: fewer models on the shelf.

Ink chemistry is chosen by the surface the code lands on, not by the machine that fires it. Aqueous and solvent formulations are built for different behaviour on the substrate, and moving a machine onto a chemistry that suits a neighbouring line rather than the material in front of it trades a shelf slot for a printing problem. The specific warning to hold onto is the one about hardware rather than print quality: mixing chemistries on the same print engine risks cross-contamination and inconsistent adhesion, which is why our cartridge type comparison for coding machines makes the point that you can run different chemistries at different stations but should standardise within a station.

The practical rule that follows is narrow and easy to apply. Convergence happens across machines that code the same material family. It does not happen across materials. Two handheld units coding the same corrugated carton are a genuine convergence candidate. A handheld unit coding cartons and an inline machine coding a coated film are not, even if the cartridge format is identical, and the fact that they take the same shell is exactly what makes the mistake tempting.

If the materials on your lines are the part of this you are least sure about, the substrate question is worth settling before the model question, not after. Our comparison of water-based and solvent cartridges is the starting point for that decision rather than this article.

Build the Model Matrix, Then Keep It Honest

The output of all this is a single page: one row per printer, one column per deciding variable, and one column naming the cartridge model that row will run. It is not a sophisticated document. Its value comes from being written down, agreed, and revisited.

Ownership matters more than format. A matrix that lives in the purchasing department drifts, because purchasing sees the orders and not the shop floor. A matrix that lives only in engineering drifts the other way, because engineering sees the machines and not the reorder history. The version that survives has both names on it, and a stated review trigger rather than a review date.

Two triggers are worth writing into it from the start. The first is a new material, because a new substrate can invalidate the ink chemistry decision on a row without changing anything about the machine. The second is a batch of repeat complaints that trace back to one line, because in our experience that is more often a sign that the row no longer describes what is actually being run than a sign that the cartridge changed.

Keeping the matrix current also means keeping the routine that supports it current. Codes that drift, codes that smudge, and codes that scan badly are usually caught by operators long before they reach a purchasing conversation, and the parameter and motion reference for handheld inkjets is a reminder that a print problem and a cartridge problem are not the same diagnosis. Our guide to TIJ printer maintenance sets out the daily and periodic routines that keep a fleet in a state where the matrix remains a fair description of it.

The Breakdown Question: What Substitutes for What

Every convergence plan eventually meets the same question at an inconvenient hour. A line has stopped, the model it needs is not on the shelf, and someone wants to load whatever is available. The answer needs to exist in writing before that moment, not be improvised during it.

There are three honest positions a row can carry, and it is worth labelling every row with one of them.

The first is substitutable without qualification, meaning a second model is already proven on the same material at the same speed, and loading it is a documented routine rather than a decision. The second is substitutable for a defined period with a known consequence, typically because the alternative gives a shorter run time or a wider band than the job needs, and someone has decided the consequence is acceptable to keep the line moving. The third is not substitutable, and the correct action is to stop and wait rather than improvise, usually because the chemistry is wrong for the material and the resulting codes will not survive the downstream process or will not scan.

The reason to write these labels down is that the person facing the stopped line is rarely the person who ran the trial. A plant that has not decided in advance is not making a risk-based choice during the breakdown; it is transferring the decision to whoever is closest to the machine at three in the morning, with no information.

A Worked Example: Five Coders, Seven Models, Two Retired

The following is an illustrative scenario built from a pattern we see repeatedly, not a report of measured results on a specific site.

A contract packer runs five coding machines: three handheld units and two inline stations. The store holds seven cartridge models. Nobody can explain the set, but each of the seven has a story.

The audit produces a surprise on the first pass. Two of the three handheld units are on different models even though they are the same machine from the same purchase, and the reason turns out to be historical: one was commissioned a year later and was supplied with whatever was current at the time. Both are now coding the same corrugated carton at the same band width. That pair is a genuine convergence candidate, and it retires one model with nothing more than a trial on carton stock.

The second candidate is less obvious and more valuable. An inline station codes a coated film, and a third handheld unit is used mostly for rework on the same film. They take the same band width. They need the same chemistry, because the material is the same. The only obstacle is that the handheld has been running on the inline machine’s settings copied across, which nobody had checked. That pair retires a second model, but only after a trial on the film at both speeds.

The remaining three models survive the exercise, and the reason each one survives is worth recording. One covers a second material family that appears on two lines. One is held as the qualified alternative for the highest-consequence work, which is a deliberate redundancy rather than an oversight. The third exists because a single machine is a generation behind its neighbours and cannot be trusted to recognise the configuration the others use.

MachineBandInterfaceMain materialsOutcome
Handheld 1Half-inchCurrent generationCorrugated cartonConverged with handheld 2
Handheld 2Half-inchCurrent generationCorrugated cartonConverged with handheld 1
Handheld 3Half-inchCurrent generationCoated film, reworkConverged with the inline station
Inline 1Half-inchCurrent generationCoated filmConverged with handheld 3
Inline 2Half-inchOne generation behindCorrugated cartonKept separate, cannot be converged

Five machines end up on five models instead of seven, and the two retired part numbers were never held for a technical reason. They existed because nobody had compared the rows.

Adding a Printer Without Adding a Model

The matrix is only tested when the fleet grows, and the default outcome of buying a machine is that it arrives with its own cartridge and the model count goes up by one. That default is worth interrupting.

When a new machine is specified, the first question is not which cartridge it takes. It is which existing row it can join. If it matches an existing row on band width, interface generation, and material, then the correct result of the purchase is that the model count stays exactly where it was, and the specification should say so before the order is placed. The second question is whether the new machine’s arrival retires anything, which happens more often than expected when an older unit is being replaced rather than added to.

Handheld units deserve a specific note here, because they are the machines most likely to be bought without a specification process at all. A handheld coder is often purchased as a spare or a stopgap, which means nobody asks what it will run. Loading it with an existing model costs a trial. Not asking costs a new part number on the shelf for the next five years, and a fleet that grows by drift rather than by decision.

Conclusion

Converging a cartridge set is not about buying less. It is about knowing what you are buying and why, so that the number of models on the shelf is a decision with a reason behind it rather than a residue of decisions nobody remembers making. Two variables do the work: band width, and the interface generation the printer expects. One variable must be left alone, and that is ink chemistry, because it belongs to the substrate and not to the machine.

The practical sequence is short enough to keep in your head. Audit the printers. Compare the rows. Converge only on the variables that agree, and never across materials. Write down what may substitute for what before you need to know. Then review the matrix when a material changes, not on a calendar.

At FirstColor we supply TIJ cartridges and supporting consumables to equipment builders, integrators, and factories running mixed coding fleets — industrial coding solutions that range from a single handheld unit on a packing bench to five or six machines spread across several lines. The model-set conversation is one we have often enough that it is worth having properly. Contact us here with your printer list and the materials each machine codes, and we will tell you plainly which rows can be converged and which should be left where they are.

FAQ

How many cartridge models should a plant with several coding machines carry?

There is no universal number, but the right count is derived from the printers rather than from the catalogue. Audit every machine for band width, interface generation, and the materials it codes; convergent rows usually collapse to fewer models than the site is currently holding, while genuinely different materials keep their own. A useful sanity check is whether every model on the shelf can be connected to a specific row with a stated reason.

Can two different coding machines share the same TIJ cartridge?

They can if they agree on band width and on the electrical interface generation the printer expects, and if they code the same material family. A shared shell shape on its own proves very little, because a cartridge can seat correctly and still not be recognised reliably when the contact arrangement or configuration sits a generation apart from what the machine is looking for.

Is it a mistake to standardise on a single cartridge model?

As a target, usually yes. Concentrating an entire coding operation onto one part number concentrates risk: one batch problem or one shortage stops every line at once. A more robust position is a small set with a documented alternative qualified for the highest-consequence work, kept on the shelf deliberately rather than by accident.

Why can I not converge ink chemistry to reduce the number of models?

Because chemistry is set by the surface the code lands on, not by the machine firing it. Aqueous and solvent formulations behave differently on absorbent and non-absorbent materials, and moving a machine onto the neighbouring line’s chemistry to save a shelf slot trades a stocking problem for a printing problem. There is also a hardware reason: mixing chemistries on the same print engine risks cross-contamination and inconsistent adhesion, so standardising within a station matters as much as standardising across the site.

What should happen when a new coding machine is added to the fleet?

Treat it as a question about the existing matrix before it becomes a question about a cartridge. If the new machine matches an existing row on band width, interface generation, and material, the model count should not increase, and the specification should state that before the order is placed. If the arrival replaces an older unit, check whether it retires a model as well — that outcome is common and is routinely missed.