The distance between a working prototype and a shippable point-of-sale unit is mostly paperwork, tolerances and test coverage, not electronics. A 1.28-inch round TFT that lights up on an evaluation board will still need a defined flex length, an agreed adhesive, a bonding specification and a test that catches the unit that works at room temperature but fails after the printer has been running for an hour. Treat the display as a mechanical part with an electrical interface, and the production questions become predictable. Treat it as a component you buy at the end of the design, and every one of those questions becomes a schedule risk.
Key takeaways
- Freeze the mechanical drawing first, including the flex exit direction, because every later change to it invalidates tooling and fixtures.
- Write down what remains customisable and what is fixed, so the supplier and your own production team are working from the same list.
- Define the incoming inspection test at the module level, not the finished-product level, so a bad batch is caught before it is built into a housing.
- Confirm the substance and waste documentation you will need for the markets you ship to before the first production order. [1]
- Agree an end-of-life and notification process in writing, because a display change late in a product's life is more disruptive than any electrical fault.
Products built around a small round display tend to look simple from the outside. A single panel, a host controller and a plastic housing. The complexity sits in the interfaces between those three, and it only becomes visible when the first hundred units are assembled and a handful of them behave differently from the samples.
The usual reason is that the prototype was built from a development board and a hand-soldered connector, while production uses a flex tail, a jig and a reflow profile. Each of those changes the mechanical and thermal environment of the display. A design that ran at 400 nits on the bench may be perfectly reliable in production, but only if the stack was specified with production in mind.
This article covers the practical transition for a 1.28-inch, 240x240 round TFT with a four-wire SPI interface, from the first sample order to a repeatable production test.
What the supplier actually customises
Customisation on a small display module usually means a defined set of choices rather than an open design. The panel itself is fixed; what varies is the flex or connector, the touch layer if one is fitted, the cover glass or window, the adhesive, the backlight arrangement and the marking. Understanding which parts of that list are on the table is what turns a vague enquiry into a quotation.
It also helps to write the fixed part down. If the driver, the resolution and the interface are already settled, then the conversation is about mechanics and logistics, and both sides can move faster. A specification that leaves the interface open invites a sample that cannot be qualified against anything.
For a product that ships in modest volumes, it is usually better to adopt an existing module geometry than to commission a new one. The time saved in tooling and qualification is worth more than the millimetres saved on the housing.
Interface choices that survive production
Production teams tend to prefer modules that keep the host interface narrow, and a four-wire serial interface such as the one carried by a 1.28" SPI TFT display module needs few host pins and short traces. That is an advantage in production as well as on the bench: fewer nets means fewer opportunities for a rework station to introduce a fault, and a simpler test fixture. A serial peripheral interface uses a separate clock line and shifts data synchronously, so the timing conversation becomes a question of clock rate and setup time rather than of addressing or arbitration. [3] The trade-off is bandwidth, which matters only if the product wants to animate the screen.
Whatever interface is chosen, the timing margins should be measured on a real production board rather than only on the development kit. Track length, connector quality and the host's drive strength all shift the margin, and a timing failure that appears only on a fraction of units is the most expensive kind to find later.
A productionisation order that keeps risk visible
- Lock the mechanical drawingAgree the active area, the flex exit, the bend radius and the connector position in a controlled drawing. This is the document that fixtures and housings are built from.
- Qualify the first article against the drawingMeasure the sample, don't eyeball it. Active area position, thickness and flex length are the dimensions that decide whether the unit assembles without stress on the bond.
- Define the module-level testScreen for open lines, stuck pixels and backlight function on every module. A test that only runs on the finished product hides which supplier is responsible when failures appear.
- Run the environmental check on assembled unitsTest at the temperature extremes of the mounting location with the printer or motor running, because heat soak changes the failure mode compared with a bare panel on a bench.
- Agree change notificationAsk in writing how you will be told about a process change, a material substitution or a discontinuation. This single clause protects the product's whole service life.
- Keep a golden sampleRetain a sealed reference unit from the qualified batch, including its measured luminance and a photo of the displayed test pattern, as the comparison point for later deliveries.
Questions worth answering before the first purchase order
| Area | Question to the supplier | Why it matters later |
|---|---|---|
| Mechanics | Which dimensions are tooled and which can still move? | Tolerance stack with the housing and the cover window |
| Interface | What timing margin is documented for the serial bus? | Intermittent faults on a fraction of units are the hardest to diagnose |
| Optics | How is luminance and colour defined and measured? | Two deliveries that differ visibly in tint will be noticed by customers |
| Compliance | Which substance and packaging declarations are available? | Market access depends on documents, not on the panel itself [1] |
| Supply | What is the notification process for changes and end of life? | Service spares and repair parts for the whole product life |
| Test | What does the outgoing quality check cover? | Determines how much incoming inspection you still have to do |
Sealing, cleaning and the environment the panel lives in
Retail equipment gets cleaned. Alcohol wipes, spray cleaners and the occasional spill are normal operating conditions for a terminal, and they attack adhesives and exposed flex before they attack the glass. If the front of the unit is expected to resist water and dust, the enclosure rather than the display carries that requirement, and the relevant protection level should be stated against a recognised classification so the test method is unambiguous. [2]
The practical consequence for the display is that the bond line, the flex exit and any gasket are the parts that need attention. A display mounted behind a sealed window with a controlled air gap behaves differently from one bonded directly to the front plastic, and the choice affects both cleanability and optical performance.
Ask for the change notification clause early. Teams spend months on electrical qualification and then discover that the module they qualified has been quietly revised. A written notification process, plus a sealed golden sample, is the cheapest insurance available on a display programme.
Sources
Frequently asked questions
At what volume does a custom round TFT become worthwhile?
It depends on what is being customised. A different flex or a cover glass change is routine at modest volumes, while a new panel geometry or a new tooling set is normally justified only at high volume. Ask the supplier which items carry tooling cost and which are configuration choices.
How should the display be tested in production?
Test at module level and again on the assembled unit. The module test catches electrical and cosmetic faults before value is added; the assembled test catches mechanical stress from the flex bend, the housing and the cover glass, plus any optical problem introduced by the window.
What causes panels to fail only after assembly?
Mechanical stress on the bond, a flex bent tighter than its rated radius, a connector that is not fully seated after the housing is closed, or heat from another subsystem. All four appear only on the finished unit, which is why the assembled environmental check cannot be skipped.
Do we need our own incoming inspection?
Yes, at least a sampling check. Use the supplier's outgoing criteria as the reference, then verify a sample against the golden unit for luminance, tint, cosmetics and the interface working at the specified timing. This keeps the responsibility boundary clear if a delivery is disputed.