The Day I Banded 40 Doors the Wrong Way Round

It was a Thursday, and I had forty doors stacked on edge against the workshop wall, all banded, all trimmed, all looking perfect under the shop lights. A woodgrain melamine, the kind that's supposed to look like real oak if you don't get too close, for a client who'd specifically asked for a finish that didn't look like plastic in her kitchen. I'd run them through the edgebander that morning, feeling good about saving time doing them in one long batch instead of stopping and starting between smaller jobs.

Then I picked one up to pack it and turned it sideways in the light from the roller door, the way you do with any finished piece, half out of habit. The grain on the face ran one way. The grain on the band running down the long edge ran the other way. Not by a lot. Just enough that once you saw it, you couldn't stop seeing it, the way a crooked picture frame nags at you from across a room.

I checked the next door. Same thing. Then the next. All forty, same thing. An entire day's banding, and every single door had a seam where the pattern didn't line up with itself. For a second I actually stood there hoping I was being fussy, that it wouldn't matter once they were hung. It mattered.

Why grain direction catches people out

Melamine and foil-wrapped boards get printed with a grain pattern, and it runs one direction across the sheet, same as real timber would. A door blank needs banding that runs the same way as the face grain on each edge. Most rolls are printed to do exactly that, if you feed them the right way round.

The trouble is, nothing on the machine tells you it's wrong. The edgebander doesn't know or care which way the pattern runs. It just glues on whatever roll is loaded, however it's threaded, and the result is a clean, well-glued band facing the wrong way. The glue line is perfect. The trim is perfect. The only thing wrong with it is the one thing a machine can't check for you. On plain white melamine none of this matters, because there's no pattern to get backwards. On a woodgrain finish, where the whole point is the illusion of real timber, it's the first thing anyone notices, usually standing at the sink with nothing else to look at.

The thickness decision I used to skip

For a long time I used whatever band was already loaded on the machine, because swapping rolls mid-job is a hassle. That's a bad habit, and the door mix-up made me actually think about it properly.

Thin band, down around half a millimetre, suits edges nobody sees, like the inside of a carcass or the underside of a shelf. It's cheaper by the roll, and on hidden edges that's money well saved. On a visible door front though, it shows every dip in the board underneath as a wavy reflection you can see across the kitchen, not just up close. Most shops settle on something around 0.8mm to 1mm for doors, thick enough to hide minor defects without looking like a slab stuck on the side. Go up to 2mm and you're in commercial territory, built for knocks and heavy traffic, and on a domestic kitchen front it tends to look clunky, a visible ridge where your fingers run along the edge instead of a smooth transition.

Order of operations, not just order of edges

After that day I started banding long edges first, then short edges over the top, every time, instead of whatever felt quickest for that particular batch. Do it the other way round and you leave a small open joint at each corner where moisture and dirt can work their way in behind the band. It's barely visible when the door's new. Next to a sink or a dishwasher, where steam and splashback are a daily thing, that's the difference between an edge that looks the same in five years and one that's swollen and lifted at the corner by then.

The same logic applies to drilling. Hinge cup holes and shelf pin holes should go in after banding, not before. Drill first and the banding process can chip the edge of the hole, or the roller pressure can distort material you've already weakened by removing a chunk of it. Drill after, and the bit has solid, fully banded material on both sides, which means a cleaner hole and far less chance of blowout tearing out the back face where nobody's checking until the hinge won't sit flush.

Neither fix costs anything. They're just sequences, not tools or materials, not an upgrade to anything on the machine. You have to actually decide to do them in that order every time, rather than picking a job up again wherever your hands happen to land that morning.

It's not one door, it's the whole run

Here's the part that made the mistake worse than it first looked. In a real kitchen, doors sit next to each other, bank after bank along a run of units, and the eye reads them as one surface, the same way a row of floorboards reads as one floor. If the grain on door three runs a touch differently to door four right beside it, people notice, even if they couldn't tell you exactly why something feels off about their new kitchen.

Some shops number door blanks in sequence straight off the sheet for this reason, so the pattern flows left to right the way it would on one wide board, instead of jumping around because doors got shuffled on the way to the machine or pulled out of the stack in a random order. It's a five-minute habit at the cutting stage that only matters once the whole run is sitting on a wall together, which is exactly the moment nobody wants to discover it was skipped.

Rolls of band have the same trap hiding in them. Band and board are printed and coated in batches, and a new roll bought partway through a job can come from a different print run than the board you ordered months back. Under workshop lights, with everything lit flat and even, the difference can be completely invisible. Under daylight through a kitchen window, with the sun raking across the surface at an angle, a slightly warmer or cooler tone jumps straight out and makes an otherwise perfect door look like an odd one out. Hold a scrap of any new roll against an offcut of the real board, in real daylight, before trusting the label on it.

Where small errors turn into visible ones

Banding on and glued is only half the job. The trim and scraper pass that follows is where sloppy work gets shown up, often on a door that was glued down perfectly well. Feed too fast and the trimmer chatters, leaving a faint rippled line along the edge that you'll feel with a thumbnail before you ever see it in normal light. Feed too slow, or run a dull blade past its useful life, and friction heat can soften the band right where it needs to look crisp, leaving a dull smeared patch instead of a sharp line.

Corner rounding should come out even across every door in a batch. If it doesn't, usually because a scraper wheel is wearing unevenly or a door wasn't sitting square going into the machine, you get doors that are each individually fine but inconsistent once they're lined up on a wall of cabinets, corners that look slightly different from one door to the next. Worth checking the first and last door off a long run side by side, not just glancing at the first one and assuming the rest followed.

What I actually check now, before the machine starts

• Hold the roll against a face-grain sample and confirm the pattern runs the same direction before it's threaded
• Band long edges first, short edges second, every time, no exceptions for a quick job
• Match band thickness to the part's job: thin for hidden edges, 0.8-1mm for visible fronts, thicker only for genuinely high-impact pieces
• Drill hinge and shelf-pin holes after banding, never before
• Keep door blanks numbered off the sheet so grain flows across a whole bank of doors, not just within one
• Check a new roll against the actual board offcut in daylight before trusting the label

What that one day actually cost

Forty doors, re-banded. A day's labour gone twice over, once to do it wrong and once to strip the band off and do it again properly. New band, because you can't always reuse what you've pulled off without it looking patched. A delay on a job where the client had already picked a delivery date and told her kitchen fitter as much. And the quieter cost, the one that doesn't show up on an invoice: having to ring a client and explain why the doors aren't ready, without making it sound like you don't know what you're doing.

None of that came from a bad machine or bad material. It came from not checking one thing before I hit start, on a batch big enough that the mistake multiplied forty times before anyone looked closely enough to catch it. Small checks matter more as the batch gets bigger, not less. One door, you'd spot it in your hands before it ever left the bench. Forty doors stacked against a wall, you spot it once, right at the end, and by then it's already everywhere.