Ramblings of an aging IT geek
← Ramblings of an aging IT geek
hardware

multi-material printing, and the honest cost of those colours

My experience running multi-material prints with an AMS-style changer, what it's genuinely good for, and the waste and time you pay for the privilege.

A 3D printer mid-print

I've had a multi-material setup running for a few months now, the four-spool changer bolted to the side of the printer, and the question I keep getting is whether it's worth it. The honest answer is: for the thing you think you want it for, probably not. For a narrower thing, absolutely.

Let's start with what it actually does well. Two materials in one print is the real prize, not four colours. A part printed in your structural filament with dissolvable supports underneath comes off the bed with overhangs that would otherwise need either careful design or a lot of sanding. Embedding a flexible gasket into a rigid housing in a single job is genuinely lovely, the sort of thing that used to mean glue and swearing. When the second material earns its place by being a different material, the changer pays for itself.

A cluttered workshop bench

The colour case is where the romance dies a little.

The poop chute problem

Every time the printer swaps filament, it purges the old colour out of the nozzle before laying down the new one. On a model with frequent colour changes, that purge waste can rival or exceed the weight of the actual part. I've pulled finished prints off the bed next to a pile of stringy purge towers heavier than the thing I was making. You can tune the purge volumes down, and you should, but there's a floor below which you get colour bleed, and bleed on a multi-colour print is exactly the failure you were trying to avoid.

Then there's time. A four-colour print isn't a bit slower, it's dramatically slower, because the machine spends a large fraction of its life loading, purging, and wiping rather than printing. A job that's twenty minutes in one colour can be an hour and a half across four. You feel that most on small, fiddly models where the change frequency is high relative to the print time, which is, frustratingly, exactly the sort of trinket people reach for the colour feature to make.

There's also a reliability tax nobody mentions in the adverts. More moving parts means more ways for a print to fail at hour three: a filament tangle on a spool, a jam in the changer's path, a swap that doesn't seat cleanly. A single-colour print of the same part has none of those failure modes. I've had multi-colour jobs die two hours in over a feed problem that a single spool would never have hit, and there is a particular flavour of despair in cleaning a half-purged nozzle at midnight.

So, worth it?

Here's how I've settled it:

  • Two materials, functional: yes, every time. Dissolvable supports and rigid-plus-flexible parts are the killer feature.
  • Two colours, occasional: fine. The waste is tolerable and the result is clean.
  • Four colours, decorative: only when the print genuinely needs it, and brace yourself for the purge and the wait.

I don't regret the setup. I just use it far more for materials than for colour, which is the opposite of what the marketing photos sell you. As ever, the boring engineering use turned out to be the one with the real payoff, and the rainbow benchy was a nice afternoon I'm not in a hurry to repeat.