FEP on Ember - Preliminary results:


I successfully printed my first model on Ember using FEP instead of PDMS.

I challenged myself to do a proof of concept. As any other preliminary test, the first results only show potential, not quality nor speed or ease of use.

My process:

  • Unscrewed and detached the glass/PDMS assembly.
  • Used scalpel to carve a window on PDMS.
  • Used a chisel to remove the inside part of PDMS from glass, leaving a border in place.
  • Used a diamond encrusted drill bit to open a hole in the glass, for air pressure to equalise. Managed to break the glass not once but twice !
  • I sandwiched a sheet of FEP film and tightened the bottom screws.
  • I could not get the tension right, Adhesion was not possible, since Ember’s build platform is large I printed a 2mm pedestal in an effort to achieve tension. To do that I used an unmodified tray, and let the printed plate on the platform.
  • I increased and Z axis overlift, set 0 degrees rotation and disabled jamming detection.
  • Clicked print*

Result shows good promise. There was some sweating of resin in the bottom, I advise caution to anybody who tries to replicate this test.

I am sharing this publicly, in an effort to urge others to pick up the torch and produce an aftermarket VAT for sale, to users like me who want Ember to reach its potential.

  • Firmware needs to be edited, we need to disable homing before job start and after job end.
    Also, Pause function did not work correctly, platform did not raise at inspection height. It might be a bug.


Nice! This one is promising.


Excellent! I just read your other post and was thinking about trying this.

I don’t have any of the new trays, so I’m curious about a couple of things:

  1. Why carve a window? Wouldn’t have been easier to just lay the FEP over the PDMS - or remove it entirely? Was it to maintain the original PDMS height?
  2. Why did you sandwich the FEP? Were you using FEP with no adhesive backing?
  3. Why drill holes in the glass?


Glad you like it Scott !

I just followed the principal of other DIYers, the only challenge was thinking inside Ember’s box.

To your questions:
1 - You need empty space bellow the film in order for it to become concave shaped during pull and release the slice in a non volatile way. The correct height is also essential as you very well pointed out.

2 - Yes, no adhesive backing on FEP.

3 - Holes are needed, otherwise air pressure differential won’t allow FEP to balloon up for the release.

PDMS release relies on a property of the resin to not bond when in contact with oxygen molecules that are present in its porous structure. Then you just slide to counteract the vacuum forces. With FEP, a bond is occurring but vacuum force is almost eliminated by the concave shaped that forms during pull.

There is another film product out there, called Teflon AF. It is hundreds of times more efficient than PDMS and costs thousands of $. Carbon 3D printers use it and it is the foundation for their layer less printing.


Thanks for the additional info Yianni. Very promising!

It sounds like your goal is to remove the suction force and eliminate rotation - thereby increasing your print speed dramatically. To this end, having the ‘trampoline’ effect of a membrane should work great.
The edges would start to peel away first, flowing toward the center.

With the FEP not supported however, do you get thickness variations in the center?

I have seen several posts by other DIYers that use the FEP with an adhesive back, stuck directly to the PDMS window. Their main goal is to prevent the clouding and extend the tray life. This would prevent thickness variations, but still require the rotation for separation.

Perhaps as a ‘best of both worlds’ idea:

  • remove the PDMS entirely
  • add the bleed holes in the glass inside the edge sealed area to allow air to flow
  • stretch the FEP flat across the glass, giving it the support it needs for flat prints
  • Seal the edges of the FEP against the glass and tray, leaving the center window of the FEP unattached / unbonded to the glass.

This may require an additional smaller glass slide to be bonded to the original glass to build up the Z height.


My main objective is to eliminate problems from clouding.
It makes every next print come out a little worst until the print quality is unacceptable. And that happens in days !

Having worked with tilt printers previously (ET Aureus), I noticed that an added benefit was that if you have one model fail during printing (let’s say from bad supports) the rest will not be influenced. So there is an additional upside of not having to swipe.

Speed increase is welcomed, but not of importance to me. Z speed needs to be slower and overlift longer. Wait time too. So it’s fast, but not lightning fast.

I don’t want to come off as I m inventing the wheel here. The only challenge here is to fit this system in Ember.

FEP release is being adopted in many printers, both DIY and commercial ones. They just use different names for it, ET calls it PreStrechedAssembly. Commercial printers need to pay license fees to incorporate this patent. Thickness variations should not occur due to tension, combined with enough wait time for the excess resin to drain out so film levels flat again.

I am afraid your idea will not work with passive release, it does work perfectly with tiltling though. Aureus works on that principal (FEP adhered on PDMS) and you need to change the VAT every 6 months, but I have seen people stress it up to a year or more.