Feature breakdown – The CEL Robox!

Many features in a small box – too many?
Printer manufacturers listen up, these are the things that would improve any printer!

CEL Robox

Hey everyone, Tom here, and with me today i have the CEL Robox (Registered Trademark), which i’ve been testing out over the last few weeks, and it’s hands-down the most innovative printer you can possibly buy. Not all of the features they packed into this box work that well, but i still want to show you what they are, how they work and maybe, we’ll see them in an improved or different variety in a lot more printers. Which i’d really appreciate.

If you’re looking for a full-on consumer-oriented review, i did that one for All 3D printing, so check that out on their site, and let me know in the comments below this video if that less-technical style is something you want to see on my channel as well.

But for this one, let’s head right in, starting with most apparent feature: The SmartReel (™).

It is basically the same thing as UFID, the Universal Filament Identification idea developed by the community which, in that form, never came to fruition since there were never any actual manufacturers involved. The idea with UFID and the Robox SmartReels is that each filament spool has a tiny EEPROM that core material parameters can be written to, like the glass transition temperature, actual filament diameter, ideal printing temperatures, color and so on. The SmartReel also stores how much of the spool has been used up already, and, thankfully, that information as well as the info on what type of filament, including totally custom profiles, can be written to any SmartReel, so once you use it up, you can wind your own filament onto it and have the correct custom profile loaded each time you clip that Reel into the printer.

Next up, even the single-filament Robox comes with two nozzles, there’s a single filament path going to both of them, and to control pressure inside the melting zone as well as which nozzle is used, each nozzle has a valve pin that physically constrains or fully blocks the nozzle when not it’s in use or when material ooze needs to be controlled. And this movement right here, i believe, is what’s controlling the valve opening and closing. As a side effect, it also lifts the unused nozzle away from the print.

Now, the auto-bed-compensation thing is also integrated into, actually, this entire assembly. As the head probes single points on the bed, it actually presses the nozzle against the surface and you can see the entire print head rocking back and forth as it does so. And this is intentional. Instead of using any sort of dedicated sensor, what’s registering the “bed hit” event is the top guide lifting from the rail, and this opens an electrical circuit between the print head and the rail, just like the Lulzbot Mini closes one as it touches the hotend to one of the shims around the bed. On the Robox, i’m sure if i’m liking this system since it’s very susceptible to contamination and actually has a sliding metal to metal contact, which i’d be afraid that it wears down eventually. But for now it works, and i do have to applaud the engineers behind this for finding what seems to the cheapest possible to implement automated bed compensation.

One other interesting feature is the way the enclosed build envelope is handled. The Robox has one, and even though it isnt’s totally watertight, i mean, after all, there’s open fan mounted up top, it does a decent job of keeping heat in and, probably more importantly, fingers # out. In this particular unit, at least i’m told it’s an isolated issue, neither the safety functions of keeping the printer from working with the door open nor the latch to keep the door closed worked all that well. The “door open/closed” detection is simply a small switch on the top of the door, but the latch is actually implemented quite efficiently again. Because it’s not actually like an independent motor or anything, it’s simply controlled by the bed hitting this lever right here as it moves forward. Pretty neat!

The last thing i want to touch on is the filament path from the reel to the hotend. This is a bowden system with a relatively short length of tubing, which, i guess, became necessary to keep the Robox’es mechanics on the lightweight side, but the way filament feeding is implemented is something i haven’t seen before, either. Because basically, the only things that are special to what makes this better than everything else is 1) a sensor that detects when filament is inserted or when it runs out, so it basically just senses whether or not there’s currently filament in the drive path. And with the procedure that starts as you insert the filament, which heats the hotend and pulls the filament into the extruder, that takes out a lot of the fiddling when it comes to swapping filament quickly. There’s also a sensor in here that detects if the filament is actually being moved forward to detect filament stalls and keep the “loading” process from pulling in too much or too little filament, but really, that’s just the icing on cake. Now, number 2) is a simple clicky button that does that process in reverse. Some 3D printer manufacturers have implemented the same thing in their LCD menus, but if the printer doesn’t have one of those, i think that’s probably the one functionality that gives the most bang for the buck.

So those are what i think are the most noteworthy features of the CEL Robox that I’d like to see more manufacturers pick up on. It does come with a lot more stuff that’s marketed as the best thing since sliced bread, but really, a fine pitch spindle for the Z-axis, an RGB LED strip that only ever glows blue or white / or the promised, announced expandability of the platform to, like, a mini CNC mill, even though the mechanics are far too weak for that, those aren’t exactly new things that only the Robox can offer.

So thanks for liking, commenting and subscribing (totally stole that line), I’ll see you all in the next one.



Dual Nozzle

Fine-Pitch Z drive


Extruder feedback

Nozzle valve

RGB lighting

Build chamber door


  • Easy setup, mostly automated
  • Auto-tramming works flawlessly
  • Reasonable print quality
  • Cartridges work well, but mine shows 0m left. Also, expensive per kg
  • Software & Firmware buggy
  • Will print with door open
  • Crashes during purge process
  • Custom filament profiles disappear & only allow printing with cartridges
  • Bed with good adhesion, but sensitive surface
  • Dual-nozzle rarely used and results in poor print quality
  • Perfect machine if everything worked as intended
  • Polycarbonate printing is false advertisement (nozzle only does 260C)
  • Software okay to use, but could be more intuitive

Buy one on Amazon: http://amzn.to/1QlCpWg [US]
http://amzn.to/1N13QEc [DE]


You can support me without spending a single penny!