5th Gen - Review and Head to Head with Rep 2

[American 3D Printing]

This is the back of the X carriage, and shows a several things. You can see that round metal locating pin I referred to when I discussed how difficult it is to attach the extruder. You can see the mating pads for the pogo pins. And you can see how the fan's output is split between the cool part of the extruder above the hot end, and the ducting on the extruder that blows onto the build. You can also see where those 4 permanent magnets connect to.

If you are REALLY familiar with the Rep 2, you will notice that the X axis belt is about twice as wide as the Rep 2's belt. We haven't had any issues with our belt wearing out, but maybe somebody has? Another difference is that the pulleys at each end are about twice the diameter, which will result in a less sharp radius for the belt to flex. It is hard to see because it's painted black, but instead of the two rods for the carriage to ride back and forth on like you see in most printers, including the Rep 2, the gantry is made of a piece of sheet metal bent into an inverted U shape. Attached to that are some glide rails that the carriage rides on. These type of rails are also used for the Y axis for the bed of the Z-Morph.

[American 3D Printing]

There are two wheels to turn to level the bed, unlike the 3 on the Rep 2. One is in the front, and the other is on the side. I should also mention here that the bed is glass. We've purchased aftermarket glass for our Rep 2s that we use for PLA, and we use the original acrylic bed covered in PVA glue for our nylon and PET builds.

When you invoke the bed leveling procedure, it raises the bed with the extruder in the middle until the extruder just touches, then it goes down a couple mm and then just sits there. At first I thought it was hung or something, but every few seconds you could hear that familiar "thunk" that the Rep 2 does when it's lowering to the next layer, in this case it's raising the bed.

Ever.


So.


Slowly.


Go get a cup of coffee, because it really takes a long time to finally decide that it's happy. I think it's like 10 minutes or so. Then the extruder comes to the front of the bed and an animated graphic on the display tells you to turn the front knob in the direction shown until the LED on the extruder illuminates. It was telling me to turn the knob CCW, so I did.

Until the knob fell off.

[American 3D Printing]

Huh? I looked underneath and saw that there is a threaded stud sticking down, and the knob has threaded female threads. Okay, so I put the knob back on and turn it CW, until it is completely finger tight. The last couple of turns. I could observe the front of the bed lowering, opposite from what the firmware is telling me it wants (and it's also visible that it is the wrong way, but hey, I got the knob back on!).

So I started turning the knob CCW like the firmware wanted and I could see the front of the bed raising up as we wished it would, but then you could feel the resistance of turning the knob go to nothing, the bed stops raising and the light never came on. I kept going until the knob fell off again.

I repeated the procedure above AGAIN, but this time, just before the knob got all loose again, I pressed upward on the bed with my fingers and got the LED to come on. The display said it was "calibrating", then it moved the extruder over to the right side of the bed, middle of the Y axis, and again wanted me to turn the knob, but I aborted, because I knew that I had cheated to get it past that checkpoint in the procedure and it just wasn't right.

I figured I was just not doing something right so I call Makerbot tech support.

[American 3D Printing]

Hold time was only about 15 minutes before I got a tech. I explained the situation, he asked me some questions and I performed a few diagnostics and things for him. He somehow concluded that the smart extruder was faulty and got my shipping info for a new extruder (I have one customer who's on his 3rd extruder on his 5th gen, and I have another 3D buddy who returned his entire machine after the 2nd extruder quit working). I figure that the drones in tech support are just programmed to send out a new extruder instead of using real logic to troubleshoot stuff, but that's probably what they're taught. I know the company has been growing like crazy and they're probably having trouble finding good help. I get that. But i don't see how a new extruder is going to fix a bed that won't mechanically level. Who knows, maybe it will?

But, while on the phone I did learn a few useful tidbits.

The business part of the extruder moves up and down on the Z axis, about a mm or so. It can sense when it touches the bed, and it can also go up and down. That's pretty cool. Teaser - when you build, when most printers raise the extruder or lower the bed to do a Z-lift, the 5th gen raises the extruder - it's cool to watch.

I asked him how the machine senses Z limit, as I didn't see any limit switches like all our other printer, including the Rep 2. He said that it was a laser. Later in the conversation he said it was a hall effect sensor. I don't know exactly what it is, but there certainly aren't any mechanical switches visible anywhere on the machine, not just for the Z axis, but for the X and the Y as well.

As for how it is actuated to move up and down, I'm not sure. I know that on the high speed pens plotters I used to design, we used a voice coil driven by a linear amplifier to moved the pen up and down.

[American 3D Printing]

I shrugged my shoulders and decided to make something, the ubiquitous stretchlet that comes with all Rep 2s and 5th Gens. Turns out the one inside the 5th Gen's memory is on a raft.

How nice.

The raft itself is an identical slice to the one that good old Makerware 2.2 and 2.4 make:

[American 3D Printing]

Drum roll:

It took ONE HOUR AND 8 MINUTES!!!! To make what the Rep 2 does in 16 minutes! WTF??

Bear in mind this is on top of that darn Z-height calibration and heating which total around 15 minutes per build.

Okay, calm down, maybe it's making .05mm layers at 15mm/sec or something? I can't see the slice because it is inside the machine. Maybe it is something in the firmware? The support tech wanted to know what version firmware I was on, and I have read that there have been a LOT of firmware updates on these things.

[American 3D Printing]

Oh and for the benefit of the doubt, maybe the build quality is better?



Please forgive the point and shoot image quality, the good Nikon's at home, but trust me, it is absolutely impossible to distinguish which one was made on the Rep 2 and which one was made on the 5th Gen. Remember, not only does the actual build take 4 times longer, but you have that 15 minute Z-height and warmup procedure before the timer starts ticking.

A couple other random observations that i hadn't mentioned before...

The 5th Gen is noticeably louder than the Rep 2. In fact it is now our 2nd loudest machine after the Z-Morph, which is in a class by itself. But then, the Z-Morph is about 5 times faster and every bit as good. The stepper motors are really musical though, second only to our plywood Type A machines Series 1.

[American 3D Printing]

This rook from Thingiverse takes about 56 minutes to build on the Rep 2. One hour 44 minutes on the 5th Gen, not including that 15 minute pre-print overhead.



As with the stretchlet, the one from the Rep 2 is indistinguishable from the one from the 5th Gen that took twice as long to build.



Right around now the guy that owns the 5th Gen saunters into the store and I explain what I've found so far. He almost started crying. I showed him the stretchlets and the rooks and asked if he could tell them apart. The closest thing he could find was the arch over the door of the rook that the Rep 2 built had a little dangly artifact, which was quickly scraped off with an Xacto knife.

[American 3D Printing]

Another random observation - we can't figure out how to see what the camera sees. The printer's owner told me he stumbled on it once but can't remember what to click to make it visible on the host's screen. I'll have to do a little research on that, unless someone reading this can post up with the answer?

Anyway, here's another standard object we make. This is just a test object I designed that we use to help dial in the slice and temperature settings for new materials on various printers. It normally takes about 36 minutes on the Rep 2. As you can see, it took 58 minutes on the 5th Gen.



As before, you can not tell any difference from the Rep 2 version.

[American 3D Printing]

Next up is another standard test object we make, a dodecahedron from Thingiverse.

As you can see below, it is big enough that the non-level build plate is causing issues. The back stuck and the front didn't, of course the front of the build plate is too low. We aborted the build after snapping this picture.