Welcome to the SeeMeCNC Community

I have some money to burn and I really like this printer and I also like that the company is fairly local to me. I had an intern try the V2 version and his feedback was below. I am hoping active users can convince me this is no longer the case and v3 works better. The imperfections were because he used USB and not SD so I understand that one.

extremely slow to heat bed

non modular components. if any part breaks, you have to disassemble the extremely complex extruder.

Low quality resistors to heat the extruder. They burn out and you have to replace them. I didn't even know this issue existed anymore.

Not an all metal hotend (though I think this has changed in newer models).

Have to unscrew the bowden tube in order to change filament

uneven bed. Hassle to level manually with no way to level it in software. Even the auto level results in an uneven first layer for big parts (and therefore also bed adhesion issues).

systematic small print imperfections

The bed is somewhat faster to heat, and it can be made even faster with a 24V power supply (As you're assembling it, it's easy to put an SSR and 24-12V converter in the base. with the SSR, you don't need a big return wire to the Rambo for the heated bed, even without a 24V power supply, loosing the resistance from the feed wire is a definite benefit). But it is still somewhat slower than people would like, but it's a larger bed.

The hotend is now largely E3D compatible, and is as all-metal as the E3D, although it has some issues with the bowden not setting well. But the extruder is not complex, and while the hotend is somewhat complex, it is relatively easy to unplug and work on outside the machine (Or swap a spare in). The resistors are also gone. The Bowden tube never really needed to be unscrewed to change filaments. You may have had the ring on the old PTC fitting come off, or an intern who didn't push very much.

The bed is pretty even, and has software leveling that is pretty good, as well as built in probing for calibration.

Mind you, I own a V2 (and have played some with a V3 a friend owns, as well as troubleshooting them for others on the forums), and most of the issues mentioned were easily mitigated (Print from SD card, use PEI and lower bed temps or print with polymers that don't need higher temps, manually calibrated bed, 5 minutes with a drill and tap to install a heater cartridge and threaded thermistor), while the PEAK hotend gave pretty darned excellent PLA performance and acceptable ABS performance, especially compared to the PLA performance some reported with the E3D V6.

Hello Xenocrates, thank you for always posting. My crew is very dedicated, hard working, and always strive toward improvement. The community and support has always amazed me and is big part of why we've grown so much!

The bed circuit is separate, so a second power supply can be added to improve performance. The honest answer is, our 12VDC heated bed does struggle some at higher temps. Due to the power (Watts) to area ratio. So 24VDC solves it, but you'll need to back off max output in the firmware. There's so many modifications you can do with SeeMeCNC machines, you'd be on your own here.

As for the rest of our machine, our stock V3 is not comparable to a stock V2 at all. The V3 is a very mature reliable design that doesn't need anything out of the box. Sure, lots of cool mods make it even more fun, but it works great right out of the box!

The HE280 is an excellent hot end. Seat the PTFE. If (for some reason) you feel a need to take the PTFE bowden tube in and out, trim an 1/8" off the end, keep the cut square. The reason to cut the end fresh is to give the Push-To-Connect fitting a new place the "bite" on tubing. If you don't cut the end, the PTC fitting wears out the place where it "bites" the tube, and the PTFE bowden tube slips up. This causes a gap down at the heat break, and a usually jam follows. It pretty rare to have our hot end jam if you seat the PTFE bowden tube and use the locking clip on the PTC fitting. A September release will help make the PTFE seating be less finiky.
Leveling the bed is fully automatic. There are no screws to adjust. Probing is excellent. Myth, install a ferrite bead. Not true and I have no idea where this started, ugh. Someone did not understand how it works. Here's how it works in a nut shell. Send I2C programming to accelerometer), turn off I2C. Probe. The HE280 accelerometer acts like a switch, triggering an output when the g-force threshold is met. It works excellent! A clever algorythm (by pounching iguanna) automatically adjusts the sensitivity of the accelerometer (among other things). It is very reliable. A recent firmware update has further improved accuracy of the machine auto-calibration.

The whole purpose of my wanting the nozzle to actually touch the bed for leveling is because it's easier. You could install a nozzle that's 5mm longer and it'll work. Or put a thicker bed material down, probe, it will work. No special adjusting anything.

We test 100% of the heater cartridges for proper resistance. Of shipments we receive (from China) 40% normally fail our resistance tolerances. Bummer huh, but we make sure you get the good ones so they perform great. It's pretty unlikely you'll get a bad one. Resistors haven't been a thing since the V2. Neither has PEEK.

The HE280 is designed to continually operate at 280C if you need it to. Fail testing at 320C, the thermistor fails and I could not get it to catch fire.

@ ejschuh and anyone else. Stop by our new plant. play with a printer, ask questions.

Presently, the V3 is at full maturity and performs very well out of the box with high degree of reliability.

I bought a V2 and have been following the improvements on the V3:

Here's some thoughts:

• extremely slow to heat bed

I also found the V2 bed was slow to heat, but they provide a better power supply in the V3. However, if you want to run over 90C, it'll be slow and hard to get to 110.

The heated bed on the V2 and V3 are both 1.1 ohm, and my testing found that with an adequate power supply AND a dish towel set on top of it (adequate insulation), the heat bed reaches 90C in about 10 minutes. For the surface area, that's not an untenable wait.

• non modular components. if any part breaks, you have to disassemble the extremely complex extruder.

I'm confused. I routinely take off parts and put on new ones. Homemade effector platforms, several different hot ends, a dual extrusion setup, fancy blue aluminum aftermarket parts, (now unnecessary and arguably not even an upgrade) carbon fiber arms, a touchscreen controller faceplate, most of them printed on the V2 before I installed on. How is this not modular? (Also, doesn't the V3 have a quick-connect built into the new hotend circuit board thingey?)

BTW, this is NOT an endorsement of swapping out a new machine with a million different bits right away! Every smart person will tell you to build stock and enjoy using the machine, unless changing parts on the machine is what you want, more than actually printing things.

One positive experience I had in particular was that SeeMeCNC made some upgrade parts backwards-compatible. The new injection molded carriages were drop-in replacements for my original parts and have been great. No more checking carriage wheel tension!

• Have to unscrew the bowden tube in order to change filament

I've occasionally hit this (minor, 15-second-to-resolve) issue on the V2. The V3 has a redesigned extruder with a better filament path.

• uneven bed. Hassle to level manually with no way to level it in software. Even the auto level results in an uneven first layer for big parts (and therefore also bed adhesion issues).

I find this criticism hard to understand. The borosilicate glass is extremely flat. Even on my V2, before the autolevel, most people's experience is that you level the printer once, and then it just stays level until you mess with some other part of the printer.

I damaged my Rambo board , so I can't speak to the quality of the autolevel on that controller.

There IS a semi-frequently encountered phenomenon with delta printers that makes getting a perfect first layer difficult near the bed edges in a weird triangle pattern. I haven't hit that particular problem. My first layer problems are almost always to do with my filament/bed surface/bed temperature combination, and print-specific problems like ABS edges curling up in a drafty room.

You'd be shocked how many first-layer problems are resolved by tricks like wrapping a shower curtain around the printer to trap a little heat and protect it from drafts. I just printed a 9.5" diagonal rectangular ABS part with no corner pads or brim to hold the edges down, square sides, tall enough to risk curling, and none of the corners lifted at all, and the entire first layer was visibly equal height.

I can see an argument that a delta is, theoretically, more complicated to diagnose bed leveling issues, but on the other hand, you only have three identical moving parts, the towers, and your main challenge is to make sure each is perfectly square. Compare that to a reprap-style build where the Y-axis could be slightly tilted on one end vs. the other, or uses z-screws that can bind. (Just to cite one painful example I helped a friend work through.)

• systematic small print imperfections

(I'm writing this assuming you're newish to 3d printing. Apologies if it's something you already know.)

For any open-source 3d printer, you will quickly discover that the biggest obstacle to print quality will be the particulars of the slicing software and the quirks of whatever filament you're using. I fight with small imperfections most days and 98% of the time they are traceable back to something stupid the slicer software decided to do, or wasn't quite right for that batch/color of filament. (The other 2% is because I pathologically change out parts on the printer for the fun of experimenting.)

For any high quality DIY kit (Prusa, SeeMeCNC in particular), the next biggest challenge to print quality will be the ability of the person assembling it to follow instructions very carefully.

For any cheap DIY kit, you must both come armed with electrical and mechanical expertise and be prepared for headaches stemming from imprecisely manufactured parts. I've gone this route and a lot of time with calipers, levels, and needle files was involved.

In General

I've assembled Reprap and Delta kits, and two things that stand out with the Rostock were the quality of the documentation and support. They are far above anything else I've ever seen from a small-company open-hardware project.

The biggest downside I can think of for the Rostock is simply that it's a delta printer and will be very tall relative to the printable area. The round, very tall print area is arguably a downside because many Things are rectangular, and a print that uses the entire gigantic Rostock printable area is, almost by definition, an epic long-running print.