gCreate's gMax 1.5 & gMax 1.5 XT

[jlgtx]

Review summary: The large print volume is really the only impressive feature. The printer itself is hobbyist-grade; the high-quality aluminum frame is betrayed by cheap PLA plastic parts. The gCreate support team is responsive and as helpful as they can be, but they can’t fix the firmware problems. Do not expect to do any production-level work (e.g., using any material other than PLA) with this printer without a significant investment in additional customizations, and even then it might not be up to the task.

What we like:
* At this price point, 16x16x21” is an impressive print volume.
* gCreate support is very responsive and helpful.
* The overall build quality is very good. Electronics & moving parts seem to be quality. The aluminum frame is very solid.

What we don’t like:
* At this price point, a printer should have significant usability improvements over DIY hardware; the gMax doesn't.
* gCreate can’t help you with firmware issues—and there are issues.
* Plastic parts on the printer are made from PLA instead of production-quality materials like ABS or nGen.
* The BLTouch Leveling Sensor has no connection to the actual Z position of the nozzle, so every print requires manual “babystep Z” first-layer height adjustments.
* gCreate does not provide consistent-quality profiles for common slicing apps. For example, their Simplify3D profiles are pretty good, but their Cura profiles are essentially worthless.

—
Unpacking was straightforward, assembly was mostly straightforward, nothing too difficult. We ordered the optional cart (https://shop.gcreate.com/products/gm...ocking-casters), and the completed assembly is very sturdy and easy to move around.

The heated build plate (https://shop.gcreate.com/products/16...tal-controller) is well-built, but we had two significant issues with it. The plate comes with four 3d-printed clips to secure it to the gMax printer bed frame. We quickly found that four isn’t enough, because the bed can easily shift around. We horked several prints before we finally figured out what was causing the random X-Y shifts. The bed frame has 8 holes for bed support clips, so I’m not sure why gCreate doesn’t include 8 clips to begin with. I emailed gCreate support (support@gcreate.com) and they sent me the STL file for the clip. Theirs are printed in Polymax Polyflex, but we just used nGen, and they actually hold the bed more securely than the flexible clips.

When you install the clips at all 8 points, you will probably have to reposition the bed rails slightly, because the added width of the rearmost set of clips just barely fits between the Z-axis screws. The rails on our printer were slightly misaligned to begin with, causing the bed to jitter at extreme Y travel, so we needed to realign the rails anyway.

The other issue with the heated bed is that it has no connection with the printer, and it has no master power switch. If it’s plugged in, it’s on. We ended up addressing this by running the bed power through a relay controlled by a Raspberry Pi.

We also bought the Dual E3D extruder (https://shop.gcreate.com/collections...3d-hotends-set), and my advice is…don’t bother, unless you really need two heads.We thought we needed two heads, so we could print support material on a specific large part we needed to print. It turned out to be easier to revise the large part and eliminate the horizontal overhangs than to make the dual head configuration work reliably.

One issue with the dual head is that the second extruder is always in the way. If the primary extruder drops a blob of filament, the second nozzle will eventually bump into the blob, causing the primary nozzle to create a new blemish at another location. If the second nozzle is hot, it can mar the finish of already-printed material as it passes over, causing “ghosting” effects in the finish. Worse, if the printed material warps, the second nozzle can hang on an already-printed part and cause print-ruining X/Y jumps.

The more problematic issue has to do with the fact that the Marlin firmware is not written specifically for the gMax and its hardware permutations, so unless you’re adept at tweaking the firmware (or have the time to learn how), you will run into multiple problems with the printer. For example, on large prints, we repeatedly encountered random X/Y/Z shifts. Some were attributable to hardware error (binding, bed shift, etc.), but many were not. And let’s say you want to add a filament sensor to your configuration--because, let’s face it, even a $50 inkjet knows when it’s out of paper or ink, for crying out loud. According to gCreate support, the firmware has filament sensor support enabled by default—all you have to do is install a filament sensor, connect it to the proper pins on the control board, and go. (For a how-to, see http://bit.ly/2AWnDnI). But the firmware has no idea that your printer has a dual extruder, so if the filament runs out on the second extruder, you change it, and press the button to resume the print, it will resume printing at the wrong spot. We also found that the filament sensor code in the firmware was prone to false positives. And gCreate support has no real control over the firmware, which means they have no ability to help solve these problems.

It’s beyond me why a $4k printer can’t determine the exact height of the bed (the Lulzbot has no problem doing it, every time). There’s no reason anyone should have to tweak the Z height for the first layer; that’s just ridiculous. What good is a 36-point bed-leveling procedure if the printer still doesn’t know where the bed is in relation to the nozzle? The Lulzbot nails the first layer every time, all by itself, because it uses the nozzle to level the bed. The gMax’s BLTouch sensor is a neat gizmo, but it has no predetermined spatial relationship to the nozzles…it’s mounted on spring-loaded screws and is adjustable. The upshot of it all is that you can expect to have first-layer issues with the gMax.

Fortunately, we found that most of the firmware-related issues can be sidestepped by using OctoPrint (http://octoprint.org) as a print server. In addition to offloading the printer control to a smarter device, OctoPrint offers greater configurability and reliability for add-ons like filament sensors, heated enclosure, etc. Our filament sensor false positives disappeared when we connected them to the OctoPrint server instead of the gMax controller. You'll have to make some adjustments to the BLTouch apparatus, and the stored Z offset in the firmware, to try to get a consistent first-layer position. Also, due to yet another firmware bug, you cannot print from an SD card while the USB port is connected (even if OctoPrint isn't using it).

Because gCreate makes many of their printer parts out of PLA instead of production-quality materials like ABS, nylon, or even nGen, the gMax will literally melt apart on you if it gets too much above room temperature. PLA will basically start falling apart above ~40^C. This means that the gMax starts falling apart above ~40^C, which we can confirm from personal experience. One of the the first maintenance issues we encountered was during an early test print using nGen, which requires an 85^C bed temperature. That was hot enough to cause the extruder’s cooling fan shroud to soften, droop into the print path, and get broken off.

Because of the PLA parts, gCreate officially does not support the use of a heated enclosure—but, as we learned, even just high bed temps can damage the PLA parts, without a heated enclosure. So, if you want to print parts in anything but PLA, you will have to print better parts to replace the gMax’s PLA parts (which includes every red-colored part in the product photos, plus a few others). If you want to print large parts in anything but PLA, you’ll also have to build your own heated enclosure, and void your warranty.

The aluminum frame is so solid, why does gCreate kill it with junk-plastic parts? Why do they claim that the gMax can print in ABS and other materials that require high bed temps, when it clearly cannot?

We purchased the gMax 1.5 XT+ specifically to print an accessory part for a device we sell. The part is 14x14x12” in overall size, our quantities don't justify injection molding, and Shapeways wants $1k to print the part in their standard white nylon. We’d been using a Lulzbot Mini with great success, so we figured we could spend $4k on a pre-built, well-supported large-format printer, print our part in ABS or nGen, and it would pay for itself with just a few prints, plus we’d be able to use it for prototyping and other things.

The decision to purchase the gMax was heavily influenced by two things: our positive 3D-printing experience with the Lulzbot, and the glowing online reviews of the gMax. In retrospect, we now know that Lulzbot is way ahead of their competition when it comes to ease-of-use. They’re definitely way ahead of gCreate, at any rate. If you want a solid, dependable workhorse 3D printer, we can recommend Lulzbot. We’ve also learned that the online reviewers of the gMax were apparently so overawed by the large print volume that they failed to really put the printer through its paces. One huge PLA print should not have been enough to convince them that it was a great printer, because the gMax isn’t great. It’s just big. Caveat emptor.

Speaking of PLA, I asked gCreate support why they use PLA parts on their printer. They replied, and I quote, "We are not set up to print the parts in a different material. We do not have the ABS on hand nor do we have the setting for all our parts to be printed in ABS.” So how did they test ABS and the other filaments listed on their website (http://www.gcreate.com/gmax-15-2016-series) in a printer that melts at high temps? ABS requires higher bed temps than nGen, so they should have caught the problem with the fan shroud. They certainly didn’t test printing large objects with anything but PLA. They didn’t test their large-format printer with high-capacity filament spools, because the stock (PLA) spool supports deformed when we parked a 2.2kg spool atop them. We had to reprint the parts using real plastic, and add horizontal support rods for more strength. Bottom line: gCreate really hasn’t tested this printer.

Perhaps gCreate should invest in a couple of Lulzbot Minis. That’s what we used to print new nGen parts for the gMax. Good thing we had a high-quality 3D printer on hand! With the new nGen parts, all of our re-engineering work, and remembering to adjust the belt tension regularly (which should be easy to do, but isn’t), the printer seems to be holding up reasonably well, even in the heated enclosure.

Speaking of Lulzbot, they provide their customers with a customized version of Cura, optimized for their printers. The stuff works. You have to try pretty hard to screw up a print. In stark contrast, gCreate provides profiles for various slicing software, but the quality isn’t consistent. Their Cura profiles, for example, don’t work with the latest version of Cura, and don’t include dual extruder support. After many unsuccessful attempts to come up with a working Cura profile that would make the gMax happy, we finally gave up and tried Simplify3D, and have had much better success. Apparently gCreate puts quite a bit more effort into their Simplify3D profiles. In our opinion, Simplify3D’s price tag should be factored into the street price of the gMax, because you’re going to need it.

If we hadn't gone past the 14-day no-questions-asked return window, we probably would have sent the gMax 1.5 XT+ back to gCreate. We thought we were buying a printer; what we got was a new hobby…and we really don’t have time for this hobby. But we're stuck with it, and we’ve invested a lot of time, energy, and money toward making it do things that a $4k printer should have been able to do out of the box…so at this point, we'll just keep trying. I finally got a successful print of our large-format part. Just one, though. So far.

I really want to like this printer. The frame is solid, the build volume is great, and gCreate support is helpful, but the printer needs a lot more work, primarily usability engineering and overall product testing. The gMax is just not user-friendly. We’ve spent far more time fixing the printer than actually getting good prints out of it. Out of the box, the gMax is not a workhorse, it’s not reliable, it’s just big. And that’s a shame, especially considering the price tag.