Geeetech G2s Printer Kit Review

[Roxy]

Review of the Geeetech G2s Printer Kit

People... I ordered a Geeetech G2s Printer kit off of eBay. It shipped quickly and my initial reaction upon opening the kit was that it was very very nice. I was impressed with the precision of the laser cut pieces. Everything looked well engineered. All of the pieces came clearly marked in separate bags so parts were easy to find and there was no confusion about what went where. With that said, right now I have a sour taste in my mouth.

I'm halfway through the build of the kit and I intend to publish a full review of everything. Initially, I asked that a Geeetech section be added to the RepRap's next to the MakerFarm printer section because I expected a lot of interest from people buying Geeetech 3D-Printers. I'm starting to worry that this migh
t actually be the case.

Like I say, I fully plan on doing a full review with comments on everything. But for now, suffice it to say: I recommend you DO NOT BUY A Geeetech 3D-Printer that uses any form of Plexi-Glass in its construction. There are a number of other problems I have run into also. But the one that really concerns me is the fact the plastics crack so easily. I'm only 1/2 through the build and I've had two important pieces of the printer crack and I wasn't doing anything wrong.

Please check out the first picture at http://www.geeetech.com/wiki/images/...nstruction.pdf to see the pieces that have cracked so far.

The first piece to crack was one of the triangular pieces that holds the filament spools up in the air. They are shown in more detail on page 32. They look pretty stout in the picture. What they don't show is how the locking mechanism is constructed. The locking mechanism is the same as the LCD Support on page 9. A small nut is inserted into the cut out space and then a screw from the far side of the mounting piece pulls hard against the nut to pull the two pieces together. For much of the design this is fine. But the Filament holder pieces are much too long to use this technique without some reinforcements that prevent wiggling. And flexing of the top will put much too much force on those little shelves of plastic. Any wiggling of the part will magnify the force on that little plastic shelf. But that isn't what broke the plastic. The precision laser cuts were not perfectly vertical. So as I was tightening the screw to pull the pieces together the filament holder piece was being forced into a tighter and tighter wedge. I wasn't even all the way seated and the little plastic shelf snapped off. And remember, we haven't even gotten to trying to replace empty spools of filament with heavy new spools. Do you think those pieces are going to be wiggled a little bit? For that matter, what about the shaking the spools will do as the printer is moving around back and forth? Well, I don't know about printing yet, but I can tell you that they didn't survive long enough to make it through a very careful and methodical building process. It would certainly make sense to put some kind of cross brace next to those tall filament holders to keep them from wiggling and cracking the plastic.

The second piece to crack is more worrisome. Please look at the picture on page 8. This is a corner of the base plate that holds everything. I got to the part where I needed to add the 6 smooth shafts that go to the top of the printer. (Incidentally, the instructions and the kit do not have the right mounting parts. This was fixed with a trip to the hardware store and some time shaping flat washers. Much more on this in the full review.) The real problem I ran into is those round holes to hold the vertical shafts are precision cut but with very little material strength around them. I got four of the shafts mounted. I rotated the printer base so I could do the back shafts. Just turning the printer base on the table caused enough force to crack the base. The crack runs from the cut out notch for the belt to the square hole next to it, then onto the next square hole and then onto the empty hole further away in the picture. This plastic they are using is SUPER BRITTLE! And once again, I didn't do anything wrong. I think this is similar to the first failure. The motor mount pieces and the display support pieces would only fit into the base plate from one direction. The holes were too tight to position them on the other side of the base plate. I think what happened was these pieces were acting like a wedge when they are installed and fully seated. I suspect there was a lot of force trying to tear the plastic apart and split it before I even got to the point where the tower rods are installed. They really need to check the precision of those laser cuts and make sure they are not forming tight fitting wedges in places they shouldn't. What is frustrating is I was carefully assembling something I spent a lot of money for and was planning on using for a long time. I didn't even get through the construction process without the base plate cracking. Come on....

My advice is to hold off on any Geeetech 3D Printer purchases until you hear this has been resolved.

Mean while, if you have a Geeetech 3D Printer, this section has been opened up to help support you. I would prefer the problems are firmware related, but given what I've experienced so far, my guess is we are going to have a lot of cracked plastic issues.

Notes for complete review so items don't get forgotten:

• Plastics are brittle! Precision laser cutting not perfectly vertical may have compounded the problem because previously installed pieces exerted forces on the sides as they are pulled in flush.
• Limit switch connectors are nicely done, but there is no clearance to install the switches without bending the connectors on each wire. And bending the connectors makes the connectors fail and fall off. I had to solder the wires directly onto the switches
• M4 washers are called out on page 33 and supposed to be used to secure the smooth rods in the plastic. But the M4 washers in Bag #6 were too small in diameter to accomplish the task. They were a smaller outer diameter than the actual rods. The rods are about 10 mm in diameter and the M4 washers in Bag #6 were about 9 mm. There was no overlap to cause the smooth rods to get anchored into the plastic plates.
• I think I'm going to close this one. Even though I sent pictures of the eBay auction showing the rod extending out from each side (And their very own instruction manual shows the same thing) I'm being told that was the "Old design" and the pictures have not been updated. Certainly, it can be argued that what was shipped was different than what I bought. But I don't really care because I'm going to put a length of All-Thread in place of the rod anyway. By doing that I can put big fender washers on each side of the Filament Support Risers and tighten things down so the spacing at the top can not wiggle. I'm hoping that will add a lot of strength and keep the bottom of the Support Risers from cracking. This is the original text: Wrong length rod for the filament rolls was shipped with the G2s kit. The length of rod shipped seems to be the correct one for the single filament machines. But the dual filament machines require more length because it needs to project out the sides of the support piece enough that a roll can be put on each side as documented in the first picture at: http://www.geeetech.com/wiki/images/...nstruction.pdf
• The carriage pieces did not allow the linear bearings to snap in correctly. Once the problem was realized, it was not a big deal to 'Post Process' the 3D Printed pieces to make the linear bearings snap fully into position. For somebody that has done some 3D Printing, this isn't a big deal if you realize it is an issue. For somebody that has never done any 3D Printing this is going to be a problem. Most likely they won't even realize that their machine's geometry is compromised. Some details in the instructions would be helpful to avoid sending people into the ditch here. Or better yet... How about Geeetech using their dual filament machines to print these parts using PVA water dissoluble support material? That is the whole reason I wanted a dual filament machine in the first place. If they "ate their own dog food" these pieces would have come out pristine. But instead, the bridging where there was no support came out flawed and made it impossible to snap the linear bearings into position.
• Thinking I needed some help I ended up at: http://www.geeetech.com/contact_us.html Scrolling down, I find this text with the 'please contact:' section left blank. Perhaps I'm being cynical, but leaving the contact information blank in the part of the page to address complaints does seem a little bit strange! This is the kind of stuff that happens when some senior executive mandates "We need less complaints! Your performance will be evaluated based on the number of complaints we get!" So instead of spending time trying to figure out where to send a complaint, I wrote this page and settled for sending an email to Technical Support pointing to this page.

Make Complaints

If you have any suggestions or complaints to our products or service, please contact:
We take customers as priority and hope your every shopping experience rewarding. If you've any queries, please email our team and we'll get back to you as soon as we can!

• There are four micro-switches used in the kit. The micro-switches they ship seem to be very good quality. One is used on each tower and the last one is used for the Auto Bed Leveling probe. The problem is most people are not going to succeed with mounting the switches where they need to be. The instructions say to use 3 mm screws to mount the switches and the mounting holes in the switches I got are 2.4 mm. There is no way to even start the screw threading through the switch. I used a Dremel tool with a small drill bit to expand the size of the holes. On the last switch (for the Auto Bed Leveling probe) it would seem I did not expand the holes enough. As I was turning the screws to tighten it to the Spider Ring the switch cracked apart. It is very clear that what ever screws and switches were used to prototype this design are not what they are shipping. If over sized screws are going to be used, how about using 2.5 mm screws instead of 3.0 mm screws ??? It would still be too tight of a fit but the customer MIGHT have a chance of getting the parts mounted without destroying the switches.
• This one is resolved. The mounting pieces only fit together one way and I didn't realize it was changing the spacing on the Z-Probe. I 'Post Processed' the pieces until they would fit together nicely and now they assemble correctly. I damaged my Platform piece trying to work through these issues and need to print a new one. I am waiting for the source code so we can take this one off the list. This is the original text: The Allen Key piece is too short. With it fully extended it is above the tip of the nozzles. According to Configuration.h the Z Probe switch should get triggered at -1.2 mm below the nozzle tips. That is clearly not the case. It can not extend that far. Some unsuspecting customer is going to destroy their nozzles because of this. It is not yet clear how I'm going to fix this issue. A longer Allen Key would eliminate that issue, but the key already wiggles a lot in the guide because it extends so far from the guide. I'm worried about repeat-ability of the measurements with all the slop in the mechanism. How did this happen? My guess is it has something to do with the fact I ordered the printer setup for 3mm filament with .3mm nozzles. I'm wondering if these nozzles are a little bit longer than the standard nozzles. I don't know, but this kind of screw up should not happen. But then again, it is kind of related to the next Bullet Point:
• They sent two small pieces of filament with the kit. I think this was for the purpose of helping the kit builder get everything setup and check out all the functions. They sent enough filament to make sure the nozzles are heating up and that the extruder stepper motors can push the filament through the nozzle. But, like mentioned in the previous Bullet Point, I ordered the printer set up for 3mm filament. They sent 1.75mm filament. Nice touch, total fail!
• On about page 49 of the Assembly Instructions, it has these pictures showing how to assemble the rods into the carriage pieces that hold the spider legs. They really should clean up the manual because they have stuff drastically out of sequence. The End stop screws really ought to be added to the carriage pieces first. But ignoring that issue, what they say here flat out does not work. I assembled everything and it will not produce a usable printer. Several more trips to the hardware store were required to figure things out. Please take a look at this picture from their assembly instructions:

Important_M6_washer.jpg Important_M6_washer_picture_2.jpg

I believe what is supposed to happen is the M6 washer is placed next to the plastic to keep the flared part of the metal rod from pulling into the plastic and delaminating it. It is also there to accurately set the spacing of the Diagonal Rods because this affects the machines Kinematics. This is some what confirmed by the previous (out of order) pictures where they show the Z-Probe switch being installed into the Effector. Please notice the inconsistency. If you put things together like they say "is very important" the bearings will not be able to pivot the full amount and your print area will be around 80 mm to 100 mm square. You very possibly will destroy some plastic pieces if you try to move the Effector more than this. In the picture above you can see the Diagonal Rod is just slightly at an angle. Even so, with this small angle, it is almost colliding with the M6 washer in the picture. This setup will not give you the desired results.

• On page 63 of the Building Instructions, it says: For G2s, connect the second Z motor here. And it shows a wiring harness being plugged in. First, there is not an extra wiring harness supplied to do that. But on a Delta Printer, there is not an extra Z motor. Come on guys. You deleted the extra wiring harness from the Bill of Materials. Can't you delete a little bit of text from the Building Instructions so your customers don't get confused by non-sense directions.
• Fan problems! The directions for connecting up the fans start on page 73 of the Build Instructions. Unfortunately, they seem to be aimed at a single extruder machine with a cooling fan for the plastic that the nozzle is laying down onto the part. The G2s has two extruders and each one has a fan. There is no cooling fan for the plastic coming out of the nozzle. The G2s does have a fan for the nozzles and it also has a fan to cool the control logic for a total of 4 fans. The wiring harness for the fan that cools the nozzles has the wrong connector on it. It has a two pin connector and it needs to be a three pin connector. They do provide instructions to use a two pin connector but they did not put that kind of connector on the end of the wiring harness. I need to give it a little bit of thought but I can splice into the +12 volt and Ground of the fan that is cooling the control logic to power the nozzle fan. I'll have to run to the hardware store to get suitable connectors and then break out the soldering iron.
• At the point where you are told to connect up the LCD Panel there is bad documentation. At Page 76 of the Build Instructions it says: "There are two cables, one is for LCD encoder, the other is for SD card, do not connect them reversed." The GT2560 Controller Board is clearly labeled. It is obvious which connector is for the SD Memory card and which connector is for the LCD Encoder. Unfortunately, this is not the case on the LCD2004 board. One connector is labeled EXP1 and EXP2. Searching around the internet I find this thread on Geeetech's forum: http://www.geeetech.com/forum/viewtopic.php?f=23&t=16742&p=26738&hilit=LCD2004#p 26738 The answer given in the forum is: "LCD is connexted to exp1 and sd is to exp2." The next post asks: "BUT THE INSTRUCTIONS FOR MINIG2 SAY EXACTLY OPPOSITE - "Connect SDcard to EXP1 and LCD to EXP2". ONE MONTH LATER AND NOT CORRECTED So everyone carefully connects them up wrong and has to find out why it does not work." Come on Geeetech!!! Add a couple of lines to the G2s build instructions and tell people which cable goes where! Telling us not to reverse the connection for the LCD and SD card is really not helpful when you won't tell us the correct way to connect things. And apparently this problem has been going on for quite some time and been causing complaints in your very own forum. Do you care about your customer??
• On Page 78 of the Build Instructions you are told to connect up the Power Supply. The power supply I recieved with the kit does not match the one in the pictures. They show a power supply with 7 screw down connectors. I received a power supply with 9 screw down connectors. The power supply I got apparently has an extra +12 volt and Ground pair. I was able to look through the schematics of the GT2560 controller and figure things out. If I had connected the wires as shown in the picture I probably would have fried all of my electronics. Come on Geeetech!!! How hard is it to explicitly state that the Red wires need to be connected up to +12 Volt rails and the Black wires need to be connected to Common Grounds? And remember, the customer is going to be wiring up 120 VAC power lines to this supply too. Wouldn't a little more text make sense here, especially if you are going to randomly ship different parts than what your Build Instructions.
• There is a package with 6 little heat sinks and some double sided sticky thermal compound. It is clear to somebody that knows about electronics this is supposed to go on the chips that drive the stepper motors. But there are no mention or instructions about what to do with these in the Build Instruction
• On the G2s there are two nozzles. On my machine one of the nozzles is about 1mm higher than the other. I suspect every two nozzle machine runs into this type of problem. There are no instructions on how to get the two nozzles level with each other. Obviously, if they are not level you won't be able to use two colors or one color and support material. Worse yet, depending upon which nozzle you are using, the other nozzle might be plowing through the plastic that the other nozzle put down. Most likely ruining the print but also very possibly destroying the nozzles and Effector. There seems to be no instructions on how to deal with this issue. I'm going to try changing the length of the diagonal rods such that the nozzles end up at the same height. It is very possible this will affect the machine's geometry so I'm not sure that is the correct way to address the issue.
• I am currently waiting for the replacement plastic pieces. But I did do a 'Smoke Test'. I kind of ducked under the table and flipped the power switch. I didn't smell anything bad and I didn't hear any sparks or craziness happening. I peeked up on top of the table and saw the LCD Panel lit up. But no text or graphics on it. I kind of started to mutter a curse under my breath and Surprise Surprise!!!! I had text on my LCD Screen. Apparently it takes a 4 or 5 seconds for the Aurdino board to wake up start doing its things. I told it to start pre-heating for PLA and watched the LCD Panel. I'm not perfectly sure I have the right thermistor paired with the right extruder but both extruders and the bed started pre-heating. SCORE!!!
• Power supply failed. I still haven't printed anything but I've been trying to calibrate the machine. I have a couple of high wattage ATX power supplies laying around. I'm just going to use one of those instead of trying to get Geeetech to replace it.
• With new power supply in place and functioning, the nozzles can not reach 230 degrees C. I get a Thermal Runaway or Heating Failure every time. I have put a volt meter on the 12 volt lines going to the GT2560 board. They are well over 11 volts with the bed and both nozzles turned on. I guess I'll see if it is possible to get just one nozzle up to 230 C. with nothing else turned on. With some more investigation I find the Geeetech firmware has turned off all the thermal protection in the Configuration.adv.h file. The default is to have this turned on as it is very important. Why did they turn it off? Probably the answer is the GT2560 board isn't strong enough to drive the heaters without tripping the safety shutdown logic. I'll try relaxing the parameters and see if I can have it turned on.

[3DPBuser]

Those are the plastic problems I was afraid of. And if I didn't break it to start, I was scared that the heatbed would warp with the heat. And lastly, that the whole frame would shift over time, especially when moved from room to room. So the steel route was better for me; completely eliminates these problems, and I expect it to be a good frame for many years:
http://3dprintboard.com/showthread.p...eatec-in-Spain

[ezrec]

If you have any workaround for the G2/G2s, please add them to the Wiki page I'm maintaining on reprap.org:

http://reprap.org/wiki/Delta_Rostock_mini_G2s

I've included fixes for a bunch of mechanical issues, including:

Mechanical Rigidity:

* Ensure the frame is rigid - if not, add washers to the smooth rods
* If you have a "wiggly spider", buy or print out the shim washers for the rod-end holders.
Frame
* If you can't finger-tighten the 4mm bolts into the smooth rods, don't force them - the blind thread may be too short. Use washers to shim up the bolt instead.

Belt Jams:

* Make sure both set screws in the driving pulley are secured.
* Make sure the idler pulley wingnuts are fully secured.
* Add idler pulley washers.

Homing Issues:

* Verify that you are on the correct pins for your endstops (see end of manual for correct pinout - earlier photos in the manual are incorrect!)
* Separate and twist the endstop wires to prevent false homing triggers

Bed Levelling

* Put lock washers under the bed wingnuts, so that you can adjust the bed from the top using the allen bolts.
* Ignore the autolevel feature for now - the printed guide for the probe can be too tight (or loose), and cause a head crash on autolevel.

PEEK J-Head

* Verify that the teflon guide is the correct length (if your PEEK cylinder is touching your heating block, it's too short!)
* Heat-tighten the PEEK heads to prevent leaks
* G2s: Align the nozzles to prevent interference during prints

Extruder:

* If you are getting under-extrusion (unexpected random holes in your print), make sure the Mk-8 extruder's tension is set correctly.

PLA Specific:

* Put a fan next to the bed to improving bridging and prevent blobbiness

ABS Specific:

* Use the correct Beta (3950) for your thermistor, and PID tune it (someone with Marlin experience can expand on this) - otherwise the PEEK heads will melt with the default firmware at 240C
* Add insulation under the heat bed to prevent heat loss

[Roxy]

* Ignore the autolevel feature for now - the printed guide for the probe can be too tight (or loose), and cause a head crash on autolevel.

I can certainly see why you say that. There are a number of issues with the Auto Bed Leveling. First, when I gave it a G29 command it moved halfway down and then started moving super slow in a sideways motion. I was wondering what in the world was going on. Looking at the code, I think I understand why that is. Who ever warmed over Marlin for this printer needed to provide a way for somebody to reach in and manually let down the Allen Key Z-Probe.

But come on people... This is totally lame. It would be just as easy to use the LCD Panel's switch to say the probe is down (and at the end, back up) rather than doing something totally Mickey Mouse like that. And as far as the probe goes, I ended up redesigning the Spider because the Allen Key is too far from the switch. The switch starts to exert a torque on the Allen Key and makes it bind in the hole that it goes through. Instead, I set it up with a piece of stiff coat hanger wire immediately next to the micro-switch. That seems to be much better. But once again, there is no way to know because they used such an old version of Marlin the M48 command is not in it. I'll know soon what the repeat-ability of my setup is. I'll probably add the wait for the LCD Panel switch in the G29 at the same time.

Here is the modified Effector. It seems to work well. But I've been looking at mounting a servo on it so I can kick down a probe leg instead of this manual stuff. I think I can use either the X-Min or Y-Min connector to drive the servo.

UPDATE:
I pretty much have the G29 working the way I want it. But I need to cross over the Topographical map stuff. And the M48 is going to be harder to get going than I realized. On Delta printers the X, Y & Z axis don't always refer to the Cartesian coordinates. I have to rework the M48 so Delta's don't go crazy. I suspect you are going to want the updated G29 when I get it done.

[ezrec]

Nice variant of the spider. How is the 'corner anchor' used? It's not clear from the .scad (I usually have a 'ghost' version in the scad all the components, and my build process makes 'bed-ready' versions).

I have a pile of mods for Geeetech G2 in SCAD format, if you are interested:

https://github.com/ezrec/geeetech-g2-mods

[Roxy]

The corner anchor is just positioned on the grid at the right spot in Slic3r. You would replicate it 6 times if you are going to use it and put one under each leg. Once the part is printed and off of the glass, you just take a razor blade and slice it off.

I'm going to print up your rod_end_shims(). I have a lot of wiggle on my effector.

[Forrest]

I just bought one of these without reading all this first ... put it together.
Seems problems with plastic separating are solved, all fit together well and strongly.
I do have the new "pro" version with metal spider and all metal for moving parts.
Printed PLA no problem after I remembered to change power supply switch from 220 to 110 v.
The is my first try with Kossel Delta printer, and seems to be working OK if I follow setup procedures closely, but I am having problems getting ABS to stick to heated glass with hairspray, mostly so far seems to be issue with keeping the tight clearance needed for first layer, although could also be my cold garage ....
This is my 3rd 3d printer, and in general quality of kit and instructions are very good, but not yet sure Auto Bed Leveling is going to work well as spring is a bit stiff to work with the mechanical limit switch, so I plan to change this to optical switch later this week.
All in all, I think this can now be recommended as a reasonable priced good set of parts and instructions.
I ordered mine direct, paid $100 for DHL and had it in hand in less than one week.

[Bobby Lin]

I heard they have some nice printers, but their customer support is just horrible. So no. I would stay away from this company and never do any transactions with them.

[DennisH.]

Just put one of the M201 "2 in 1 out" printers together. ALL of the acrylic parts fit perfectly. None of them cracked or got damaged during assembly. Two rods slip into metal mounts. The holes were too small. In one case it as just the thickness of the paint and once scraped off the rod fit fine. The other one was that the rod was 8.04 mm in diameter. That rod was also bent more than I though acceptable. New rods are being shipped at this time. Customer service was good. Once they understood how I measured the bend and which rod it was they shipped.

There were a few minor issues that made this kit not "perfect" out of the box. 2 of the limit switches could not be tripped by the carriage with out extensions being made and installed. One of the hardware packages was mislabeled in the bom. A couple of the electrical connections were described in the instructions, but were easy to figure out.

I'm glad I bought the printer and if I had it to do over I would again.