I recently received a message from “Thomas” regarding my thoughts post-purchase on the Taz 5 since he was in the process of eliminations similar to what I shared earlier in the year. Like me, Thomas “Keep(s) coming back to Taz 5 as well”. Unfortunately when I respond to his email the server rejects it so I’m going to outline my thoughts here.
First I will briefly summarize my new printer wish list from my March blog post:
- Improved Extruder
- More accurate prints
- Larger Parts
- Open Source flexibility
- A degree of turnkey capabilities
So four months in, multiple spools of wasted material calibrating the printer and some good, bad and ugly prints… what do I think?
Let me suffice it to say, there is more to a 3d printer than the hot end… and the Taz 5 makes it abundantly clear. The unit is rock solid and the extruder is tight to say the least. The extruder gears have virtually no slippage and new All-Metal Hexagon hot end holds temps very close and I can tell the unit is capable of pretty good speeds and when you slow things down, very good prints. And though I haven’t done it yet, the X Axis mount is awesome so upgrading to a e3d or 1.75mm filament extruder would be possible.
If I were critical, the only constructive things I might add are:
The small extruder cooling fan to keep the heat creep down started developing some noise and I wind up flicking it some to get it to stop at times.
The larger hot end fan is great, but it would be nice if there was a solution that would cool two sides of the hot end since printing with PLA and ABS gets mushy when doing small details such as gears - this could possibly be eliminated with cooling on two sides.
More Accurate Prints
The frame of the Taz 5 is solid. Very stiff with lead screws that really seem to improve the accuracy given the hefty weight of the extruder. with the above extruder points and very tight build I am getting really solid prints. I dropped ABS early on because the open frame really limited my ability to do large prints without a ton of warp… however the PEI sheet did help a ton. However with PLA - night and day difference. Stick and warp isn’t an issue as much as eliminating ooze and getting retraction and perimeter width right.
The Lulzbot CURA and Slic3r profiles are great. They get you up and running quickly. Part fitting has been a challenge. I have spent hours and hours trying to get some standard calibration part fitting prints to work. I still have a ton of work to do, but I’m starting to get some better prints - but they don’t always fit. I have tried KisSlicer, Slic3r and CURA. I have landed on Slic3r for slicing and occasionally CURA when I need supports. The recommended Lulz edition of CURA doesn’t have enough control for my slice tweaking tendencies. Much of this may be additional learning curve I need to cover - albeit not for lack of trying.
UPDATE - after doing some heated bed cleaning per an email from Lulzbot support, I have not only gotten better adherence, but it seems to be level without slight dip in the middle of the bed (possibly in part due to removing the bed for the cleanup and putting it back). Could be a perfect bed leveling calibration storm, but I tried a .075 layer height print last night of a Kinect Scan I did of my son… the results in my opinion are remarkable. This was roughly a 12 hour print and the surface, without cleanup (I literally tore the support off), is satin smooth. NOTE: I sanded the bed with 10% isopropyl alcohol and 1200 grit sandpaper per Lulzbot support (they said 1200-2000) and my adhesion is insanely good again like the PEI sheet was new. I need to back it off some which might be as simple as loosing up the first layer height because even with a sharpened spatula it was way to hard to pull off jeopardizing my bed leveling efforts.
My first .075mm Layer Height Print:
Per my previous post, I am working on a 3d printed pinhole camera. The body and lid for the camera are the largest prints I had and once I slowed things down and really got a solid first layer… things were great.
Some recent issues: The past two days I started printing a sized down version Winged Victory of Samothrace with little luck. To this point I have been using Black Velleman PLA and really feel good having some decent .2mm layer height settings. I switched to White Velleman PLA and it seems to act completely different. Suddenly the first layer isn’t sticking, it oozes more and my first and latter layers are getting stuck to the hot end. Then when I felt like I had a decent profile… the print of the Winged Victory is stopping about half way in. I printed a smaller version successfully but when printing a 50% sized section it stops again. I realize as I type this is may be my Octopi setup causing it. Some troubleshooting ahead. One of the things I may try is going back to White PLA now that I have the bed cleaned up and level, use CURA for slicing and see if I can get a decent print.
Open Source Flexibility
A few things that are really positive about the Taz 5. Completely Open… I can use whatever software, filament and even hardware that I choose. I have currently upgraded my Taz to include an Octopi/Octoprint setup that allows me to wirelessly connect to the printer. And I am able to use any slicer I choose. I also have future plans for a 1.75mm filament version of the printer and am considering possibly building a all-metal hotend multi extruder (Kraken, Chimera or Cyclops maybe?) down the road… maybe down the highway if I keep biting off more than I can chew.
The down side… although Lulzbot does a great job of providing out of the box printing with open source softwares there is room for improvement. Though my Afinia 479 fell short in a number of areas, it did a great job of printing .1mm layer prints without fail immediately out of the box. I have had to do more tweaking to the Lulzbot to get there but I’m happy with my above mentioned .075 layer height efforts and feel confident enough to try .05 soon and see what happens.
Easy to setup, easy to print and definitely easy to stick prints. There is something to be desired in the way of auto-bed leveling in such an expensive printer but it isn’t too hard manually. Out of the box, I could get above average prints with the recommended software and settings…keeping in mind many of the provided profiles are really tweaked for max speed as well. Oh, and did I mention prints stick. And the all metal extruder can print almost anything.
What it lacks (that I was aware of and considered worth giving up prior to purchase):
- Auto Bed Leveling… again, wasn’t a deal breaker before making the purchase but worth adding if there were a Lulzbot edition upgrade
- Wireless connectivity… just seems to be where the trends are going so I’ve added octoprint/octopi which was cheap enough
- Some seriously tweaked profiles for various slicers to get high quality fitting parts. It still requires a ton of the user experimentation and though I scour the forums and other user feedback, the Taz 5 doesn’t really have this degree of part-fitting accuracy down in my opinion “out of the box”. Though I know it is capable from the work of users like Micheal Hackney is getting with his 3d printed fly reels and I’m getting really close to consistent settings on at least one filament.
I would recommend the Taz 5 to anyone that wants a large build area, high quality prints and the ability to add your own personal touches via open source components. They also have very responsive support via email.
My own personal wish list of add-ons in the future:
- Setting up my Flexy Dually printer with PVA support filament as a second extruder and using NinjaFlex and various other flexible filaments
- Building a 1.75mm filament extruder - useful for using my stockpile of 1.75mm filament and buying more filament locally
- Designing (or finding) a Taz 5 fan that surrounds the hot end for better cooling -really helpful for small parts, gears etc
- Redoing my octopi/raspberry pi case - the one I have is good, but not very permanent… I would also like a better location for the camera