Lair Of The Multimedia Guru

When i experimented with measuring the amount of particles from my 3d printer, i noticed the majority originated from when the heat-bed was turned on. I had cleaned the inside of the printer roughly before, so this was slightly odd. More investigation shortly later revealed the cause though.

The use of tinned multi-core wire without ferrules. Why it was done this way you may ask? Because between the incomplete and poor official documentation, the consumer made videos of how to assemble it and my plain lack of knowledge about how multi-core wire with and without solder behaves in this case. This way of connecting the wires seemed sound enough and there also where no ferrules included with the printer kit.
That said if you did connect high current wires in a similar fashion, you probably want to consider redoing it with ferrules or in another reliable way, like directly soldering to the board or soldering a solid core wire onto the multi-core as an intermediate. Also keep in mind iam a programmer not a expert for this so do your own research if failure of a connection could have some real consequences.
My printer is fine, the heat-bed also didn’t behave erratically or anything. And i never trust/trusted it anyway in terms of safety to ever run it without someone being at home, so in my case it was impossible for this to cause any real harm. But i still didn’t expect that …

I didn’t had any fitting ferrules, so the way i fixed it was to take some terminal connector thingies, which when one cuts half of each off are basically ferrules. I also didn’t had a new terminal block that was rated for enough amps to replace the roasted one. I will properly replace these once i have proper replacement and have time. Quite likely next time that printer needs some kind of “service”.

A while ago i read about 3D printers spewing out nano-particles into the air, especially ABS. I was curious if i can actually measure this on my own printer and how to fix it.
My plan was to use a cheap air quality detector, that i bought previously. It has to be noted here that the device detects particles 300nm and larger while what i read seems more about 100nm and smaller particles so its not 1:1 comparable.
Before anything i used my vacuum cleaner to very roughly remove large visible crap from the inside of the printer enclosure.

First baseline measurement, with the printer on, no heating yet.


Can we reduce that particles/dust ? Lets try properly cleaning the enclosure with the vacuum cleanerto use a stream of negative ions to ionize the particles, so it sticks to something and doesnt stay in the air. As can be seen this is a effective way to clean the air, or the measurement device is affected by the ions.


There is a slight disadvantage, in that it seems the well designed printer (which lacks grounded components). Has its metal frame charged up by the ions and that then eventually discharges into something …


Luckily no permanent damage is done and a reboot of the printer clears this. Until it quickly happens again. But after a few minutes all the particles appear to be eliminated


Adding a wire between ground and the frame (with a 1k resistor between) stops it from charging up and seems to resolve this problem.


Next after heating up the heatbed to 65°C without the ions. There is really alot of particles/dust in the air of the enclosure:


After also heating up the nozzle to 215°C with PLA in it:


After finishing the print 38 minutes later


Waiting 30min (the heatbed is not disabled by the printer, the nozzle cools down), there are still significant particles in the air


Switching the negative ion generator on, 5 minutes later the particles in the air have decreased significantly


Another 5 minutes later the particles are largely gone


Now question, what happens if we try to print with the thing on?
First opening the printers enclosure brings some of our particles/dust back (from the airflow caused by the door presumably). After all it didnt magically disappear and theres nothing that would truly capture it (the heatbed is at 65°C here)


After the nozzle heated back up to 215°C the particles actually decreased


Taking an image every minute from the air quality meter shows that within about 6 minutes after the print started the particles go to 0, it does rebound slightly and fluctuate around 0 after that.


The print finishes without problems after 36 minutes. Originally i intended to repeat this with other filaments but as the biggest dust/particle creator seems the heatbed or rather air convection from its heat. This would need to be setup differently to eliminate that. As it drowns out new particles from the heated filament. But still this shows that for less than 4 euro shipped one can get a device to eliminate at least larger particles from the air within the enclosure of a 3d printer. Compared to a filter this is probably easier to setup as well. Further experimentation would definitely make sense. Though simply having an enclosure and opening it together with a window for 5 minutes probably removes most particles, if one is concerned about it.

When i installed the vibration dampers a few months ago on my 3D printer. I wondered how long the metal to rubber connection would last. One or 2 days ago my extruder started to skip. After checking it and the hot-end for clogging, i found the vibration damper was no longer in good shape. about 50% of the rubber had separated from the metal.
Thus the lifetime of that one was about 2 months with low to moderate use.

Somewhat expecting this i already had a bag of replacements and replaced it. Lets see if this is a one time failure or if this repeats. Opening the arm of the extruder when not in use should reduce the stress on the part though. But until i see a repetition of this failure mode ill leave things as they are. So as to exclude that this is a one off faulty part.

Low cost 3D printers (in 2017) are “fun”, theres “always” something that needs adjusting or fixing.

The extruder of the tarantula is capable of printing flexible filaments. But its not good at that.
What worked best with the original extruder was to push the bowden tube up to the gear of the cold-end and to disable retracts.
With that it was possible to print stuff at 15mm/sec with some success. 30mm/sec did not work for me, at that speed the filament found a path of less resistance than the intended. That was with 0.4mm nozzle 0.2mm layer height.

To solve that, the cold-end of the extruder needs to be replaced. Looking at thingiverse i found several printable extruders that looked like they should be better. The one i printed and tried first was the RSE-2 extruder. Which worked a lot better but required hacks to make it fit. It needed a spacer, to not hit the x carriage, required a separate piece of smaller diameter PTFE tubing, M5 coupling instead of the M6 i had from the original extruder and was not really well mountable with the 2 screw based vibration damper i had.
So i modified the RSE-2 extruder to correct all the issues i did run into. Printed it, printed it again with correct settings ;). And replaced it.

And with that, my tarantula can print flex filament (TPE88 to be precise) at 50mm/sec at 0.2mm layer height, 0.4mm nozzle, Example print:

100mm/sec at 0.1mm layer height with 0.4 nozzle works too but makes not much sense as the higher speed adds too much artifacts and 100mm/sec at 0.2mm is beyond what the hotend of the tarantula can melt, so i did not try that. Also it seems retracts work fine though were disabled in the example shown above, the 100mm/sec test i did had retracts enabled.

The modified RSE2 extruder is available on thingiverse, if you want to print it. It works with MK8 drive gears as well as UM2 Knurled drive gears. The MK8 seems to work better of the 2 in terms of extrusion error per backpressure. and both provide 40-50% more force than the original tevo tarantula one with the same stepper and driver. Some cheap ones from aliexpress:

On the quest to improve and tinker with my 3d printer, one area to improve is noise. And a while ago i did buy a set of Stepper motor vibration dampers. The idea is quite obvious, the steppers produce noise and if they are rigidly attached to something solid and big theres alot of area that can radiate the noise. Adding a flexible damper between the source and the rest of the machine reduces the area that can radiate the sound and makes it less noisy.

The vibration dampers i got have two raised spots that have no matching holes in the 3d printers parts. They have to be removed or holes be added to where they are attached.

Filing them off by hand is very easy though, the metal seems quite soft.

Ive attached one on the X and one on the Y steppers, these 2 are the most active an noisiest steppers. The Z stepper moves so little that its inaudible except when the printer starts up and runs the auto homing sequence. The extruder stepper too is inaudible except during retracts. I plan to add a damper to it too when i take it apart for some other reason.

Testing with a microphone inside the enclosure printing a 20mm calibration cube before and after each damper addition

The difference after the 2 added dampers is quite noticeable. The top row is the original and bottom row is with the 2 dampers. The remaining noise is primarily from the hotend cooling fan and less so from the part cooling fan. I plan to replace the later with a noctua fan. The hotend fan is a 30mm fan though and 30mm fans seem rare, i am not sure from where to get a silent quality 30mm fan, noctua does not make one. I could use a 30 to 40mm adapter and different fan duct with a 2nd 40mm noctua fan though, have not decided yet.
Also it seems the addition of the dampers has slightly reduced the waviness of the sides on the printed cube. Its hard to say this definitive though as taking the stepper off and putting it back on changed things like belt tension which may have cause the difference too.

10 years ago, in 2007 3d printer prices broke through the 10000 USD barrier. Today in begin of 2017 you can get a 3d printer express delivered anywhere for less than 200 USD or € from aliexpress and probably ebay too. So obviously i had to get one :) iam not good at resisting “cheap” and cute tech.

The printer i bought was the tevo tarantula which at the time had the most orders in the price range i looked at on aliexpress. I got the version with a large print bed and auto leveling for about 250 €. I almost also got dual extrusion for free as the sloppy seller had a copy and paste error in the offer, sadly she insisted on not including that but she instead added some free replacement nozzles and more free filament.
There were not too many alternative options as i did not want a printer made from primarily acrylic plates (not strong yet brittle and not very hackable). One alternative would have been the flsun 3d i3 printer, my main reason i decided against that was its smaller build area and apparent smaller user base resulting in lack of reviews. I decided against delta style printers simply to keep the number of choices down and with the thought that i might buy a delta style printer at some point in the future.

The tarantula is primarly made of aluminum extrusions with a few acrylic plates. At the time when i bought it, it was shipped with DHL express for free. I was quite surprised about the express shipment and quickly found why, apparently express shipments appear to be subventioned in hong kong. Before receiving the printer i decided to buy a cheap set of random replacement parts for 3d printers, having read of quality issues and having experience with buying things from china ;).

The manual for assembling the printer starts quite good, the first page tells you how to put the base together suggesting to place the horizontal part in the middle, which is not optimal if you want to maximize the print area.

The 2nd page explains how to mount the y stepper motor so that it does not work with the included firmware, it needs to be mounted on the other side or the firmware changed and rebuild, also if you look carefull you will see that on the bottom there are 2 identical acrylic parts depicted, one serving no function and IIUC a left over from a older revission, they do include this part physically, its useless though.

A few pages in, the author apparently decided to stop annotating parts in the middle.

And 2 pages later all annotations are gone, and one has to use trial and error to find out that no included screw fits in some places.

Thanks to the community there are 2 video series on youtube though that explain how to put the printer together. Searching for “tevo tarantula assembly” finds these and might also find future ones. I recommand watching both and reading the manual before and while assembling, that way no mistakes should happen.

2 nuts included lacked threads entirely, one pulley was bent, the speculatively bought replacement parts included these luckily :)

The power supply works fine, sadly theres no real enclosure. The main wire terminals can be touched if the user puts his fingers on the power supply. Theres also no power switch, to switch it off you pull the plug from the wall. The fan of the power supply is quite noisy. Its output is 12V 25A (didnt test if its actually capable of sustained 25A thats just what the sticker says). I quickly replaced it by a random old ATX power supply which i had laying around, that way it has a nearly inaudible fan, no exposed mains voltage parts, a power switch and a few more voltages for free.

One of the first things i printed was a fan duct for part cooling, that improves print quality. I also covered the acrylic plate below the heated bed with aluminium tape as it appeared to soften from the heat and bend slightly requiring level adjustments between initial power on and when it had heated up. I also screwed the printer onto a piece of wood. That makes it independant of the flatness of the surface its standing on, it makes no difference in print quality though.

The LCD display as it comes with the printer isnt attached to anything and there seems no really great place to put it wthout extra parts, so i cut and drilled a piece of aluminum and attached it with that on the top. That location works. Slightly lower and in front of the al extrusion it would be in the way of the cables and tube from the print head if it moves higher up.

To better see what the printer is doing i added some 12V LED stripe with some kapton tape for isolation and stuck into the top alumiunium extrusion, it exactly fits. Powering this off the 12V line results in flicker as the heater elements being switched on and off cause +-1% voltage fluctuations. I thus stuck a cheap ebay voltage booster in line (12->13v) which fixed this until i powered it off and on, at which point the magic smoke left the voltage booster. Assuming that it didnt like the rather small boost of around 1 volt i took a new booster and this time added a fuse and fuse holder and powered it off the 5v line. After it worked fine over many hours and some aggressive power cycling i put it in a nicer box with ugly hot glue.

Ive also added a glass plate on top of the heated bed taken from a ultra cheap picture frame that was hanging on the wall, temperature cycling it to 130°C and printing a few things on it had no ill effects, no need for borosilicate glass here apparently.

And last ive build an enclosure for the printer out of wood and polycarbonate plates, some screws, silicon and 3d printed barrel hinges from thingyverse. I failed to find hinges in a shop locally that i liked. With these hinges one can also just simply remove the front door by just lifting it up. With the enclosure the print(er) wont get dusty, and is alot less noisy also it appears to have improved the results with ABS.

The door locking mechanism is build out of 3 times 2 screws and a ball neodymium magnet, and works very nicely.