Big Mama Cetus in Heat Again!
Tiertime - Cetus sent me (gratis) repair parts for my heated bed cable fire. It is a new accessory board kit which includes the accessory controller board and a new 20 conductor ribbon cable as well as the other kit parts like the white printed case and a power feed cord for the main controller board. All the parts in the standard purchase kit.
It was the 20-conductor ribbon cable that melted and burned. The end of the ribbon cable that connected to the accessory board showed evidence that it too had overheated. Indicating there was probable damage to the ribbon connector device on the accessory board. (large white horizontal bar left beow center in picture)
It was very good they sent both the board and the cables. I assume they are part of the standard accessory board kit.
I have replaced the parts and again operating Cetus with a heated bed.
There was a slight issue when I tried the first test print with the heated bed. After UP Studio sliced the print and sent it to Cetus, there was a small audible “clink”, Cetus stopped dead and UP Studio reported that Cetus lost connection. Very strange behavior. This behavior continued on three more attempts to run the test print.
Re-booting everything did not fix the issue. Then I used the manual pre-heat to test heating the bed. It heated the Cetus build platform exactly as expected. Then I went back to trying the test print again. This time it continued past the “die” point and the test print completed perfectly.
I believe the Cetus Processor does some very strange “caching” of its previous experience and/or setup. Its’s not been the first time I have seen the Cetus printer OS acting strangely. I believe it tends to use cached data without checking if a new “read” from UP Studio is available.
I usually do a “clear the SD” within printer setup and I know I didn’t do that this time. Probably the cause of the crash. I’ll remember next time.
Big MAMA Cetus is again a hot babe. Fingers crossed she doesn’t get too hot!
Three-dimensional printing can be far more than printing one-piece Junque trinkets and plastic dragon skulls. <grin> Here, I go head-to-head with conventional CNC subtractive manufacturing.
I use a CAD/CNC program call Vectric Aspire. Wonderful application for 3D style 3 and 4 axis design and production using overhead routers and milling machines. This project doesn’t use Vectric Aspire. Rather the user forum for this product generated a 3D print design idea.
A fellow in the “Vectric Design Lab” named Todd created this desk lamp. He uses “subtractive manufacturing” (CNC router) to create his design. I took one look and decided it would make a good “additive manufacturing” project for 3D printing.
If you open the video from Todd, the first part his microphone is off. Don’t be alarmed about no sound. He figures it out…
The video is rather long winded for non-Vectric users (and me) but skip towards the end and you can see the finished design. He says to modify and build it any way you want. So, I did…
My first version is an exact copy of Todd’s design. I had to dig deep in his Vectric carve files for dimensions. He uses 2D and 2.5D routing so there are no 3D parts. I redrew all parts from scratch in Autodesk Fusion360 (F360) to create 3D models of the components.
I had to experiment a lot in F360 to get good 7/8-9 thread design to print properly. I made the (prototype) screw knobs round like the original in RED PLA. Rather ugly. I called it my clown lamp. I later did three more sets of “bolts” in white with my new original design and using the very nice working 7/8-9 thread I developed in F360. PLA required a 0.040 gap between 1A and 1B dimensions. Thread design in F360 is a whole other topic for another day.
My STL files have the correct threads.
The lamp uses a battery “puck” lamp so there is no line voltage wiring. If you are interested in this design, you can decide alternative electrics if you desire. The head opening is 60mm and uses the “AAA” battery puck lamps of that size. Source Amazon.com.
The lamp is larger than it first looked to me. All parts were printed on the Cetus build platform, so a large printer is not required. Longest part is 6 inches (152.4 mm) The original lamp head nearly maxed out the build height on the Cetus, but the new design should not be an issue for any printer capable of printing the base or arms.
I built the original “Todd” design in white PLA with the red bolt connections. The color combination did not “grow” on me. I was re-designing the lamp and decided the white Ver.1 lamp needed the new design “knobs” in white. I now call it Ver.1.1.
Lamp Ver.2.0 (in black) removes a lot of material from Ver.1.1. I think it is more suitable for 3D additive manufacturing. The open jointed arm is plenty strong for the application. Version one was intended to be a simple project routed from solid wood stock. Todd recommends builders modify his initial design.
There can be many ongoing variations from my Version 2. The limitation question is: “How may battery lamps does one need?” I am good at two and hope to give them to someone. I can always print more if required. No need to build an inventory.
This lamp is a durable product. Not a minimalist material, lightweight display piece. I don’t try to see how much material I can save when printing.
Printing Specifications
I print at 0.20 mm layer height with (minimum) 3 layers top and bottom. Four would be better. Sidewalls are 0.50 mm print width with a 0.40 mm nozzle. The threaded connector bolts and nuts are printed with five (5) shells (layers) 2.5 mm (total) thick sidewalls. This is to produce good solid threads.
All other parts are three (3) sidewall layers (1.5 mm, total) thick.
Infill on arm components is 50% - 60% for added strength. Base and lamp head are printed 10% -15% fill.
PLA plastic is more than adequate strength and prints nice and flat. Feel free to experiment. Exotic materials like carbon fiber are not required but could be a printing brag point… Ha!
Other Design and Application Ideas
The head with puck light design can be used with a smaller base with no arm links. Just the two short stubs. Could be a good accent light for seasonal decorations. The puck lights can be purchased with colored LEDs and will produce many colors and variations as well as remote control.
The pucks are also made in various diameters. The larger use AA batteries for longer life. There are also low voltage A/C wired puck lights for more permanent lighting use. Variations on this theme are many.
 |
 |
| Version 1, Exact copy of CNC version. Gaudy red knobs! | All the printed parts. Red knobs replaced next day. |
 |
 |
| Original head maxes out Cetus print height. | Base maxes out Cetus build plate at 6 inches dia.. |
 |
 |
| Version 1.1 with improved joint knobs | Version 2.0 Nice improvement IMHO! |
The new design (Ver 2.) looks less like assembled from 2x4 lunber. I am happy with both. Puck lights can have other uses. Some of them can randomly change color and would be great for holiday decorative lighting. A base and head (no arm) could be used for decorative up-lighting. Many ways to use the lights and 3D printing.
20 conductor Cetus Mk3 heat bed cable with vaporized center trace.Big Mama Cetus blew her cool again right in the middle of a normal print. This Hot Mama is going to be sent to a quite religious Convent where she can live the life of a reformed sinner… with no more hot flashes.
Good thing I was sitting right in front of the printer, when about halfway through a normal operating print, the heated bed ribbon cable suddenly burst into a large puff of smoke. This is the second time, so I knew exactly what was happening. I immediately lunged for the “power off” switch on the side of the printer.
A power trace in the middle of the ribbon cable had again (2nd time) vaporized the plastic coating.
I took micro-photos of the cable ends. The problem originated in the controller board end of the cable and NOT at the heated bed. The picture shows a group of “cooked” contact points.
The cable connectors and the cable itself is NOT CAPABLE of sustaining the load of the high power that it is required to pass. While multiple traces in parallel are used to distribute the power load in the cable, it is clearly evident the tiny terminal contact points are not intended for such power loading long term. In my case rather short term.
If I had not been present when this happened, I could have easily suffered a tragic fire and perhaps severe loss of property and worse case, life.
The Tiertime Cetus III Heat bed system is a serious defect in design and engineering and all users should disconnect the ribbon cable immediately and cease heated bed operations.
An outstanding minimalist printer has become a possible life threatening monster with the upgrade to the high power heated build plate. IT IS DANGEROUS!
I will continue to use this printer without the heated bed ribbon cable. All other functions and control have tested normal, and no collateral damage to the printer has been detected.
I have suffered two (2) exact same cable failures on the Cetus Mk3 high power heated print bed. There will not be a third.
 |
 |
| Heat bed end, no apparent damage. | Controller end. Severly overheated connection traces. |
UPDATE 8/26/2020:
Tiertime (Cetus) has contacted me. They will replace the damaged controller board and failed 20 conductor cable under warranty. This will be the third board and cable. This repairs the damage. Not sure that it will cure the problem. Best they can do for now. Thank-you Tiertime.
I have to take Cetus MK3 off my recommended 3D printer list. At lease until it gets some improved firmware The Mk1 and MK2 are a great bit of hardware Extremely accurate and capable movement, but still have some faults, like having no physical bed leveling.
The leveling, including the the electronic software leveling are not true bed leveling. The software leveling is simply for determining a nozzle height so an over-thick raft can be built level with head movement. Thicker in low areas, thiner in high areas. The plan is to compensate for a few tenths out of level. In practice the Cetus bed can be several millimeters off level.
This all-but-eliminates printing without a raft.
What amazes me is all the Tiertime printers use this same method. The more costly printers DO HAVE mechanical bed leveling, so the auto leveling really is a belt and suspenders feature. Extra measure but not really needed.
The Cetus MK3 ROM can not handle g-code like the MK2 and MK3. Trying to run g-code, even the gcode made by their own "Catfish” advanced slicer, will crash the bed temperature sensing algorithm in the MK3’s ROM. Temperature reading and control go spastic, even after returning to UPStudio.
It happened when I first got my MK3 and it Just happened again after I changed hardware to fix the first issue. This time I was able (so it seems) to have recovered the bed temperature control. I had to reset and refresh the ROM, Reload UPStudio, and let Cetus set for about a hour to cool down and it finally reset.
I tried so many things I can not be sure exactly what provided the “fix”. I don’t want to deliberately crash the OP system again just to find the simple fix, if there is one.
Cetus requires two major improvements to work properly beyond stock operation.
- Mechanical bed leveling
- Proper g-code operation especially with Catfish. Tiertime’s own advanced slicer
Other issues are the feeble flat cable feeding the heated bed. I have had two cable fails. The new perforated build plate is very difficult in raft removal and simply can not be used for no raft printing.
The heated bed is bolted directly to the X-Axis linear bearing. The possible 100C+ (212F) bed heat is transfered to the bearing, linear rail, and into the printer case. I expect short life for the bearing if constantly running a heated bed. (Bearing lubicant vaporization)
I have had "run-away" Build Surface heating up to 180C and the printer case got so hot it shiut down on thermal overload.
All major Cetus issues stem from the primitive build surface plate and it’s simple bolt-to-the-bearing mounting system. No problem with an unheated build surface. A heat-isolated mechanically adjustable bed is needed.
Cetus was/is an experiment in minimal hardware, so as not to compete with the major Tiertime product line. The MK1 popularity succeeded beyond expectation. However, the "advanced" Mk3 has become a Kludge. (Clik link if you do not know the meaning...)
I love the prints that Cetus will produce, but the MK3 machine can not (now) be recomended for the biginner.
Search
Random Pix
