polar

I did my tests with the thermodisc 16A 104C. As they are cylindircal I machined a small aluminium part to have a good heat conductivity. To fix it to the heated bed I used on of the 3mm holes at the corners. Between the bed and the alu part I put some thermal paste. Around the fuse a little Kapton tape as the fuse-casing is electrically conductive and "connected" to the power (dont know why, found this rather strange). Between alu and Kapton again some thermal paste, cant hurt I blew several fuses on purpose. I monitored the temperature given by the thermistor and also measured the temp as close to the fuse as possible with a thermocouple. Cutoff themp was always between 104 and 106C at the fuse, which corresponded to a bed temp of 130C, starting form a 70C bed temp (which was is like 65-68 at the fuse). My bed heats up quite quickly due to the 320W PSU. This explains the large difference between the center temperature and the temperature at the fuse, it takes some time for the heat to heat up the aluminium part and "flow" to the fuse All in all it worked perfectly and I consider it to be a good safety system. (I am not happy with the cable routing, but I will change that soon. Also another connector, 90degree or something totally different)

@tsp42: When I google SEFUSE SF139E it shows up with 10A? But never mind, I will try THERMODISC 16A 250V, 121C. The manufacturer writes the following in the datasheet: Direct Current (DC) Applications MICROTEMP® thermal cutoffs do not have published electrical ratings for direct current (DC) applications. Current interruption capacity in DC circuits is highly application sensitive. Therm-O-Disc recommends thorough testing of DC electrical applications using the testing guidelines in Therm-O-Disc’s MICROTEMP® thermal cutoff technical information section. So I will do some testing.

@tsp42 I was told that under DC there is the possibility that you get like an electrical "ark" that is not estinguished. for the elcut the amps increase, for the setfuse they dont . But be that as it may, your setfuse r series are rated for 50VDC, so this is ok for me. Btw, the elcut 22 I also found but could not find a store to purchase them. Now it seems to be the same which the setfuse or do you have a store?

@tsp42: DC and fuses seems quite difficult, look at how they have to design DC fuses, often springloaded and other tricks. the "simply melts" is not working as far as I was told. @Johnny+Martin: As I wrote, the one from Raychem triggers at too high temperature. I was aiming at 110-130C for triggering (which again is very simple to find for AC but not DC). Maybe I just buy a couple of them and do some testruns.

@tsp42: this is a AC fuse, no guaranteed to work on DC, as I wrote a few posts earlier. AC fuses are easy to find, DC I am still looking for a good one. http://il.mouser.com/ProductDetail/TE-Connectivity-Raychem/RTP200R060SA-2/?qs=n8MrIBhnc4RacQBDqxfQ7Q== This would be ok as far as DC, amps are concerned, but it goes at at too high temperature... still looking

nono, not 240VAC, I don't like that near a metalplate I have to touch sometimes ... would be much lower like 20VAC. throwing the relays away and use mosfets is a good idea, however this will only solve the "stuck relay" problem, but not the "failed thermistor" or a potential the "software" problem

maybe a stupid question, but why did we all go for DC to heat the bed? ok, DC PSUs are easier to find but using AC the relay would not "weld" so easily and tons of thermal fuses would be available. Only negative effect I can think of is EMI, but again I am not an electrics/electronics guy, maybe there is a very obvious reason to go for DC

a main switch for the psu I already have, but this is of course only useful when the printer is not printing the LM27 looks interesting, but as I am not an electronics guy I would rather go for a pure "mechanical" soultion like a bimetall-switch that switches directly the 24V. I found switches in an acceptable temperature range e.g. from ESKA, also rated at at large enough currents, but this rating is only specified for 120/240 VAC. For DC it would probably not work, the switch would not "separate" but rather "stick" ?

hi I have the 2mm JasonHK heated bed up and running, so far no problems. Before I used a heated bed I often let the printer print large model (10h or more printing time) unattended, I wasn't even at home. Haven't done it yet with the heated bed, I don't like the idea of having a 250W heater that depends on a relay and the readout of a single thermistor and let it unattended. So I would like to add like a thermal fuse or thermal switch. Anybody else done that? Other ideas? Or am I too paranoid (got an aeronautical background, used to have redundancy ).

on sunday I wanted to print a small wing (using an 13.x cura version) and discovered the noise in z direction (z being the span direction), and the trailing edge was kinda jagged. Not so nice, already thought of using netfabb as this prints in z perfectly and a nice sharp and regular trailing edge but on the other hand seems to do some decimation in xy, resulting in a airfoil which is not totally smooth (and nf has some additional issues) But then I found (http://umforum.ultimaker.com/index.php?/topic/4601-cura-toleranceaccuracy/) discussing exactly this and your accuracy fix in 14.03, very happy I downloaded it and sliced (ok it took longer than with 13.x but who cares, still fast enough). then I started printing and the printer sounded like it's going to die at the leadingedge and at some small 4mm diameter holes in the wing, After finishing the print the z direction was much better than in 13.x but still not as perfect as in netfabb, trailing edge was nice. Now the 14.03_test sliced code is running, the speed issue is gone! Thanks will comment on overall quality after is is finished, but in the graphic preview it looked fine.

ok, now the file is is fine will test it....

hi Daid the beta exe seems to be corrupted, gives error before installing stating that the integrity check has failed. can you re-upload please?

It seems that they finally removed the ":" in the comments. This is good news to me as I print with pronterface which did not like the ":". So I had to remove them before printing.

hmm there is the level of detail, however this is only for the display, not the slicing as it seems

I'm not sure if I understood the idea behind this option correctly but I thought this means that the z is continuously increased and the extrusion always on. So one would get something like a helix. As I tried to print a cylinder I got a "seam" over the whole height. The reason for this can be seen in the code below. E.g. it increased up to Z8.3, then stopped the extrusion, went down to 8.225 and began extruding again. Is this the way it is supposed to happen? G1 X93.603 Y96.483 Z8.294 E0.4079 G1 X94.094 Y95.709 Z8.297 E0.4121 G1 X94.679 Y95.003 Z8.3 E0.4163 ;TYPE:FILL G92 E0 ;LAYER:54 ;TYPE:WALL-OUTER G1 X94.679 Y95.003 Z8.225 F9000.0 G1 X95.347 Y94.376 Z8.228 F1834.0086 E0.0042 G1 X96.089 Y93.837 Z8.231 E0.0083

One particular problem I have with netfabb is that the "polygonal resolution" of the slices is quite low. Meaning that when I slice e.g. cylinder the resulting print has visible edges .The model itself however has a very high resolution. Sliced with cura I get a much rounder result (but other problems) Any ideas how to increase the xy resolution in netfabb slices?