- User Since
- Apr 14 2016, 1:35 PM (183 w, 2 d)
Dec 28 2017
@benmalouf does the part shown in the picture above suit you cosmetic standards?
@samantha I think this would require a note to be able to reject gears for broken teeth. I don't believe you could reject this part based on the call out for the diameter, will you look into this please?
Dec 27 2017
@tutley @kent @karrad It appears the reason for the filure is due to the block being located too low on the nozzle shaft. The air from the extrusion fan system cools the block and causes the system to be over-damped (PID values were found with the fan not running.). The nozzle has <0.5mm gap between the nozzle flange and block but needs to be >1.5mm. This means the block is one full revolution too low. for this test i lowered the end effector to 3mm off the heat bed and turned the fan on all the way. The target temperature was set to 200C so it is clear to see the block could not reach the target and timed out causing the error. I will continue testing solutions and values.
@kent as I'm sure you'll find out if aren't already aware, parts can change. Even parts you think for sure won't. I would have waited to create a beta branch myself, but it shouldn't hurt anything. I myself would hold off on pushing files to the beta branch for sake of ease. To me it seems like more work than necessary with no added benefit. Also, doing this doesn't conform to the agreed upon form.
Dec 26 2017
@kent @EricNugent I think we need to log the output from the thermistor, and overlay it with data from a thermocouple. Then we can tell if there is something wrong with the thermistor, and at what temp. We will be able to better determine a path forward with more data. I will look into a good way to automate this.
@kent The rate of change seems very sporadic. Have you looked in to the thermistor at all? possibly a bad thermistor or connection somewhere?
Dec 19 2017
@kent How is the testing going in the cluster on this machine?
@MikeR do we have another machine that we can drop? Or are all the petg machines in the cluster?
@kent Have you looked into this yet?
@benmalouf here is what the part would look like with less gap.
Dec 18 2017
That's ok. It is slightly more than the previous generations, but this does not impact the functionality of the machine to our knowledge. If you find a reason to eliminate this gap let us know it should be a simple fix.
Dec 14 2017
Seems as though the bearings in the petg parts fit especially loose in the bores.
@karrad the end gcode is in the same order as the gladiola. The only modification is the z height. would you be willing to try @nickp's gcode? I had a very different result than he did. My machine actually had multiple pauses throughout the print.
Dec 13 2017
@kent the supplier is patiently awaiting your response. Please respond soon to avoid any unnecessary delay in supply chain.
@Steven @west Testing shows a significant drop in sound level between 1600Hz and 3000Hz using the new blower shroud. It looks as though we need to work on some other frequencies to bring our overall noise levels down. I will run some tests to determine the source of the larger spikes and work to reduce them.
Dec 12 2017
@karrad I did not. but i will.
Dec 11 2017
@logan which parts do you need still?
@logan what is the status on getting one of these machines into the cluster?
@logan have you finished the modifications to EL-HR0114? If so, please close this ticket as resolved, and link your commit.
@kent X belt can now be reinstalled without removing the tool head. I'm closing this ticket as resolved because the new part resolves the problem described in the ticket. If you are having problems with the slice please create a new ticket.
Dec 7 2017
@logan the new firmware is working. and the fan modification discussed in other tickets does help. We have not yet tested this in the cluster though.
@logan we are unable to get the amount of the final colors at this time. Let's go ahead and fix one of the dropped machines and put it in the hottest spot in the cluster we can for testing. Like @tutley and @karrad said, the glass transition temperature is the main concern for this first trial. Please let me know what you need.
@MikeR which pieces came loose? we may need to look into how they are oriented.
@lansky It appears as though the printed extension mount isn't fully seated on the home switch. Also, it looks as though the nozzle has melted into it a few times. I would reprint the home switch mount if possible and see if sits flat on the switch.
Dec 6 2017
@logan Try this gcode and let me know if you have any problems.
Either we can make the overhang less drastic, or a break away would be fine.
@kent @logan If you want to change the part lets go ahead and do all the edges that fall under that parameter. I would hate to have the other holes crack as well after changing the ones the ticket was originally created for. Let me know when you have the ECR process started.
Dec 5 2017
@MikeR To my knowledge the hole is the same size as before. Were we having this problem before? if so, we should look into changing the hole size. If not I think a re-slice would be the best approach.
@MikeR Do you think we need to make a change to method of insertion or the printed part?
The drop test went well. The only broken part was the extruder mount.
@kent sounds good. make a recommendation based on what you are seeing in production. Also, when did the chip occur, and do you have pictures?
@DRobertson Will this work for you?
Dec 4 2017
@logan looks the reference came off of the chamfer diameter. should actually be the same now. lo siento...
Dec 1 2017
@kent I'm afraid you may still see cracking on the top side.
Adding thickness to this part will require revising the body plate printed part as well. If i were to add some thickness to the walls i would like to do so all the way around the part. The body plate would interfere doing this.
@kent this is less than 4% fall out. Do you still wish to make this change?
@kenhara During development there was a feature to prevent that exact thing. However in some cases the feature interfered with the tension mechanism not allowing the idler to close fully. The feature was removed because that can cause stripping out of the filament. Also, you can always unscrew the tension screw, open the latch and insert the filament. However, this would require a slightly smaller spacer to set tension with. We decided to model the tension spacer to set tension before filament is inserted because with varying durometers of filament it would easy to over tighten if tensioned after.
@kent what is the fall out rate?
@logan The tighter fit is ideal for locating the motor. Judging by the example you showed i would not change this tolerance. It is OK to use the screws to pull the motor into place provide the torque value is not exceeded. Please reopen the ticket if more investigation is needed.
@nathan please let me know if this helps. Please respond either way.
Nov 30 2017
@karrad Ran a test moving a carriage from the same location (100mm) +175mm with both step values. As expected the 100.5 step value overshot the target (275mm) nearly 1mm, while the 100 step value lined exactly. In my opinion, if we can see this difference on a machinist rule this is too much error to have present in any axis.
End stop bumper created.
@kent The assembly was done in accordance with standard techniques, however the sheet metal was modeled incorrectly which is why it does not line up correctly in the assembly. To my knowledge blender cannot set a coincident constraint. If you look in the old assembly you will notice the upper and lower plates are wider than the inside of the left and right flanges in which they are supposed to rest. Therefore these are not properly aligned. In order to align the left and right sides coincident constraints were applied to the front of the upper and lower plates. There will still be a 64um overlap on the back side until we are able to correct our solid model for each sheet metal piece, but everything lines up across the YZ plane now. The prints appear to be correct, but i couldn't find a call-out for the distance between the flanges on the left or right plates. Due to the nature of sheet metal production i feel we should be cautious if we decide to change the prints. If anything just a note in the rev block notating the change to the solid models as suggested by @samantha .
Nov 29 2017
@tutley, @samantha I'm going to have to rebuild the assembly. it looks as though the left and right pieces are too narrow to fit around the bottom plate in the assembly. I need to rebuild the assembly and verify the prints and then sheet metal parts.
@kent as we were just discussing downstairs, I'm currently working on this.
Nov 28 2017
@kent We had already drop tested a machine and there wasn't any slipping of the belts that i was told about. The only thing had happened was the shipping foam slipped around the rod mounts during the drop. We are correcting the printed part so the foam can't slip around it. (https://code.alephobjects.com/T1377)
Moved the x rod down 2.5mm. Found the clearance we needed.