youtube.com/watch?v=hNgehKRg-4g
PIRACER VELOCIRAPTOR
Checking finished part's fitment. Quality is not that great, but it was moving pretty fast. Flow was a bit too high.
youtube.com/watch?v=hNgehKRg-4g
youtube.com/watch?v=hNgehKRg-4g
updated 2 years ago
youtube.com/watch?v=hNgehKRg-4g
Just sending a message :) H.F.D.
Goofing around again. This is a challenge going around in the Mammoth server. 1st attempt at this and tweaking a little more to get the top right, then I am done with this.
I am gonna turn it up and see what it does next.
This is currently too fast for my printer to achieve the quality I want. But it was a challenge.. so I accepted.
Rolling shutter is bad, oh well.. it's moving pretty quick.
cults3d.com/en/3d-model/game/apex-cube
ABS. 500mms/50K/40SCV, 275C, .01ST, .006PA.
Having a retraction of nozzle diameter was causing a delay at layer change, enough to cause under cooling for the most part.
The seam width, under higher speeds/accels, seems to get wider. I have not yet figured out how to close it. You can see the bulge in front and after, as though the plastic is out of synk.
Is the filament being dragged or is it's mass giving it some velocity that does not follow the toolhead?
13 grams of filament in 7 minutes @ 100% size or 6.4 grams in 3:50 @ 75%.
500/50KA, 40SCV, ABS, 275C
I am going to try and print one with the UFO duct and see how much it cleans up.
I think this is back on track now. Changed to 625 bearings from 695 and re-shimmed the Y blocks.
I made some mistakes in the rail risers and the traminator found them for me. I only had to re-drill 52 holes.
Created one bad carriage and rail for the frame. This thing has never been right since I built it, gremlins!
Found another issue in my shed building skills with a tilted idler shaft. I think this is the main reason for my VFA.
If I can get some acceptable results printing, I will have the motor plates made by a cutting service. With all the revisions and what not, I may be able to finalize some of this gantry. Hand building these plates is tough to do without proper equipment.
Crossing fingers
Sub 7 min quality slice.
Speeds set at 500mms, 30-50K acceleration at 20SCV. PA 0.01 all features.
Standard benchy rules: .25 layer height, .5 line width, 3 bottom/top, 2 walls, 10% infill.
Beta V0.0.1. Had an idea and went with it. Glad I did. Although this is beta, don't be fooled! If you have a cross gantry or other, this is the tool for you!
I have some strange resonance I believe is coming from miss-alignment, and thus causing VFA.
Stay tuned :)
Total time for this is 5:40.
Sunlu ABS, 0.4TC nozzle, 275C print temp, 45% cooling speed.Cooing is too high for this print and or temp too low. UFO has some velocity!
This is using variable PA by feature. Smooth time .01
Walls/internal/external .007 - this value is too low
infill .001 - who cares haha
Top and solid .001 - too low
Strange patterns, not vertical lines, but swashes of some sort. Shows better when cube is still hot, reduces visual when cold.
I have not been speed printing lately. I have had many things come up in life these last 4 months... Micro Burst tornado like thing destroys fence and trees, projects for customers, ETC. So now getting back to seeing if this printer is any better than the last one.
Slated mods for this printer will be 12mm belts, Belay filament pusher, and also the Boombox bowden extruder. It's been slow to test these, but I will get there.
If I can get this bowden setup to work at high speed... it will be very good!
This is the "BoomBox" extruder, using two nema 17's with no reduction and idlers as filament clamp. Grip is on opposed sides of the filament! Filament abrasion from gears is minimal!
I am flow limited by extruder motor torque. I wanted to prove it by switching back to the worm, but it has the same losses and performs nearly identical to this one.
It started as a test to determine usable flow, with the VZhex, straight gears, dual, 17mm wide, .5A LDO motor. After some tests failing at flows over 63 I switched to a 15:1 worm drive using the highest torque nema 14, moons motor. The results were the same on flow cap. Then I decided to install the moons motor on the VZhex. It did push a little more, but it's so marginal and would not take the PA needed. I do have an LDO 20mm wide motor I can try, but I think it would be for not, who knows, I have never ran the 20mm LDO @ 48V.
ABS 270C, 0.8CHT nozzle, 20K acceleration, 20SCV, using regular PA at .02, ST.02. I can take away a little more fan, maybe 20-30%. Feels strong as is! Super impressed by the strength!!
Sparse infill connected for one and not connected for the other.
Also consider "ensure vertical shell thickness" can also add filament to the inside walls as well.
At what wall count will reduce this effect? - Filament is deposited by pressure, created from the nozzle against the previous filament. The final shape is an oval, wider than the nozzle. That excess is used to fill the gaps of a circle, making it a square. The same way in that bridges are always round on the bottom.
I don't know the answer on how many walls will eliminate this effect. I typically use 3 walls for non functional and at least 6 on functional prints, mostly 10 (3-5mm thick walls).
** sorry for the filament choice! It's PRO ABS yellow, a mustard of sorts colored filament. I really don't like it and the perfect filament to use for tests.
Tough crowd here at times.
800mms is too fast without a lot of temp. The hot end has some issues with flow at that speed (100mms3, but 70mms3 cap in slicer). I just can't flow enough with a .4 TC nozzle, less than 275C. Looks like ~60mms3-65 @ 275-280. Race track flow tests were ~92 with .4 CHT(goliath). Interesting, because with worm and Chube I was getting ~58 for functional prints.
UFO duct is definitely better for this test.
Custom Skew Cube + Layer stack test. 60x60x60, top hashes 4mm.
ABS. 275C, 0.4 TC nozzle, .25 layer height, 800mms outer-inner walls at 30k acceleration, 600mms infill at 50K acceleration, top skin 300mms at 30k acceleration. All at 60 square corner velocity. Part cooling at 70%, overhangs 100%, bridges 40%, top solid 40%. 40 grams of filament in 17 minutes(141 grams/hour).
Upping acceleration on external and internal walls from 30k to 50k reduces overall time by 15 seconds.
Top rectangles are more rounded than I like to see, SCV is too high, but we need higher SCV to decrease print time.
Slicer capped at 70mms3 flow, but there were many instances where the flow was greater than 120mms3. So maybe the flow shown on Klipper screen takes into account PA move speed?
Non-linear advance settings:
pressure_advance_model: tanh
linear_advance: 0.015
linear_offset: 0.00
linearization_velocity: 5.0
Are they correct, no. There is still some issues with infill break up. The infill is not as clean as it should be, but this is really not a bad print by any means!
These types of tests are needed in order to get some idea of what will happen on a functional print where the hammer is put down.
* sorry for the vibration in the video. The printer is in my garage, on top of a table that rocks terribly. I have to stand on a step stool to reach the top! The bed is stiff enough that the printer rocking has minimal effect on the prints.
Some more content on the printer build, putting in some time on it.
This was a 'square build'. Meaning the design was such that the length in XY direction is the same, while trying to maintain the same volume I had previously (300x300). This allowed me to remove some moving mass as the expense of some working freedom with my hands.
The belts were moved to the front and impede some freedom when pulling the plate.
The second thing is with 4040 on top, the view of the plate from the side is obstructed.
Temperatures of the frame is higher than I want to see. As though I didn't know that from the beginning, but wanted to see what the baseline would be with what I have now. I will add some insulation to the outside once I figure out how I can keep the temp better regulated. It might be too much if I insulate it and failures as a result?
Front/rear belts have been raised a good bit and the new carriage blocks seem to be working, but at the expense of more weight. Given the speed here, they are not an issue.
It is coming together and I am really proud of myself for sticking it out with this one. I would like to give a shout out to my bud, who helped me in the endeavor. I cannot name him due to him wanting his anonymity. But you can bet he can fix anything with a roll of duct tape and a paper-clip. So, if your watching John D., thank you!
This is Sunlu ABS, 265C, 700mms commanded, 30K acceleration, 40 SCV. These motors (LDO 2504) have terrible VFA from 100-150mms. This particular print is a constant PA value and variable SCV based on feature ranging from 20-40SCV. Acceleration is variable from 20K to 60K. 20K for top solid, 30 for walls and 60 for infill.
Infill i s adaptive cubic @ 5%.
Total time ~22 minutes @ 200% size.
Measurements: Keep in mind that with PA not giving a sharp corner, I had to pick and choose where to measure to avoid any bulge.
X-59.83
Y-59.91
Z-59.92
ID circle top:31.4 - should be 32.4
I.D. second hole above 1st- 25.4 - should be 26.4
There was compensation entered in slicer for vertical holes at .522 (I missed resetting to zero).
XY being not the same means the skew is off. Not much, but it's not printing entirely square. It's close, but can be better for sure.
For starters, the fan speed is way too high for this speed. Discoloration on the top surfaces is the sign! This is the crux when printing higher speeds, too much fan and layer adhesion issues, too little and melting.
Layer stacking is not right and I have some VFA on smaller circles where the speed is too low, probably in the bad range of speed for the motors resonance. I can see some lines in bad lighting, typical for 3DP, seems worse than usual, but this part is a bit faster than I print normally. Filament diameter can affect this, so I will disregard.
PA is such a problem for some reason. There are those small ripples in the corners and flat surfaces. I can see some irregular surface finish where pressure advance kicks in during the starts and stops. I can lower the PA, but the corners suffer. Too low of PA and the corners will get those squiggly lines.
There is an improvement in the shake of the machine, probably due to the added weight of all the panels and more frame. Running the higher square corner doesn't seem to be as loud/violent.
It's a challenge, why not! I really wanted this to come out nice, so it's slowed down for bridges, since it seems to be mainly a bridge test!
I had to print 3 of these to get an understanding of what it wanted from me. Slow the bridge and overhang, reduce the accel in that area, and let it melt plastic quickly for the rest. I would like to see even better bridges, but it is what it is!
In order to get better bridges this is what I had to do:
1: Get PA right for bridges!! This usually means MORE PA than typical for the rest of the print. If you are .02PA then .028-.032 seems to work
2: SLOW DOWN BRIDGES! Here I am at 80mms for bridges and overhangs!!! This is because bridges are being detected as overhangs in some places! 35K acceleration, 85mms, 5SCV (I use change at gcode to customize features for this model. 5SCV for overhangs and bridges, 60SCV the rest.
3: Are the bridges as good as they can be, no.. I need to practice more with them!
This is ABS, 255C, 600mms, 60K acceleration, 60 square corner, 55+mms3 flow, 0.4 tungsten carbide nozzle, chambered at 57C. Playing around, as usual.
printables.com/model/1141963-3d-boaty
I have some VFA that is related to some sorcery I need to find. I was having issue with PA and smooth time, getting ripples on one corner. Reducing smooth time to .01 and lowering PA to .0025 helped a bunch! Values change from speed/acceleration, which drives me crazy.
This is fairly quick, but think it has more left in it(sub 4 seems possible, with good quality)! 50% cooling max, ramping to layer 15. This was a vase mode print, 600mms external, 40K acceleration, with a square corner of 30. ABS at 300C in a heated chamber of 55C.
Interestingly, I am able to monitor my heat break heat synk temperature in real time and cool it with an aquarium pump, using PID. That is the cat's meow! Cool the toolhead/extruder motor and the heat break synk! I love it!
ABS, 295C, 400mms, 30K acceleration, 40SCV, 30% fan(CPAP WS7040)
Issues to correct: VFA.. Is it the lack of couplers or the X carriage mounts I have not made yet?!?
Printer still in garage, some items left to complete, but getting there.
290x290x420 print volume. Printer stand 41" to top of top hat
2504 LDO motors @ 48V, extruder @ 48V, LDO 17mm wide .5A.
Octopus Pro 1.1 H723 mainboard @pi3vam4rco37+ PI3B
Chube hot end, VZhextrudort extruder, TC .4 nozzle.
1PZ Z system, using 1330 universal joint and 20100 extrusions for leverage.
Using WS7040 for part cooling, berd air for toolhead cooling and spare air pump of 100K RPM brushless motor.
2 spool active heated filament drier coupled to a BTT filament motion sensor.
Mic-6, 10mm heated bed with 750W silicone heater.
All aluminum has been hand carved at home using basic shop tools and standoffs made on a small mini lathe. So go easy on me for build quality!
Gantry blocks were giving me hell, but finally making headway and close to building the top hat. Should be complete in a few weeks- CROSSING FINGERS
This printer is composed of:
Octopus Pro, H723, 48V XYE, LDO 2504 motors with 5160 drop in's.
Z is composed of a take on the 1PZ system using a truck universal joint in the rear and assisted with a constant rate spring. The Z system is one that I have really liked of all that I have tried. It is very rigid laterally and is able to float in the front, or tilt.
All the aluminum used here was from the scrap yard and the plates made in my garage using basic tools and a small lathe for the standoffs.
The bed is driven with a 1000W silicone 120V AC bed and the hotend is one of a few I like to use.
Most electronics for the higher current is driven with external mosfets and SSR's.
The CPAP command signal are generated with PWM to voltage converters to get the 0-5V.
This system uses a WS7040 and a 100,000 RPM brushless blower.
Power supplies are cooled with aluminum plate mounting and the 120mm blower fan under the printer. Air enters the rear and under the printer and sends it through PI, then Mainboard, cooling the drivers to 10C-15C above ambient.
The pumps are located under the machine as well as an aquarium pump to cool the hotend and extruder.
youtube.com/watch?v=YuiG1NUG2ec
Need to check alignment again to make sure the height is right on the cross standoffs, belts and motor mounts are done. Had to re-do some plates.... Tough, time consuming job. The hard part is done, so should be down hill from here.
All the parts were made with common hand tools with the exception of the SLM printed worm extruder.
I decided to cut the old halo into 4 rail supports and add the corners for the stepper mounts. This system will not use motor couplers! Long shaft motors with a 3 plate motor mount. 2 plates on the bottom and one on top for the motor.
Extrusions welded and bolted. Tried to stay away from blind bolts as much as I could! Top gantry is 4040 and I am really not liking 4040! Too much of a void on the inside.
Top gantry get's a halo for the rails but the current flatness is very good as is.
This will use the 1PZ Z system I currently have, but I increased the rear extrusion to 20100. Cross braced it to stiffen the moment the 1PZ system has.
I am not a professional welder and aluminum is a bit tricky with anodized material. The main purpose of the welding was to stop any of the bolted connections from rotating and to mitigate vibration.
This will have a build volume of 275x275x420 and is mostly square design. Uses same length cross rails vs the old setup at 420mm and 500. So this machine should be a little faster with ADXL recommended and should be fairly the same shaper frequency.
I need to increase the cooling or slow down a little.
This is one of the tests I am using to compare worm gear extruder to straight cut/bull gear and pinion extruder.
thingiverse.com/thing:2795194/files
This duct is using ~8% air at this speed. Base speed 6%, min layer time .7s, max speed 40% cooling.
500mms/30K internal, 18K external, 40SCV, ABS, 285C
This is a 15:1 worm drive using a 2 start worm and a K1 extruder gear. Interestingly, this arrangement puts the extruder motor on it's back, lowering the center of mass and taking the motor heat to the base toolplate and to the extruder frame.
2.8 rotation distance (~1100 steps per mm). Moons CSE14HRA2L410A-02
Initial tests with plastic and 17mm LDO motor was ~50mms3 flow. However, there was some additional friction in the hob gear. I will go back and test the 17mm against the 20mm Moons motor. The 20mm Moons tested up to 85mms3, but to be fair I didn't run the test past 87.
The filament path was not right. The edges were shaving the filament and making powder. I went in twice and re-shaped the top and bottom of the guide. Using small dremel with round tip bits and then some valve grinding compound on a piece of filament, in and out.
There were 2 surfaces that are dimensional critical that were off ~.35mm one dimension and .25 in another. So with SLM, give yourself some leeway on design so you know where it might matter if things go wrong.
Thermals of the extruder is really good at ~42-45C on the extruder body. Much better than the 100C+ temps I had with plastic and melting.
Assembly is really easy for this one. The cat's meow is the jack screw in the back to set the worm lash. This works like a charm and repeats when the motor is removed. This ensures that when tightening the motor it does not walk into the hob gear.
BTW, that little 20mm moons motor is 101.3 grams!!!
Tested flow at 85mms3(ABS) with .4TC nozzle, 20mm Moons motor vs 50mms3 with plastic and 17mm LDO.
So, why worm you ask? I was sucker punched into doing it!
Worm has some advantages and disadvantages. The hob gear is not able to be backdriven, so pressure buildup in the hotend does not make it back to the motor to cause slip. The extruder gear runs on a bearing that is concentric. So as long as you have a quality made gear, your extrusion quality should be better. It does need more speed to run for the same extrusion, so the motor can run out of torque to drive it. Not able to load filament by hand due to cannot rotate the extruder gear (auto load only).
I will be testing this rig, making some changes to the CAD and making a Rev1.0.
thingiverse.com/thing:6779529
No, it's not a perfect benchy, but the cooling is not the problem for the most part. There are a few places could be better and bridging slowed down a bit more, but it is what it is!
Give it a try and see what it does for you!
5 minute rose( I know I said 8 in video, it was a little quicker), ~50mms3 flow. ABS Sunlu Black, 285C, 60% fan.
This worm drive is promising, at least in vase mode! Let's see what else it can do.
700mms/60SCV/40K acceleration/6 min, ABS, .5s min layer time.
I need to fix those seams! This was a little faster than the other video with Chube, maybe has something to do with cooling or extruder settings.
This is PLA, Polymaker high speed PLA, 8 minute benchy. High speed Polymaker filament does not cool well, so I imagine this will print faster with PLA+ or other.
This is the final version of this duct below using a 19mm CPAP hose, WS7040 blower, 100% fan speed. This can make a PLA Shuriken in 4:30 minutes with a min layer time of 2 seconds. At 150% Shuriken min layer time is 2.75 seconds. Maybe later I will play with this more to see if I can improve.
thingiverse.com/thing:6779529
And gooooooooo
flows ~80GPM, but focused jets to converge at nozzle. Not my style! Thought I would try it. Only 4 jets, no external/outbound air nozzles. Nozzle cooling is good but lacks a wider spray area. Needs air point away from nozzle.
Puts out ~85GPM, 32grams/force on the scale. Significant pressure compared to others I use. I believe this is a fail and will result of overcooling. Pressure is too high for bridges, but I will test and see what fan speed it needs so it does not blow it into the next county. My style is to use large flow that goes in all directions. This puppy's gonna need more holes!
Testing the UFO duct. This is interesting! With this duct and this model, 10% fan... I had it set to my normal 35% fan @ .7s min layer time, and the layers separate. Move down to 15%, increase temp to 285 and bingo.
700mms/30K acceleration/60SCV/ABS/285C
youtube.com/watch?v=XFG1Eomm_OM
550mms 1st 8mm, 700mms the rest, 40SCV, 30K acceleration.
0.025 PA 1st 8mm, 0.022 PA for the rest. Retraction 60mms, unretract 60mms, extruder acceleration 20K. Flow up to ~54mms3
265C ABS, 3 walls, .2 layer height, 2 bottom.
EDIT: problem solved! Bone head me increased the extruder acceleration beyond the extruders limits n the printer.cfg. Bone head move apparently!
VaLvNaToR can do this in ~2:30 and have been as low as 2:15, see other video on this channel. So that is my reference, for this test, in comparison. This one was 2:38!!!
I went to the 8 orifice in order to get more volume. The 4 hole was in the range of ~60GPM, and the 1st revision of the 8 hole was 78GPM. I have not tested the 3rd revision yet, but it also has some shape changes to allow more flow as well. Low speed is very uniform ring and nice center pocket.
What's holding it back now is the angle of the nozzle orifices. Trying to keep this compact so I don't have to keep increasing Z/height distance of the duct.
This is the revision 2 duct in that I abandoned the 'wedge' profile and instead added an upper torus to better allow more volume into the top section. This allows a more even pressure above the exit orifices so they bleed more homogeneous.
This is all test and tune to see overhang performance in all directions. This duct will get 8 orifices eventually, if I can balance it.
Currently tested 8.5min benchy, no issues with cooling.
This is an old file, not sure the acceleration. 500mms, maybe 30K, 20-50SCV, can't remember. Tops are slower at 300mms. Stack printed too slow, ate up 50 seconds!
Inspired from Crown Cooler duct. I think this duct has potential! Get some side air and maybe hard to beat this one! But I always get excited in the beginning. Let's see what happens in the next few months.
Commentary is purely frustration, comedy to express ones own idiotic design choices.
I used a 5mm bore, 9mm O,.D. bearing. It's too large by about 1mm/.5 per side. The door does not close all the way, but it is putting pressure on the filament. Need a 5mm or maybe even 3mm bore, 8mm O.D. bearing, 4mm wide.
I decided to get a straight gear as I hear a lot the twirl is not good. So I hope to get another perspective with straight vs twirl, or so I hope.
Gonna see if I can tell any difference.
Tramming in the rails makes machine much more quiet and less vibration.
Annex cube test for baseline.
Seams are not right, non smooth surface/rough/jagged/out of place filament. Some corners not as crisp as I like. The issue I keep seeing is that when I increase PA, the seams get wider and more corner bulge. Less pa and the corners look much better, but lost the edge. Seems predictable.
Overall it's a well crafted cube.
Changing to a Capricorn bowden tube to see if a little tighter filament path will help. If this won't work(I have never used Capricorn in that location), I may put a sherpa extruder on and see if any change in performance.
ABS, 265C, .015PA external/internal, .010PA everything else.
Noises on diagonals
Skew not right
Some issues printing and artifacts
This started with the new toolhead for Chube, Goliath, and Mosquito. Prior to this machine was printing well. I skipped the alignment and seems was a poor choice!
Y rail out ~.026", X rail out .003". Y cross rail not aligned to printer frame parallelism.
These issues generally will cause skew issues and binding of the rails, which cause VFA and mechanical wear.
Time to get out the beers, a good few hundred songs on the radio and have at at it.


