Torgeir

Hi cesern and Paul, Actually, I've made a mistake posting this one in the wrong topics. I did not use much money on my feeder, but some parts and work for drawing etc. Well for this modifiation, I’ve used between 25-25 USD including freighth. Sure, excluding the tools mentioned. You will need: 1 ea shaft diameter 5 mm and length 45 mm. The shaft for the “knurled wheel” filament drive & GT2-32 pulley. 2 Bearing 5mm (shaft) 10mm (outer diameter) 4mm (width) 1 Pulley GT2-32-5-6 ~USD 5 1 Pulley GT2-6-5-6 ~USD 5 (for the feeder stepper motor) 1 Belt 200-GT2 ~USD 6 (pack of 5 ea.) 11 ea SST Screw 3mm x 10mm (2mm hex key) 2 ea SST Screw 3mm x 30mm (2mm hex key) 2 ea SST Screw 3mm x 25mm (2mm hex key) 10 ea SST Washer 3mm (id) 6.6mm(od) 1 Hexagon (Umbraco) key 1.5mm (for the knurled wheel & GT2-16 pulley) 1 Hexagon (Umbraco) key 2.0mm (GT2-32 pulley – standard for all UM screws) Note: SST = Stainless steel (is not magnetic) Pulleye GT2-32-5-6 Belt type 2mm (tooth space) 32 tooth, 5mm shaft and 6 (6.35mm) belt width. Since the screw holes are threaded (3mm fine), it is wise to use threaded tool to clear out any "plastic remains" as this may cause crack in the wall due to the extra pressure this may cause. The belt I’ve used 200 GT2, is the same type of belt that is used from X & Y steppers (200 mm GT2 belt). Doing so, as I’ve only have two types of belt in my printer (as before) and as this will lift the bowden tube approx 3” you can cut the original tube length 2 times. This to address the wear that’s occur at both sides of this tube. Just a little warning here, if you’re using the advanced menu to feed the filament through the bowden tube (705mm), then if the tube is to short -it might go with full speed into the extruder!!! So do not do this.. Now some pictures: Feeder Unit Details 1 Feeder Unit Details 2 Feeder Unit Details 3 Feeder Unit Details 4 A few word about this modification. If you want to do this modification, you can use the old stepper motor. As there is a single stage gear, the turning direction will be reversed, however as you turn around the feeder this does not matter. The only thing to change is the number of step, as we need to double this number. So for a standard UM2, the stepper amount is 282. This modification using a 400 step/rev we'll change the rate by *2 = 564. As the track line (from the knurled wheel) now is changed to inside loop of (most of the time) the bowden tube, the benefit is less drag forces due to less friction. Friction can only be between 0 and 1, but drag can be any number! I will sure add some more in here soon. Thanks Torgeir

Hi Paul, Forgot to say, If you need to adjust the feeder tension pressure, you need to use a hex key, turning counter clock wise increase the pressure. Good luck Torgeir

Hi Paul, Have you ever tried to clean the nozzle? There is a method named atomic pull, but read it carefully. Look at this site: https://www.google.com/search?client=firefox-b-d&q=UM+nozzle+cleaning+atomic+metod#kpvalbx=_e2vSXpmBKMbc6ASDu4uQBQ24 Torgeir

Check the indicator on the feeder side, about half scale setting should be ok.. Torgeir

Hi Paul, Just saw this.. Can you see if the knurled wheel is spinning on the back, grinding the filament? Try to help feeding by pushing the filament into the feeder, if this help it might be the coupler. Let's see this. Torgeir

Sorry folks, I've made a mistake as this one should be erased. I've put it into here:

Hi Hexonx, Welcome in here. If you make an object to print, you select the prints-peed say 60 mm/sec, so this is the speed that is 100%! This is ofc assuming single wall printing! Thanks Torgeir

Hi John, You should try burtoogle's latest master version, the stringing problems is much less in this version. Here you'll find it: https://www.dropbox.com/sh/s43vqzmi4d2bqe2/AAADdYdSu9iwcKa0Knqgurm4a?dl=0&lst= Also, some of the previous versions of cura may work better. Thanks Torgeir

Hi Mari, I might also say something about printing slow. 🙂 As I'm mostly printing small and very small object, where time is not that of an issue, even if it takes 6 hour.. Here speed is very important and it has to go from slow to very slow sometime, so this is an important issue here. Anyway, using "post process script" is a great tool for improving your 3D printed objects. Thanks Torgeir

Hi, I could not get this out of my head, so I tried this out in Cura (3.6.0 and 4.6.1) and there was some difference. I think this work ok in Cura 3.6.0 (but I've not trying to print..) Anyway, I sliced a GT2-20 pulley and made a speed reduction for layer 20 (single layer). From the gcode file I've found this: Used Cura 3.6.0 and used Post proseccing Script Change AtZ.. speed reduced to 10% for layer 20. ;LAYER:20 ;MESH:GT2_2mm-20T-OD12.22.stl G0 X107.594 Y103.064 Z4.53 M605 S0;stores parameters before changing ;ChangeAtZ V5.1.1: executed at Layer 20 M117 Printing... ch@L 20 M220 S10.000000 . . . ;LAYER:21 ;MESH:GT2_2mm-20T-OD12.22.stl G0 X107.734 Y103.005 Z4.73 ;ChangeAtZ V5.1.1: reset on Layer 21 M606 S0;recalls saved settings Cura 4.6.1 ;ChangeAtZ instances: 1 (This is in the info part in the beginning of this file). No AtZ here in the beginning of line 20, as in Cura 3.6.0! ;LAYER:20 M204 S3000 M205 X20 Y20 But further down, -it is found just at the end of layer 20? M605 S0;stores parameters before changing ;ChangeAtZ V5.1.1: executed at Layer 20 M117 Printing... ch@L 20 M220 S10.000000 G0 F9000 X104.924 Y104.924 ;TIME_ELAPSED:296.302061 ;LAYER:21 ;ChangeAtZ V5.1.1: reset on Layer 21 M606 S0;recalls saved settings G0 F9000 X104.924 Y104.924 ;TIME_ELAPSED:303.686144 ;LAYER:22 To me it looks like Cura 4.6.1 is working in one layer ahead of the one selected, also the start and stop position is changed. But this is for someone else to dig into.. 🙂 Thanks Torgeir

Hi burtoogle, This was a great upgrade, work much better than before! Cura need to adapt more of your ideas. 🙂 Thanks Torgeir

Hi, I'll assume when you select a single layer with a given speed, then it will proceed with this speed.. So I'll guess you need to telling it the new speed to return to. My best guess.. Like this: But I have never used it so... Torgeir

Hi aaerelon, Oh yes., this was quite different for sure, this object is kind of difficult to print as there is no platform/bottom. I'll think someone in here mention that is better to dimension wall thickness in the stc file, as this is better for Cura slicer to recognize. So maybe define wall thickness same as you print, might be worth a try? I've also tried to use Cura 3.6.0 and found this one better, Cura also made one (or more?) version that's traveled along the mid wall. Made my prints nicer for sure. Anyway, whatever path you take, good luck. Regards Torgeir

Hi, Forgot to mention that your model was sliced for you printer; Anycubic i3 Mega.. Torgeir

Hi aaerelon, Not sure if I understood this right, -but you wanted to print your model in spiraling mode, so I made a test with Cura 4.6.1. The model was sliced ok, but a check of the "print file" with the gcode viewer, shoved two issues: a small sector of a single line was missing and one small missing overlap(?). I thought this was strange as the model is made in sprialing mode.. Today I've been using my S3D for some viewing of old sliced print I have. This made thinking about your problem, so I just loaded up your model, sliced with Cura 4.6.1. To my surprise, your model come up without any error, perfect sliced! The "simulator" in S3D showed a proper spiralized model. So I installed Cura 3.6.0 again, gcode viewer showed the same as S3D. Here's some pictures: The sliced model with Cura 4.6.1 The picture of "gcode print file" showing two errors, viewed with the viewer of Cura 4.6.1. And lastly the same gcode file viewed in S3D with no error! Here is the gcode file of your model sliced with Cura 4.6.1: AI3M_G_100Dx100HS_test_1.gcode Thanks Torgeir

Hi Folks, Welcome in here. Since I built my printer I've had some issues and want to give you some hint. The most important part for this plus upgrade is the feeder system, IMO. This cause the feeder wheel (knurled) is connected directly to the stepper motor, so during a long time print this wheel can be quite warm (as it is heated directly through the stepper shaft it is fixed to). This thing in like a "warm knife in butter", under printing with some load (high pressure in the feeder tube) this wheel start grinding the filament and will cause under extrusion as the feeder cannot deliver the amount of filament needed. If you want to go for a geared option (you can print the gearbox in PLA), make sure to turn around the original feeder unit. This mean, to have the knurled feeder wheel to grab the filament at the inside of the loop path out of the filament roll. I've used this since late 2015 and advice this many times here, but I'm not sure that people get this idea. However, I now have seen that the UMS5 is using this approach as well, -so you can make your printer quite better at this point. Further, I've also made a quick release system for the old feeder, I can feed the filament easily all the way without the need of using the stepper motor. The original stepper motor used by UM2 is a 400 step/360 degrees, but in the plus version (and kit) they use a 200 step/ degrees motor. So this is a little different, however, it is easy to change the step rate in the firmware, so no problem using the old stepper and the old feeder -as well the old firmware! In my setup, I used same belt as the short one in UM2, (200 mm GT2) belt as it was easy and inexpensive to get. Then you will need two pulleys with the right ratio differences. Here I've used a 16 tooth at the stepper shaft and a 23 tooth at the "feeder drive" shaft, this give a ratio gear of 1:2. By doing so the torque drive to the filament is doubled! You will need a 5 mm diameter shaft at 45 mm length, plus two bearing that fits into the gearbox. If you are using a 200 step/360 degrees motor, you do not need to makes any change to the firmware. The top and the bottom of the extruder is easy to print in PLA, but over time it will be better with some equivalent high temp filament. Next up will be the Ollson block, but you do not need anything else, the PT100 and the heater is good enough, no problem! But here you might be very carefully when removing the heater and the PT100 from the old block. Do not re bend the sharp 90 degrees angle, move it without disturbing the current bend, as it may break. They tend to be stuck, try to push them out via the lock screw hole using a small rod without sharp edges, a little heat might help. I will assume the plus upgrade might handle any quality issues that (was known?) and might improve the original somehow, but since I built mine with all the quality issues, shaft play etc. etc. I'll guess all that was not an issue later for me since I had no problem. But sure I do not know.. Well, that's was maybe a little too much, anyway it may perhaps give a little inspiration. Sure there it's lot's more, but later.. Here is some pictures of my setup (that's all in all contains some 11 different modifications, but still using same firmware): Feeder and top part of extruder head. Some modification to lo end of extruder head. Extension of fan to improve radiator flow through. The feeder unit with manull feed handle. Regards Torgeir

Hi, Well, I may jump in as I'm an owner of an UM2E (not a plus model), however modified to act as a plus model. 🙂 Heat bed 60 deg. C and nozzle at 200 for PLA, absolute max for PLA is approx 240 deg. C. But never go that high with PLA! Different brand of PLA temp setting might differs slightly, but this values is good estimates. For printing in the lo temperature end, you can print as lo as 180 deg. C., But just practice at 200 will be a good start. Make sure you bed is clean, wash with mild soap, and when you feel with your fingers during last rinsing that your fingers kind of stick to the glass, then it is good. Water temp (35-40) deg. C. Sometimes we found that the glass plate is not really flat at both sides, so I've marked mine up in the left forward corner using a permanent (black) withe boarder pen. When adjusting the height, I am using kind of thin paper, just as the recipt from an atm. It is important that the first layer is kind of squeezed to make it "glued" to the heat bed. To remove the printed object we need to use a tiny sharp blade, I've found that the painters paint remover to be used when removing dried paint is the best tool for this. But, before trying to remove the printed object, let it cool down -then it will be easy to remove. When the printed objects underside (bed side) can be used as a mirror, your first layer is really good. Anyway, happy 3D printing and good luck. Thanks Torgeir

Good catch. Torgeir.

That's clear up for sure, hmm... I'm kind of slow today. Torgeir

I did not reflect your model for sure, but this is a general thing all the time during printing. I have never tried to measure this "normal violent" acceleration/deceleration bu I am sure it is very hi.. 🙂 Rgeards Torgeir

Hi chrisr, I will think this is kind of normal, -as the fan is a mechanical device that will be affected by such a "violent" acceleration. Thanks Torgeir.

Hi Robert, Hmm... I'm not quite sure how this would work? A single layer of plastic directly on the platform will still be at the same temperature as the glass. I don't think I've ever heard this suggested before so I'm a bit hesitant to accept it straight off the bat ----------------------- We know that the heat radiation from the bed can be quite directed to our print object and can degrade surfaces of our print. But, by using a layer of filament at the bed surface around our print object, we can reduce the direct heat radiation. So by using this method the heat radiation will be spread over a wider angle and hence reduce the heat spot effect. Using this method may help, but will depend of the kind of the color chosen and thickness of this layer. All this is from the "thermodynamics" that's been part of my job. Thanks Torgeir

Hi, I'll think you have a coupler that's collect filament (as a ring) in between the bottom flat of the coupler and the upper flange of the heat block. I have had only one grinding of filament at the feeder, and this was the cause of my problem! This ring will build up, and become a solid barrier for the filament feeding system. As I see some black colored remains at the bottom of the coupler flange, this tells a story. Here's a magnified picture of your coupler: Good luck Thanks Torgeir

Hi denormalize, Welcome in here! When I look at your pictures, I think this is not a very much squeezed first layer. Say so because first layer always give high pressure at the nozzle. In your case, the pressure is very lo.. So this one probably is the coupler, the withe device just above the heat block in the extruder. This coupler will over time shrink and have less diameter due to temperature and compression. So first question; during normal print, how much does the nozzle temperature variate in Celsius? Normally, if your printer is good not more than plus minus 1 deg. Celsius. If this is normal on your printer or just close to this, just continue to next point.. Since you say your filaments is grinded the "only" place to look is at the coupler for reduced diameter that will (most often) stop the filament moving further. This coupler is a known issue for this problem. Thanks Good luck Torgeir

Sure! Here it is: First Cura 15.04.5 https://github.com/Ultimaker/Cura/tags?after=lulzbot-17.14 Second Cura 2.5.0 https://github.com/Ultimaker/Cura/tags?after=2.6.1 Ehm, forgot one thing, use gcode viewer in Cura 2.5.0 The viewer in 15.04.5 is not good, but the slicer is.. 🙂 Good luck Torgeir