Prusa i3 3D printer

[tekkydave]

8 hours ago, Dave_D said:

So far, 2 of the parts i've downloaded from thingiverse just don't fit... it's more reliable and fun to make my own STLs... openscad ftw

calibration cube came out at 19.99mm on 2 sides and 20mm on the other... 

Nice print

 

[Stub Mandrel]

11 hours ago, Dave_D said:

calibration cube came out at 19.99mm on 2 sides and 20mm on the other... 

You might want to try '3Dbenchy' a printer test piece:

 

[tekkydave]

Having fun here. Printer x & z axes disassembled to fit new pieces. Discovered the new x pieces slightly deform the plastic Drylin bearings so they are not smooth at all. I need to sand out the inside of the bearing holders on each side to give an extra mm or so. Watch this space......

[Dave_D]

Dont think i'll bother faffing around with little boats... getting excellent prints so far and am busy printing enhancements sold one of the kids and bought some filament lol

[Gina]

We bred goats in the past - ho ho ho

[tekkydave]

Had to give up on the upgrade and put the printer back how it was. Almost anyway - I still need to put the x-axis belt back but that's a job for tomorrow.

I realised when fitting the lead screws that they were offset from the motor shaft. I think what I did was when I mirrored the x-axis end pieces I should have also adjusted the z-axis motor and top bearing brackets. There is a slight offset in the y direction between the lead screw and the smooth rod which I hadn't also mirrored.

I need to adjust the z-axis motor and top bearing brackets so everything is in line before trying again. It's at times like this I wish I had another printer

 

[Gina]

Yes, it certainly helps to have another printer when modifying one.

[tekkydave]

I should have spent more time checking that all the components would work together in Sketchup. I'll definitely check this time when I have made my changes.

[tekkydave]

I have now redesigned the incorrect parts to put the motor spindles in line with the lead screws. This image shows the relevant parts as assembled. I have not included the motors, couplers or bearings. The rods (light grey) and screws (dark grey) are not the correct lengths, they are drawn just to show that all the component centres are in the correct place.

 

[Dave_D]

So , after much faffing about using buildtak beds and skynet3d firmware, the prints were garbage in comparison so i ripped the buildtak off, went back to blue tape and reinstalled the original firmware lololol

prints are back to being marvelous

[Stub Mandrel]

I found you can make PLA stick to build tape or blue tape much better if you wipe them over with meths first. As in virtually impossible to remove!

[tekkydave]

I find PLA sticks ok using a heated bed at about 60 deg. I occasionally get curling at corners depending on the design. This can be alleviated using a brim of 6-8mm.

[tekkydave]

Upgraded the RPi in my control box from a model B+ to an RPi 3. The startup and restart times are much shorter now - down to approx 30s whereas used to take a couple of minutes.

[Dave_D]

well, for everything that's been said about prusa i3 clones, and this Anet A8 (kit, throw it together with spit and chewing gum) in particular, one thing i can't fault it on is the quality of prints...

time to do some astro stuff...

[Stub Mandrel]

4 hours ago, Dave_D said:

Darth Bhudda?

[Dave_D]

15 minutes ago, Stub Mandrel said:

Darth Bhudda?

yep... sits on my desk at work and every time i have to debug anything, i rub his belly for a minute beforehand to 'destress' lol

[tekkydave]

I have now re-fitted the 8mm leadscrews using the redesigned parts.

I had to change the steps/mm setting in Marlin as the leadscrews are 10x faster than the 5mm threaded rods I had before:

From:
#define DEFAULT_AXIS_STEPS_PER_UNIT  {80.0,80.0,4000.0,823}
To:
#define DEFAULT_AXIS_STEPS_PER_UNIT   {80.0,80.0,400.0,823}  

I also changed the following as the leadscrews allow faster movements:

#define DEFAULT_MAX_FEEDRATE  {150, 150, 12.0, 25}    // (mm/sec)  Pluto V2.6.3
#define Z_PROBE_SPEED_FAST 400  // Pluto V2.6.3
#define Z_PROBE_SPEED_SLOW 200  // Pluto V2.6.3
#define HOMING_FEEDRATE_Z  (720)  // Pluto V2.6.3

The z-axis top speed has increased from 4.3mm/s to 12mm/s so it's a lot zippier moving up & down.

I'm still not 100% happy with some aspects of the new printed parts especially how they hold the linear bearings but they are ok for now.

[Stub Mandrel]

Will you still be able to get 0.1mm vertical resolution with a leadscrew of that pitch? Would it be better to use a single start screw?

[tekkydave]

It shouldn't be a problem. A single full step of the motor will move the z axis by 0.04mm (8/200). With 16x microstepping on top of that it comes down to 0.0025mm.

[Stub Mandrel]

That's good. I would be tempted not to use microstepping and 0.08mm, 0.16 and 0.20mm layers if the software will let you as then you will get maximum accuracy (microstepping is notoriously imperfect and full stepping gives you more torque).

[Bagnaj97]

I finally bought an ED80 (from the classifieds on here), only to realise that my previous side-by side tube rings were useless - I'd wrongly assumed that, both being made by Synta and 80mm, the ED80 and ST80 would have the same tube diameter. The ED80 actually has a 100mm tube diameter whereas the ST80 has a 90mm tube diameter. So I went back to Sketchup and designed some piggyback tube rings, they're up on thingiverse. STL files are there for 100mm rings, 90mm rings and mid section, so you can either print regular 90/100mm rings or 100mm bottoms (has a hole and cutout for mounting to dovetail), mid sections and 90mm tops. Unlike my previous side-by-side rings these don't need a large build plate (mid section is 166.5 x 121.7 x 20 according to Cura)

A single piggyback ring can hold the weight of everything (I tested it horizontally over the sofa), so once you've got a pair of rings it should be plenty strong enough. I haven't tested it on a cold and frosty night, which would make the plastic more brittle, so if you're paranoid then you can use a 100mm ring in addition to the piggyback rings as I've done.

[tekkydave]

When the Prusa is going at any speed there is a lot of vibration which can reduce print quality. The design is flawed in two respects; the moving print bed and the weight of the motor on the x carriage. Both of these can cause vibration. In order to reduce the x weight I am switching to a bowden extruder. I have all the parts from E3D and have printed a new hot-end mount using the generator here.

The printed parts

Assembled

I also bought this adapter which will let me use my old extruder with the bowden tube. If all goes well I'll make a more compact unit with more ideal gearing.

 

This the existing extruder, geared at 5.2:1

 

[tekkydave]

I started putting together the other metal parts for the hotend from some spares I bought abour 2 yrs ago. Then I realised the heatbreak I had was for 1.75mm filament. I have ordered a 3mm one. I also need to drill out the thermistor hole in the heater block to 3mm to allow the thermistor I have to fit. I had to do this with the one currently fitted in the Prusa when I originally assembled it.

The latest V6 heater blocks are different in that they accept a thermistor cartridge but I have an old-style one. At the prices E3D charge I'm determined to make use of it.

 

[tekkydave]

Drilled out the thermistor hole to 2.5mm and installed thermistor & heater cartridge:

Waiting for the 3mm heatbreak to arrive.

[Dave_D]

I'd like to support E3D but you're right, they're expensive, especially when you can get a high quality clone for 25% of the cost. Going to be switching to a bowden extruder in the next few weeks.

Have you tried the Craftware slicer btw? Just trying it at the moment and the print quality looks phenomenal compared to the same print with cura

https://craftunique.com/craftware?gclid=CjwKEAjw8ZzHBRCUwrrV59XinXUSJADSTE5kwtmxEydhix6mn0IhnS64WBad7wvXmWa1KqoKFWouZBoCCefw_wcB