Bukobot 3D Printer Forum

News: Added the Bukobot Parts List (BOM) to Wiki at http://bukobot.com/bukobot-parts-list
Welcome, Guest. Please login or register.
Did you miss your activation email?

Author Topic: Design Improvements Discussion  (Read 1319 times)

redteam316

  • Newbie
  • *
  • Posts: 9
    • View Profile
Design Improvements Discussion
« on: January 04, 2013, 06:19:20 AM »
Extruder/Fan Mount Design
Initially my Duo kit only came with one fan mount. No biggie, I used some zip ties to secure the second fan to the aluminum plate (and documented it in jitsbuild), and was planning on printing a second fan mount. After the Duo was assembled, I stripped out one of the nuts(more on this below). I emailed Diego and explained the situation and he sent me replacement parts(Thanks!) as well as a 2nd fan mount. Tonight, I got around to removing the zip ties and prepared to put the 2nd fan mount on. It doesn't align the fan properly as the zip tie solution because the bolt and wooden plate mounting the 2nd extruder is physically in the way. I can only think that the flat aluminum sides where you must mount the fans were designed this way on purpose so the fan fits nicely and are not cross-blowing on the other extruders. I took a closer look at the pictures of this area in jit's build here and it appears that jit did not align the fans with the aluminum plate at all when looking closely at the picture. The only other good picture showing the Duo here doesn't have the fans mounted at all and uses the birdstruder instead with a fan in the middle.

I did think it was odd when assembling the 2nd bukostruder that it was not an opposite-hand/mirror image of the 1st extruder. While this would have improved the aesthetics of the machine if they were, it makes good sense to have the same extruder design used for both since it reduces the files needed in production, produces a better nest of batched parts, and allows you to swap parts from another in case one gets damaged. The cost and design complexity of the extruders is much higher than the fan mounts so I'm suggesting that the fan mounts still be improved so that they will fit with the frame properly and miss the 2nd bukostruder bolt.

I have an idea on how this could be improved but haven't designed a better fan mount yet. I'll post the parts when I get them complete or further along with it.

Flexible split-seam cable tubing
The cables shown here are wrapped in the black tubing. Stuffing all of the cables in the supplied 3/8 tubing results in an extremely tight fit and doesn't protect the cables very well. If one length of tubing is to be used here(which is what comes with the kit), it really should be 1/2 tubing and the hole in the x-left-end needs to be larger. In Jit's build, he depicts much more tubing than which is supplied in the kit. The tubing supplied with the kit stopped at the hole in the x-left-end and the rest of the cables were exposed so I used twist tie to tidy them up some. That was a mistake and is what caused me to strip out the left UHMW nut when jogging the z-axis high enough the cables prevented the x-left-end from raising higher. While waiting for the replacement nuts, I bought some more 3/8 tubing from Lowes. It's not expensive, it's about $2.06 for 10ft of 3/8 tubing. So I grabbed 20ft and covered all the cables I could and have some left over. Lesson learned so hopefully others will learn from my mistake.

It may be overkill but I put the cables for each extruder and the fans in separate tubes(3 total) so it would be easier later to replace them individually or re-purpose them elsewhere. If this type of configuration is used, then actually modifying the x-left-end to have 3 cable holes would accommodate the builder better and provide extra holes for other mods they may come up with.

PhracturedBlue

  • Newbie
  • *
  • Posts: 44
    • View Profile
Re: Design Improvements Discussion
« Reply #1 on: January 06, 2013, 03:00:43 AM »
While not exactly the same thing you ran into, I sent the 'home' command without thinking, and not being next to the printer.  I forgot that I still had an object on the bed.  this prevented the z-axis from ever moving, and caused the UHWM z-couplers to come loose(I got lucky that the nuts didn't strip).  I'd prefer to have a weak link built into the z-axis (probably at the coupler) to ensure no damage to hard-to-replace parts is done.

I also found the supplied cable-wraps to be too tight (and too short) and remedied it with new wraps from Home Depot.  i have yet to find a good solution for hooking all the wires to the Azteeg-X3.  It is really a rats-nest back there.

whosawhatsis

  • Administrator
  • Hero Member
  • *****
  • Posts: 598
    • View Profile
    • Whosawhatsis on G+
Re: Design Improvements Discussion
« Reply #2 on: January 06, 2013, 03:39:07 AM »
The UHMW nuts were designed to pop out if the axis hits an obstruction to prevent damage. If that didn't happen, the screws on them are probably too tight.

redteam316

  • Newbie
  • *
  • Posts: 9
    • View Profile
Re: Design Improvements Discussion
« Reply #3 on: January 06, 2013, 05:31:47 AM »
Nuts
whosawhatsis,
That makes sense. It appears that is what happened to PhracturedBlue since the z-axis was moving downward for homing. I never looked at the design that way. I understood it that the UHMW nuts were designed to fit into the blocks and be tightened down to prevent backlash. When analyzing how a 3D printer operates, it becomes apparent that this is practically a non-issue with the z-axis since it should always be raising higher when building. With a typical 3 axis subtractive CNC, the z-axis is constantly moving up and down. Even so, wouldn't not tightening the nut to the blocks cause calibration issues over time? It sounds as if the x-axis should be loose enough that the operator can lift it off of the nuts. Can you clarify this some more?

In my case, the z-axis was moving upward, causing an upward force, and the cables were preventing it from going higher with a downward force, so there was no chance that the nut would pop itself out. In the end it was a cable management problem but may not be in other cases such as being damaged while transporting it somewhere. It doesn't seem like a good idea to let the nut pass completely thru the block, so i think the risk of it being a common problem are minimal but it did bring up a good question: How practical is it to print replacement nuts out of ABS or PLA?

Cable management
Speaking of cable management, I saw the picture you just added to the wiki that shows lower frame brackets that route the cables along the rail. I like it. I was thinking about doing something similar but running the cables vertically from the x-carriage up to the bottom of the top aluminum frame and across to the left aluminum frame and down it to the azteeg. Since the left aluminum frame has moving parts on 3 sides, I was thinking of some printable inserts that could be attached to the inside face. It might require slightly longer cables but I think it would be a very elegant solution, which would also prevent the wire sag jit mentions on the x-carriage, and would eliminate the cables possibly getting caught up on the y-axis.

Leveling the machine
The machine itself is still wobbly even when fully assembled. It likes to lean back towards the y-axis motor since that is where the weight is. I think a printed cup with an insert of a threaded rod that fits on the bottom of the motor that takes a threaded rod and leveling foot attached to the rod would work wonders. By lowering the threaded rod, it will push the back end up enough until the front y-axis leg will not lift.

whosawhatsis

  • Administrator
  • Hero Member
  • *****
  • Posts: 598
    • View Profile
    • Whosawhatsis on G+
Re: Design Improvements Discussion
« Reply #4 on: January 06, 2013, 06:56:40 AM »
If your machine is wobbling when fully assembled, the feet on the sides are probably sitting on their heels rather than their on toes as they are supposed to. This probably means that you pushed the Z beams down too hard when assembling, but the easiest way to fix it is by getting some of those stick-on rubber feet and applying them to the front foot and the toes of the side feet.
« Last Edit: January 06, 2013, 08:43:09 AM by whosawhatsis »

buildrob

  • Global Moderator
  • Full Member
  • *****
  • Posts: 123
    • View Profile
Re: Design Improvements Discussion
« Reply #5 on: February 05, 2013, 12:36:23 PM »
Hi guys,

There's some great design overall - I was really impressed how easy it was to get the tolerances tight on everything - but here is the list of design improvements I hope you guys consider:

- include two M5 washers for the Y axis mount screws. You said recently on the Frame construction page that Y-axis M5 8mm bolts had been delivered too long in some cases but I think you'll find that the bolts you did your original design with were delivered too short (unless my Misumi pieces are out of spec). The Mitsumi channel is 6mm deep and the lip of the Mitsumi channel on the other side is less than 2mm (about 1.9mm on mine) - so a standard M5 8mm screw is going to bottom out (the main problem however is if the bolt isn't held in the center of the channel and so the screw is at an angle - the nut slips down into the channel by a small amount). Cheap and simple solution: include two M5 washers, change build instructions to say to use them.
- even with the two washers, the Y axis joint just has no strength in a rotational direction. This would be easily solved with two small rectangular printed blocks which screw into the Y axis channels to keep the Y-axis perpendicular against the X-axis (one on each side). These can be quite narrow and so should be quite fast to print.
- increase the thickness of the bearing holder outer feet or redesign the location of the holes in the X & Y carriage pieces so that one foot doesn't hang off the edge of them.
- supply the Duo with two wider diameter split tubes rather than three thinner ones. The cables just don't fit well in the ones provided- most frustrating; other people have reported this as well. I found that tubing with 11-12mm ID worked well.
- longer extruder thermistor cables for the Duo. The currently included ones have trouble reaching.
- be a little more generous on the heated printer bed power cable (same length as thermistor cable) so that you can lay the cable better. If you cable tie the cable to the Y-motor mount piece (but on the other side of the Y-axis endstop), then this the cable feed very nicely backwards and forwards without catching on anything. From there it is nice to have the length to run the cable under Y-axis, and a long the bottom of the frame so that it is out of sight. I had to use the longer 16AWG 12VDC power supply cable piece to do this. Including two M2.5 x 16mm screws + two cable ties for this purpose would also help (I'll upload the pics soon - its very neat).
- make sure the 30mm screw supplied for the Z axis stop is threaded the entire length otherwise you can't fit the anti-rotation nut.
- include a minimal power-supply terminal cover. I mean come on guys - take responsibility with electrical safety. You should be able to print a very simple one based on http://www.thingiverse.com/thing:11821 except add two holes in the cover such that when you remove two of the low voltage screw terminals you can then firmly hold the cover in place by screwing two M4 screws into those terminal holes. The thing should should be very easy to print. It's irrelevant that some people might print a better cover themselves - its just not a proper product without providing a proper out-of-the-box solution here.
- the mains cable should be pre-crimped for people as well (as most people won't have a proper crimping tool and you don't want to have the case where it comes loose because of bodged up crimping). I think you can buy them like this.
- include a Bukobot sticker which people can attach to their frame or Y axis.
- include filament feed tube as standard or as an optional upgrade when you buy the printer (e.g., a "plus" kit which could include things like a spool holder, painters tape, tweezers, palette knife, small tube of bearing lubricant - basically everything you need to get started)
« Last Edit: February 07, 2013, 03:11:57 AM by buildrob »

jit

  • Global Moderator
  • Full Member
  • *****
  • Posts: 174
    • View Profile
Re: Design Improvements Discussion
« Reply #6 on: February 05, 2013, 01:53:27 PM »
I had pretty much the same issues with my Duo.  Good list.
Bukobot Duo v1

whosawhatsis

  • Administrator
  • Hero Member
  • *****
  • Posts: 598
    • View Profile
    • Whosawhatsis on G+
Re: Design Improvements Discussion
« Reply #7 on: February 05, 2013, 09:50:29 PM »
We have already started doing most of those things for the most recent kits. I couldn't figure out what you meant by this: - increase the thickness of the bearing holder outer feet or redesign the location of the holes in the X & Y carriage pieces so that one foot doesn't hang off the edge of them.

Tyson

  • Newbie
  • *
  • Posts: 42
    • View Profile
Re: Design Improvements Discussion
« Reply #8 on: February 06, 2013, 01:24:09 AM »

After completing the mechanical build I feel the biggest weakness is the X-Y axis connection.  If the design is to stay this way I think adding two proper T-slot nuts (and of course the washer) would help greatly so I could really crank down on this joint.  Also, some experimentation with the following products on the metal surface interface (not the nut or screw) could be done to see if it increases the force required for the joint to be rotated.  Since I didn't have this product on hand, I tried another bicyclers trick and put a thin coat of AquaNet hairspray on each mating part and let dry before assembly (works at times for bike parts by making the surface tacky, might have made my joint worse in this case - I have no data).

http://amzn.com/B000MAMR4U
OR
http://amzn.com/B0012RIEM6

whosawhatsis

  • Administrator
  • Hero Member
  • *****
  • Posts: 598
    • View Profile
    • Whosawhatsis on G+
Re: Design Improvements Discussion
« Reply #9 on: February 07, 2013, 01:04:01 AM »
In response to the suggestion of printed blocks to enforce the 90-degree angle, we have a part designed for that purpose, but we haven't started using it because they are such a pain to QC. Basically, if the printer that prints them is at all off-square (which ours shouldn't be, I check them periodically to be sure) OR the surface meeting the bottom extrusion has a pimple from starting/stopping the extruder (hard to avoid, especially with certain slic3r versions), the angle that it will enforce won't be the correct 90-degree angle (it will just make the problem worse, and it will be a heritable error). In order for me to trust the part (at least in this print orientation), we would have to build a precision jig and plane the mating surface after printing. We're looking at other solutions like designing the part to be printed in a different orientation, but the simple printed part you suggested is something we've already experimented with and found inadequate.

buildrob

  • Global Moderator
  • Full Member
  • *****
  • Posts: 123
    • View Profile
Re: Design Improvements Discussion
« Reply #10 on: February 07, 2013, 01:20:01 AM »
Hi Rich, The way I was thinking that these would be used is that you get the Y axis perpendicular using a set square as normal and tighten the 2 metal-on-metal screws as normal. You then slide up the printed block on each side of the frame (on the underside of the Y-axis axis) and then fasten the two (or four) screws on each block to lock it into place. Should be simple and shouldn't need exact tolerances as far as I can see (they don't even need to be that square as you would simply press them against the frame at what ever angle they've been printed - they would just need to have a relatively flat edge on one side).
[You would want a right-angle allen key to tighten the screws on the underside of the Y-axis (without turning the printer upside down) but you provide these with the kit anyway.
« Last Edit: February 07, 2013, 01:33:57 AM by buildrob »

whosawhatsis

  • Administrator
  • Hero Member
  • *****
  • Posts: 598
    • View Profile
    • Whosawhatsis on G+
Re: Design Improvements Discussion
« Reply #11 on: February 07, 2013, 01:37:29 AM »
That's one of the other possibilities I was considering. Those parts would be simple enough that we could laser-cut them instead of printing, which should produce harder, straighter edges, but they wouldn't help as much with assembly as the self-guiding parts we thought of making.

buildrob

  • Global Moderator
  • Full Member
  • *****
  • Posts: 123
    • View Profile
Re: Design Improvements Discussion
« Reply #12 on: February 09, 2013, 12:59:38 PM »
We have already started doing most of those things for the most recent kits. I couldn't figure out what you meant by this: - increase the thickness of the bearing holder outer feet or redesign the location of the holes in the X & Y carriage pieces so that one foot doesn't hang off the edge of them.

See attached. [You can also see the surface rust appearing on my linear bearing casings - hopefully that's just a cosmetic issue.]

buildrob

  • Global Moderator
  • Full Member
  • *****
  • Posts: 123
    • View Profile
Re: Design Improvements Discussion
« Reply #13 on: February 09, 2013, 03:34:26 PM »
I've written up my alternative wiring ideas: http://bukobot.com/alternative-wiring-ideas for hiding the wiring. This is the one where a slightly longer platform cable would be helpful. It seems that you guys are trying to go in a similar direction from the recent build pics you've posted.

You could probably solve the platform cable attachment point more cleanly by designing a printed part which attaches to those Y-motor mount holes (instead of using cable ties) but the cable ties still work well in the interim.
« Last Edit: February 09, 2013, 09:04:45 PM by buildrob »

buildrob

  • Global Moderator
  • Full Member
  • *****
  • Posts: 123
    • View Profile
Re: Design Improvements Discussion
« Reply #14 on: February 11, 2013, 01:34:13 PM »
One more feedback thing I just remembered:

I found that a number of my captured slot nuts in the frame corner pieces were spinning when I trying to tighten them (i.e. it had not captured the nut properly). This is obviously annoying when it is the middle nuts so that you can't poke something down the slots.
If you slightly reduced the width of the captured nut hole and/or removed/reduced the sloping angle of the wall then I think it would hold the nuts better.

 

BloQcs design by Bloc
SMF 2.0.3 | SMF © 2011, Simple Machines