Cheap-and-cheerful Taobao $150 knock-off running at 200mm/sec

Quote
fauxsoul
Wasn't aware anyone else was doing that type of remote-direct drive setup.

yes - jason is a reclusive engineer who does not talk about or promote that which he knows to be technically not up to scratch. it's taken him several years to do the materials research and engineering analysis... and zesty came along and screwed him over, copied the design *without* listening to him, and... it's... yeah.

Quote

TBH not really impressed with the zesty nimble and have no want or need for one. It looks pretty janky, the idler bearing assembly looks really flimsy and I would not pay for that kind of design.

yeah an untrained eye with an aptitude for mechanical engineering, you can look at the design, and you can tell there's something not quite right

Quote

That Flex3Drive system looks pretty cool though, maybe I'll look into that next time.

this is the drop-in jobbie: [flex3drive.com] - don't use it with a geared (wades) extruder though, because the 3x (or 4x) gearing *and* the 40:1 wormdrive gives a 120:1 or 160:1 mulltiplier which takes you OUTSIDE of the maximum rated RPM of a NEMA17. a direct-drive (hob directly onto the NEMA17) is perfectly fine though, if that's what you've got.

it can be done as an add-on later, with literally about 10 minutes of unscrewing and bolting.

Quote

Finally got my 'economy nuts' in, I'm pretty much done with the x axis, now just needs to work on the Y mounts, which should be very quick work and then the Z mounts (which are mostly done).

Still waiting on some things from china (antibacklash nuts, screws, the ball screw I ordered for the bed). So it's still going to be a while before I throw it all together. I started to put some of the extrusion pieces together, this is my first experience with it and I'm not too impressed with the metal corner brackets that I ordered, they have a lot of slop in the slots. I might just print some of my own to make getting things square easier.

oo you got a ball screw? nice.

ah i have a friend who bought a box of 1,000 40mm 2020 extrusion triangles, and another bag of i think 500 20mm triangles if you're interested? i can put you in touch with him, he's based in HK. but please don't be upset if he sets a high MOQ ok? it's not reaallly worth his time to send a few parts, i'm sure you understand.

I've got quite a few of the triangles and I ordered 2 brands and have way more than I need, but thanks though.

I just feel like the triangles don't do anything to help ensure that the actual extrusions are square and aligned, I'm just going to print or buy some that have multiple mounting points and use those where possible, probably just print some.

I ordered this ball screw which I should be getting any day now, I'll just drill a hole and use a countersink/chamfer bit to give it a place to sit. Then I will have to find a couple tensioning/pull springs and I'll put one on each side so it will stay down, at least that's the plan.

Or maybe I can just print something to keep it held down, that might be a better. I think I will do that, it seems easier.


[www.banggood.com]

Quote
fauxsoul
I've got quite a few of the triangles and I ordered 2 brands and have way more than I need, but thanks though.

I just feel like the triangles don't do anything to help ensure that the actual extrusions are square and aligned,

what have you got? can you send a picture? if you're using flat triangles (50mm x 50mm x e.g. 2mm thick say), then you need 3 per corner to get proper alignment and squaring. if you have those cast aluminium things you can get away with 2 of those, but they have to be 40x40mm to get structural strength: the 20x20mm ones will only get you square + aligned.

Quote

I'm just going to print or buy some that have multiple mounting points and use those where possible, probably just print some.

yyeah don't use printed parts and expect any kind of structural strength / rigidity, the only way that would make something decent is if you used 3 (one on each pair of extrusions) and then made damn sure that you had some form of triangular bracing on all faces.

the 40x40mm cast triangles will actually give you a pretty decent amount of structural rigidity... as long as you go completely overboard but also *do not* exceed around a 350 absolute maximum 400mm length 2020 extrusion.

Quote


I ordered this ball screw which I should be getting any day now, I'll just drill a hole and use a countersink/chamfer bit to give it a place to sit. Then I will have to find a couple tensioning/pull springs and I'll put one on each side so it will stay down, at least that's the plan.

Or maybe I can just print something to keep it held down, that might be a better. I think I will do that, it seems easier.


[www.banggood.com]

oh! that's intriguing, never seen those used before, curious what you'd do with them. i thought you meant a full 200mm-travel ball screw.... maybe like this but with a higher pitch [www.aliexpress.com]

I've got some of these.
[www.amazon.com]

I think I want to either get or print something like these, which have 2 anchor points.
[www.amazon.com]


The longest length of extrusion that I have is 380MM but it will be braced in the middle as well as the ends when I am done.

Oh no, I wish I had that kind of ball screw, but nope. Just want to use this one as a fixed height pivot for a 3 point bed leveling setup. I have an idea that I think will work, will draw it out and show you when I get a chance.

Still thinking things through but this is a simple drawing of what I have in mind for keeping the bed pressed against the ball screw, basically a simple spring pinch clamp:



The other option that I can think of would be extension springs like this [www.diss.on.ca]

I tried printing some of these brackets (in addition to using the metal ones) to see if things might feel better and it seems to be a good improvement so I think I will use these when I can.



Edited 1 time(s). Last edit at 12/25/2017 12:40AM by fauxsoul.

Quote
fauxsoul
Still thinking things through but this is a simple drawing of what I have in mind for keeping the bed pressed against the ball screw, basically a simple spring pinch clamp:

it would have to be very strong. you can only do one end like that, though. at least two ends will have to be a different arrangement so as to stop lateral travel.

Quote

The other option that I can think of would be extension springs like this [www.diss.on.ca]

that's 2040 cast corner. they'll be fine for joining lengths of say... up to 100mm joining up to.. 400mm.

Quote

I tried printing some of these brackets (in addition to using the metal ones) to see if things might feel better and it seems to be a good improvement so I think I will use these when I can.

trust me, it's not. get the two longest parts that you have (400mm to 450mm) and join them in the same way. you'll find that the plastic is completely and utterly a waste of time.

2020 cast triangles are fine for *joining* things, and they stop rotation of the parts that they are joining together down the length of the extrusion. you should in NO WAY rely on them for any other kind of rigidity.

at this point i would strongly suggest you review the page i wrote, here [reprap.org]

you need to constrain SIX degrees of freedom to get rigidity:

* rotate X Y Z
* lateral X Y Z

it's really very very specific, and you have to be very strict about the FULL analysis

so for example, two of those 2020 cast corners is not enough to give full rigidity at a corner with 3 extrusions. but if you use THREE then it is. you have stopped rotation on all THREE extrusions.

equally, a box cube is not rigid until you have ALL SIX FACES constrained. you can test this extremely by taking a closed cardboard box, it's rigid, right? then you open it up and WHAM, it's as floppy as anything, right?

so please bear in mind you cannot just consider just two pieces, or even three pieces in your hand, you must consider the ENTIRE design ok?

Thanks for the input. Haven't been working on lately but I did get a few parts in yesterday so I'm going to try and work on it soon.

I got some anti-backlash nuts for my lead screws and got that ball screw thing so I'll have that for when I am ready to make this transition, but I'm going to need to buy some more bolts and probably some more t-slot nuts before I'm able to finish things, underestimated how many I would need.

Still need to:

Design Y idler pulley mount
Design/place mounting for drag chains
Design a parts cooler
Finish the Z motor mounts
Design some holders for LED strips that I got.


Thinking about getting some TCM2130 drivers for my X and Y as they seem to have some cool features, do you have any experience with those? Right now the drivers I have are DVR8825s.

hi briefly, yes DRV8825s.... chuck them out. serious. look up the fact that you have to put 8 diodes in series per motor in the wires (2 per wire) to... yyyeah...

TMC2130s... anything from Trinamic is great... they just run a bit hotter as the extra accuracy in turn requires more power to maintain. if you don't have a controller that can do SPI don't bother with TMC2130 get TMC2100s. bear in mind you WILL need forced air cooling across stupid, stupid polulu style stepper boards *sigh*.

late night here so more another time

Yeah, after I ordered them I started seeing some things online about some of the problems people have with them. I think I will just order a set of A4988s since they are so cheap anyways.

I'm just going to be using a RAMPS board for this printer which has SPI so I think I will get 2 TMC2130s and use them for my X and Y, and maybe I will use a DRV8825 for the extruder if I don't run into any issues with them. I don't see the point in buying expensive drivers for the z axis so definitely not going to splurge on trinamics there.

Definitely going to be putting a fan in there for the RAMPS board.

RAMPS has SPI but it doesn't have the pinouts to connect to the TMC2130s properly. i've been designing a better board (RD3D) which does... but it'll need tweaking the Marlin firmware to do it.

please for f***'s sake do not use RAMPS. please read the warning i put on the page. if you absolutely must use RAMPS then UNDER ABSOLUTELY NO CIRCUMSTANCES exceed 6 amps at ANY time. that means you cannot safely use a standard heated bed in a cold environment. under NO circumstances should you try temperatures of 65C or above. RAMPS 1.4 due to a basic mistake is incredibly dangerous, and there are reports from a Canadian company ending up being blacklisted by the FCC because one of their customers had their house burned to the ground due to a fire caused by a standard off-the-shelf RAMPS 1.4 board.

[reprap.org]

i REALLY MEAN IT. check the video.

I'm going to be using an external mosfet like the one below for the heated bed, I've read about some of the power issues but those are only related to the heated bed power draw right?


[www.amazon.com]

Quote
fauxsoul
I'm going to be using an external mosfet like the one below for the heated bed, I've read about some of the power issues but those are only related to the heated bed power draw right?


[www.amazon.com]

actually... if you use those DRV8825s at max power, they draw up to 2A, if you have 4 of those that is 8A and that is far in excess of the safe amount that the piss-poorly-designed tracks of the RAMPS 1.4 design can handle.

please look at the video - it is really important that you understand quite how dangerous RAMPS 1.4 is.

... here's a sample of slightly better "power+thermal"-design, how I'm using it for my own drivers with up to 20Amps (for more than 10Amps I'm aditionally soldering solid copper wires along the thick tracks):



But maybe even bigger of a problem are the cheap terminal pins (there are some designed for 20Amps, but not used for higher costs) ... and too, the bad habit to solder the the wires ends

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
VDX
... here's a sample of slightly better "power+thermal"-design, how I'm using it for my own drivers with up to 20Amps (for more than 10Amps I'm aditionally soldering solid copper wires along the thick tracks):

[attachment 101125 Thermo-Design1.jpg]

But maybe even bigger of a problem are the cheap terminal pins (there are some designed for 20Amps, but not used for higher costs) ... and too, the bad habit to solder the the wires ends

fauxsoul, so, like VDX is doing: if you're prepared to solder very thick (10A+) wires between connections on the bottom of the RAMPS 1.4 PCB, you should be ok. you'll need, obviously, the circuit diagram.

can you see in the bottom left of the image that VDX posted, there's that maltese-cross-style pattern? the stupid mistake that was made with RAMPS 1.4 is that on **POWER** tracks those "thermal spokes" as they're called are so thin that they only carry about 0.5 amps **EACH**... and in most places the tracks are laid out so poorly or components so close together that there are only SIX such "spokes"... total is 6A absolute maximum SAFE power.

all this just to use a $3.50 board.... sigh...

Will definitely watch your video, aside from what you are working on right now are there any boards that you could recommend that are currently available?

Quote
fauxsoul
Will definitely watch your video, aside from what you are working on right now are there any boards that you could recommend that are currently available?

*sigh* the whole reason why i designed RD3D is because the alternative 32-bit boards are either non-free (proprietary, i.e. completely against the principles of the entire reprap movement), criminally GPL violating (most of these are chinese), or just damn f******g expensive. so... no, after about a year's worth of research, i actually *can't* recommend anything!

Yeah, it seems like there would be more good options available.

I want to have a decent set up and a good board, but a RAMPS setup is like $15 VS something like a smoothie board or duet or any of the nicer boards are multiple times that amount.

The printer that I have right now I got on sale was $230 and it has a melzi board, which I don't really dig (I know that it has some of the same problems as RAMPS). I just can't fathom spending 100-150 on a board for this printer. Eventually when I go ahead and do a from scratch build (coreXY) then I will fork out and get a nicer board but until then I just can't stomach it.

I just don't understand why we are now in 2018 and someone hasn't come up with a good (decently priced) option, I definitely don't have the right to complain though because I surely don't have the skills or knowledge to do it.

When a customer buys a Duet, as well as getting a 32-bit controller board they are getting 5 high current stepper drivers that stay cool at currents that cause ordinary drivers to overheat, high microstepping with optional microstep interpolation, power monitoring so that you can easily implement resume-after-power fail, a bed heater circuit capable of driving an 18A bed without the mosfet overheating (whereas RAMPS often struggles to provide 10A), 2 extruder heater outputs AND 3 controlled fan outputs, firmware configuration by editing a text file in the web interface (no need to rebuild and re-upload firmware when you want to make a change), plenty of expansion capability including up to 7 more stepper drivers, 5 more heaters and 5 more fans, 24V compatibility, optional digital daughter boards to support PT100 sensors or thermocouples, the awesome Duet Web Control web interface without the need to add a Pi or other additional hardware, SD card socket built in, Western standard design and assembly, continued firmware development, a good warranty policy, and support. I think most Duet customers consider that these things make the Duet board good value, but you can ask the question on the duet3d.com forum if you don't believe me.

I appreciate that if you are building a budget printer, it may be sensible to forego some of these features for the sake of keeping the cost down, and to hope that most of the hardware will work, and that the community will provide you with sufficient support. I recommend Arduino/RAMPS to people in that situation, not because it is good, but because it is cheap and better supported by the community than other cheap 8-bit boards.

We could cut the cost of a Duet somewhat if we used cheaper stepper drivers, removed most of the expansion capacity, and perhaps dropped some of the other features too. But quality and support are not cheap to provide, so we couldn't hope to match the price of the cheap Chinese 8-bit boards. The best 8-bit boards (Ultimachine RAMBo) cost almost as much as the Duet, despite being far less capable.

Edited 2 time(s). Last edit at 01/05/2018 05:02PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I would never dispute the value of a duet board, they seem really great and I've heard a lot of good things about them. I might even buy one eventually when I move on from this printer, that being said it surely doesn't fall in line with my usage scenario with this printer (which is a budget printer).

What surprises me really is the lack of options, you've got RAMPs which is $15-$20 then if you want anything that is nicer at all there is a huge leap in price and essentially no justifiable benefit until you get to over $100 or so which is around 5x more expensive.

I'm not saying it's not worth it or anything like that. I'm just some another guy bitching about things on an internet forum, but it sure would be nice if there was a decent economical controller board that didn't have any major flaws and didn't have the added expense of features that your standard 3d printer might not need.

Quote
dc42
(whereas RAMPS often struggles to provide 10A),

.... the extremely dangerous "thermals" design flaw limits the safe (10C rise) current provision in several places - all the main terminals, the fuses, the MOSFETs, the MOSFET terminals (bed, fan, extruder) and many other places all of which are in series - to 6A. back-calculating using 7pcb.com's trace-width calculator which includes temperature rise, 10A will result in a THIRTY (30) Centigrade increase in the temperature of the tracks around the pins that they're connected to. over time that will easily cause a runaway cascading failure and this has (fact) caused fires that resulted in people's houses burning to the ground.

the most obvious thing that tells you there's a serious problem is that a 1.4 ohm heated bed, which draws around 9A @ 12V, cannot reach in many cases over around 75C. the resistance of the power tracks is simply too high.

Quote

I recommend Arduino/RAMPS to people in that situation,

please, dc42, don't. it's simply too dangerous, until there's a released version that has the design flaw fixed and it becomes commonplace, which could take anything up to a year to propagate through the system. i'm amazed that the person who maintains the github repository - with their name over it - hasn't noticed the modification to the wiki page with the warning on it, and bothered to fix it.

if you are recommending RAMPS please for goodness sake warn them and advise them that they need to add large copper bridging wires under every single power trace, and every single MOSFET where they intend to use over 6A. given that the heated beds are around 9A just on their own that's pretty much essential.

Edited 2 time(s). Last edit at 01/05/2018 10:42PM by lkcl.