Electronics for a new design?

Hi folks!

Im fairly new to the 3-d printing world. But not new to electronics (well, been a while..) and fine mechanics.
Have a small workshop with a watchmakers lathe and some other stuff.

I got my self a Anet A8 for Christmas. I new it wouldn't be the best quality in the world..
And. its not.
In my mind. the thing is close to crap.. But still, works better than one can expect from one! Still, it does what
It was intended to do. Learn 3D
Have been a professional large format, 2D digital producer/printer dude for close to 20 years now. So
3D is getting me exited again! Feels cool!

Anyway.
Im started to fiddle with my own design. Mechanic wise. I have 2 designs on the works.
But have been looking around and researching electronic bit. mainboard, steppers etc.
But, its not really clear to me, what the differences are.

So. what would YOU go for today. and why??

Cheers to all #D printing dudes and dudhttes!!

Edited 1 time(s). Last edit at 01/22/2017 05:27PM by SatorCodex.

State of the art is duet wifi; if you don't mind the price you won't go wrong with that. The general consensus is that 32 bit (usually arm) processors are the way to go, but I'm still making do with the previous generation using Chinese manufactured mega2560+ramps boards. It requires some care and attention to get the most out of them, but the value proposition is unbeatable (i.e. it's dirt cheap).

An arduino due and RADDS board is a cheaper way to get into 32 bit controllers, smoothieboard is also another option and like JamesK mentioned the duet, then after if and only if you are good with programming a Beagle bone black with a 3d printer cape, this last one can actually become the true high end of the controllers once firmware development pass the birth stage.

Edited 1 time(s). Last edit at 01/22/2017 07:03PM by ggherbaz.

One thing you should research before purchasing any electronics 32-bit or 8-bit is the firmware development.

The Ramps/Mega combination has multiple firmwares to choose from although most are not fully released and always in a state of release candidates. I'm not sure if those MKS Base boards use the same firmware or not.

The Duet boards are 32-bit and have excellent firmware support from DC42 with numerous released firmware upgrades with bug fixes and new features. This was the route I took to the point I have 4 CoreXY printers running on Duet 0.8.5 primarily because this controller is decently priced for the options is was built with.

The Arduino Due with RADDS also runs similar firmware to the Duet boards, but up until recently the firmware was not on par with the Duet boards until DC42 started to release untested RADDS firmware.

I believe Smoothieboard is also 32-bit, but know nothing about firmware support or development. And then there is the new Re-arm for Ramps project that runs Smoothieware with the Ramps plugin board.

No matter how you decide to go, my biggest suggestion is to check what firmware options you will have and how active is the development for new features and bug fixes and what operator interfaces exist with the route you decided to go.

I can highly recommend Duet for active development and being a well thought out board with all mod cons. David is releasing new firmware updates on a weekly basis based on our feedback as users.

Ramps/Mega will do the job and Marlin/Repetier etc.. are all very functional, it just feels a little old fashioned right now but as James said its very cheap.

I'm waiting on a re-arm which will run Smoothieware, this I have not used, but I believe is fairly active on the development side also.

Steppers - nema 17 are the popular choice and running them on 24v seems sensible especially 0.9 deg motors which offer higher resolution. Some people prefer Nema 23 on larger machines. Personally I'd stay clear of bowden extruders they are inherently flawed, as filament is elastic and it doesn't perfectly fit a bowden tube so no matter how well tuned it is, they cannot beat a direct extruder (many of which are small and light now), or a remote direct extruder with a cable drive.

My personal opinion, being a fan of the delta and corexy machines, is that shooting for the optimum between lightweight printheads and rigidity is the sweet spot. Too many designs I see are massively overbuilt and heavy, and will print beautifully if you have days to wait for the object, and if quality matters more than everything else including your time, that's fine. On the other hand some designs like the i3 type machines are too flimsy and whilst cheap, won't yield great results no matter how much you tweak them.

I'd spend time and money on making a very stiff frame, with low mass but as-rigid-as-possible-axes within it. This is where deltas have a big selling point for me, no moving motors, fixed print bed, lightweight motion and therefore fast. Getting them accurate however is frustrating you're always chasing that next 0.1mm, then 0.01mm of precision which gets further away the harder you chase.

And use linear guide rails such as MGN-12H or larger etc... they are very reasonable from China, might need a few ball bearings swapping but they beat any other method of achieving precise linear motion that I've used so far.

Whatever you do, the satisfaction when your machine you've just built prints something is really worth it.

Edited 1 time(s). Last edit at 01/23/2017 12:17PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Yep, agree, DuetWifi for me.
I am very impressed with the feedback, responsiveness and help they deliver on a daily basis.
This is support as it should be. Stellar.

Hi!
Thank you all for the response. Sorry for the slow response back..

I checked out due and duet. Seems like the combo with radds is they way forward. If one put som serious effort to the frame ect. I do not
want to let the electronics be the limiting factor.
Same with steppers. 24v is smart in several ways. 0.9 degree steps gives more precision. should also give less vibrations?
Se alot of talking about micro stepping? Thinking of that. The large format printers I work with, is only using servos. Must
be because smoother running? No one tried that on a reprap?

I was thinking of building a new frame for the Anet. But considering the amount of work, and the components with it. Its not
worth the effort. So I will improve/stiffening the frame. change some components etc. so it prints decent. Im at the moment learning
about g-code. figuring out the firmware, or will change it(need to change/adjust acc, de acc, and jerk). maby. a new board (cheap 8 bit) and octoprint. Btw, octoprint looks good!
One thing is for sure. The frame is so flimsy. That a Bowden is the only way here.
In my research (as mentioned above) there is (as always) pros and cons with with both Bowden and direct drive.
But Ive never heard about direct drive/cable!? Is that something using the classic Bowden wire?? Intresting!
Drawback must in that case, be retraction? (if Bowden wire)

Well, as you can see. that is really crazy. Close to build a own design anyway

Regarding weight in frame. My philosophy is sturdy frame with weight. All moving parts lightweight and stiff.

So, my old school maschinging says Iron/steel/cast iron for the frame. Aluminium/carbonfibre for all moving parts.
Have one cartesian 30-35 cm build plate design on the works. Will probably use a combo with quality rods and linear rails.

Core xy design will be much larger. Im not sure how big is feasible/realistic. But, that will use quality rails. Base will probably
a lapped steel surface. with Z travel. I have a flat granit surface for precise measuring. So that is perfect for get things in precise alignment.
We will have to c what I build first. one is defenetly faster to build.. the other more fun and cooler..

chers all!

Edited 1 time(s). Last edit at 01/24/2017 04:01PM by SatorCodex.

Quote
SatorCodex
...
I checked out due and duet. Seems like the combo with radds is they way forward.
...

The RADDS + Due + RAPS128 drivers looks like a nice setup and you get a choice of firmware, but if you want networking then you have to add an external device e.g. RPi + Octoprint. The Duets have networking built-in.

---

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].

It's also worth pointing out that RADDS + Due is not a common platform - there won't be many people to answer questions if the OP goes that route.

I have to thank you all for your efforts and charing your experience.

Have been looking around @ the usual suspect shopping sites.
Looking at steppers. Its seems to be a jungle out there..
Feeling a bit lost. What to actually buy!? Imean. Id rather spend a few
bucks extra to get something with reasonable quality. Size does matter too..

Also, going to replace the belts on the poor Anet. Plasic crap.
Shoud I go with the rubber and "fiber" reinforced type. Or the Pu type
with steel reinforcement? Latest sounds like overkill maby.
6mm gt2 belts is used.

Thank god its Monday soon. So, I can't spend more money on #d printer stuff!

Cheers!

The steel/PU belts are IMO better for precision but depending on which pulleys you are using (16t or less might be problematic) and how far the belt wraps around them, they might be less good for longevity. My delta developed a slanting print, as the steel wires were breaking and the belts stretching in only one part of the belt, which meant the slanted print straightened itself out after 80mm or so. This had everyone including me, baffled for a week or two. Now back to using fibre/neoprene belts again. If you have time to kill its quite a saga [forums.reprap.org]

I have heard of a kevlar/PU belt but it seems to be unicorn horn, as far as actually getting hold of any goes.

For steppers we are spoiled for choice, I've had good experiences with OMC stepper online, and Wantai (who made me some custom motors for not much more than the off the shelf versions).

Edited 1 time(s). Last edit at 01/29/2017 03:26PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Wow!
That must have been a headache!!

Good input on pulley diameter! didn't think of that.
So I go for the rubber type. I bought a belt for my lathe. In the search I found a company
who had a huge selection of belts. To expensive for the Anet. But have to check it for the
next projekt.
Im just now, window shopping at ali"express" So I will order a few meters of belts. Will order
a few small 16 pulleys too. Have amongst other quality problems, waviness int the prints.
think one of the problems is the secondary pulleys. Is just cheap bearings. So plastic belts, with
totas riding on thebarings. Gives vibrations.

I have to check OMC and Wantai. Ty for the tip!
The mental construction of my design goes faster than in cad.

/Johan

p.s Kossel type is interesting. The inventor of that must be a parson that thinks way out of the box!

Quote
JamesK
It's also worth pointing out that RADDS + Due is not a common platform - there won't be many people to answer questions if the OP goes that route.

Unless you have a firmware problem and you use the RRF firmware ... then all duet userbase can help you out since its the same firmware pretty much ( sometime there some option missing in old firmware version but thats it)

Hardware wise... I dont see problem there, its pretty straight forward like any others shields or all in one, hook stuff there and there and you are set.

SO my advice to SatorCodex is unless you already have the raps128 spend a little more and go duet, im a radds user and its working good but when I chose the board we had active firmware update from Dnewman and its not the case anymore. there DC that recently release an update but he is not taking the firmware update for radds officially and those firmware are untested since he dont have the radds to test with. So radds futur is not looking good firmware wise.

Edited 1 time(s). Last edit at 01/30/2017 06:51AM by GroupB.

Ty for the input. As a newbie all information is gold!
There is alot of information here on the forums. And elswere.
But as new to 3D diy. It's not easy to collect and get a good picture of what is what.
Im I wrong to say 32bit bit boards struggles with firmware contunetly? (Probably bad spelling
As most ppl. I want a good board that has a lifespan for a few years.

What do ppl think about smoothieboard?

In my opinion it's a djungle out there.
I do buy from alixepress etc. But things like this. I tend to buy from the originator
who actually has don the development. If it's a product that is at the end of the lifespan. I go cheap
But in this case, it's a serious build. So where it counts, I'll take a plunge.

Cheers!

Some 32bit development seems to have slowed down or stopped, I'm not familiar with Radds, but if no-one who can develop the firmware is doing so, then new stuff just won't happen, i.e. new sensors or new ways of configuring a printer. To the best of my knowledge Smoothie is active, Duetwifi is very active (new firmware releases every other week at the moment), Marlin are moving into 32 bit but it's new for them so it will take time to happen, but for many it will be very familiar when it does, just on a much faster board that won't struggle with high microstepping, 0.9 deg motors, 16t pulleys and all these things which are moving 3d printing on, 32 bit board configure off a text file on an sd card wheras Marlin/repetier had to be compiled onto an arduino, which is quite tedious.

If in doubt buy a ramps/mega clone its £20 and get your printer working, and if you want a bit more accuracy and performance switch over to a 32bit board later on, when you know for sure that 3D printing is something you like to do, sell the ramps or keep it as a backup.

I'd go straight to 32bit for a new machine now but I'm well and truly bitten by the 3D printing bug and I know its worth paying extra for. Its harder to say to someone who wants to dip their toe they spend £150 on a controller even if it is faster, better and more fully featured.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
DjDemonD
I'm well and truly bitten by the 3D printing bug

Yeah, that happens eh?

I got sidetracked into making a better fan duct this weekend which turns out to be quite tricky to design. I'm about 8 iterations in and still not quite there, which was getting a bit frustrating until it occurred to me how long it would be taking if I didn't have a 3d printer...