Best 3d printer board to date

Hey.

Quote
vegasloki
I'm holding off on ARM until the firmware selection becomes more available, specifically a Marlin port to ARM. Smoothieware looks promising as well.

Smoothie currently supports all of Marlin's features*, and on top of that has much more new features.
It also is designed to make advantage of the 32bits architecture and added computing power and RAM, in many places.
We have smoother/more precise stepping, better math/acceleration, longer look-ahead, much simpler configuration, ethernet, composite USB ( Mass storage + Serial ), more modular ( easier to contribute to ) code, just to cite a few.
I recently had someone comment that they wouldn't use Smoothie because it was boring as they couldn't find things to add to it ( they are wrong of course, we have tons of ideas of new/innovative things to add, and the bigger microcontroller allows for those new things where other microcontrollers are out of space/power ).

* with the exception of support for I2C panels, which we are working on.

Quote
RC-CnC
Customer support is huge for me. I build and support CnC routers. I expect the same level of support. smiling smiley
I'm just looking for a no hassle design...smiling smiley

Smoothie has a great documentation ( [smoothieware.org] ) in a centralized wiki, developped by the community and the people selling Smoothieboard. Everytime we get a new question we try to add the answer to the wiki.
We also have a very active community on IRC.
And I answer Smoothie support emails at wolf.arthur@gmail.com very fast, lots of people can attest to that ( maybe somebody here could ? ). Usually the same day, sometimes after a few days if very busy.

As you are into CNC routers, Smoothie is used on quite a few of them ( including big ones ), and we have code and documentation specific to them, and more being added all the time : [smoothieware.org]
Supporting CNC routers/mills as well as lasers and 3D printers is one of the primary goals of Smoothie.

You mention "hassle-free", that's also a primary goal, which we even put before performance ( but the bigger MCU and optimized code saves the day for performance anyway. Just saying we could have even more performance, but we don't sacrificie hassle-free to get there )

Quote
answe79
I've recently switched to Smoothieboard, and can just say that i love it. With the "premium" connectors it's a joy to wire. It can also run standalone with the addition of an ethernet jack + 5V regulator. Configuration is very easy to change, no more firmware uploads, just drag and drop onto tyeounted SD drive. Software contorl of current limit is also a big +.

I've only used RAMPS and Smoothieboard, but in my book Smoothieboard is in a different league, there is no comparison.

\o/ \o/

Quote
RC-CnC
anwe79...my only concern is if you burnout a stepper driver you have to replace the whole board.

Actually, Smoothieboard has connectors for external drivers, so if you loose a driver, you can just wire an external one ( and there are some very cheap/easy to wire ones ).
So you don't loose your whole board. You just get a few more wires ( and have to pay/wait for a new driver ).

Also, Smoothieboard's drivers have a lot of cooling area, and as they are A4982, they have more protections that you don't find for example on lots of pololu/stepstick-type drivers ( which happen to also be much harder to cool ).
So you both get much harder to burn drivers, and better current/temperature handling ( Smoothie can go up to 1.9A which is hard to achieve on lots of other setups ).

The only way to burn your Smoothieboard drivers, is to plug/unplug them while the power supply is ON. And even then that is not a guaranteed burn, it's going to depend on the current setting, and luck.
Lots of people now know not to do this ( and we insist very strongly that it is a bad thing to do so in the documentation ).

Quote
anwe79
My only gripe with smoothieboard is that it's not super easy to extend with more drivers, but it's no worse than RAMPS in that respect.

Actually, we break out pins for digital current control of external drivers, -and- it's very easy to configure Smoothie for additional drivers.
You still have to solder an additional connector and wire everything up, but it is a bit easier

Quote
vegasloki
No doubt that computationally the newer ARM hardware is more capable than the AT Mega. The part holding up moving to and exploiting higher power processors is the firmware. Most of the firmware ported from exisiting platforms to 32 bit does not offer any feature enhancement.

Oh no !
That's like -exactly- what we worked on a lot, over -several years- with Smoothie.
It was a grbl port several years ago. A LOT of work has gone into adressing exactly what you are afraid of.

We have a shitton of new features, and we have lots and lots of code that is written to take advantage of the higher computing power and additional ram/flash.
We have this huge problem where lots of people assume exactly what you did assume here, and it's so annoying to us, because we have SO MUCH more than that, and people don't know about it.

Quote
vegasloki
Those will require complete re writes to leverage any new power the processor may provide.

We did that ! We totally did ! Look at the code, look at the wiki, look at what people say about it, try it yourself !
We have a modular code structure, simple configuration without recompilation, usb composite/ethernet, contributor-friendly code, and lots and lots of of better math/accel/look-ahead/step generation.
What do -you- want us to have ?

I can say this. I am not a code guy by any means. My codding starts and stops making bash scripts for linux environments I manage. However the arduino IDE is not very dificult to use and the documentation is there, just RTFM or google it. Not dificult at all.

I had a very positive experience with my ramps/marlin combo on my delta. I also like the modularity of ramps. The price is unbeatable at this point, some are as cheap as 35-40usd shipped to your door. That includes ramps/mega/5 drivers/lcd-clickwheel-sd combo. Being that they are that cheap it makes it a good board for a beginner. As for bulk, wiring etc. Depends on the implementer.

I can say this. The mega is all tapped out as far as horsepower. Its done. No more left.

The only board on the market that seems like a worthy successor is the x5, it seems well thought out for an all in one package.

However I have high hopes for a ramps for DUE combination. That combo can come in the same price range as the ramps 1.4. I like modular components as they give me the ability to hack the hardware. The only thing that bothers me is that no one is even taking the due seriously. It seems that the coders of firmware do not know where to turn... first, there are at least 1/2 a dozen boards out there giving 32bit power. To me this is the turning point and feels like the .com craze all over again. I will not move from ramps unless one of two things happens. My performance starts to suffer too much to make it useful. Or the dust settles and we get a proper successor.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

I forgot about this thread... We've been using the X5 for a while so here's an update.

Simply put, you would have to pay me (A LOT) to go back to Ramps.
The X5's not even in the same league as the old Ramps boards, they seem archaic compared to this, it really is that much of a step up.

It just needs a bigger brother now (dual extrusion!), otherwise, we have been quite happy with it.

Quote
sheepdog43
It just needs a bigger brother now (dual extrusion!), otherwise, we have been quite happy with it.

The Smoothieboard 5XC has all you need for dual extrusion ( and was designed with that in mind ).

Quote
arthurwolf

Quote
sheepdog43
It just needs a bigger brother now (dual extrusion!), otherwise, we have been quite happy with it.

The Smoothieboard 5XC has all you need for dual extrusion ( and was designed with that in mind ).

It's close, but it doesn't do 1/32 native stepping like the Azteeg X5, which is what we use on our printers.

Quote
arthurwolf
Hey.


We did that ! We totally did ! Look at the code, look at the wiki, look at what people say about it, try it yourself !
We have a modular code structure, simple configuration without recompilation, usb composite/ethernet, contributor-friendly code, and lots and lots of of better math/accel/look-ahead/step generation.
What do -you- want us to have ?

I'm speaking specifically to a Marlin port, not native Smoothieware. While I use two different kinds of machines in production I build and support 5 machine types with 7 variants total. There are already not enough hours in the day (which is why I'm in the shop right now...) so for me being able to port Marlin files directly is an advantage. Right now, it's mandatory as I've not the time to port the machines to Smoothie and test them. Even at that, except for the two higher end machine types they are value based machines for entry level builds where ease of use, wide community support and most important for the majority of the users I see for those machines, price. I use RAMPS in many cases not because it's the best technology but rather the best solution for that application.

There are two machines on which I'm working that will ultimately use Smoothie. The first is a Shapeoko being built to bring the CNC routing in house (we use Shopbots at Techshop and they won't let us cut ACP or AL) that should be in the shop cutting before the end of the year. That is just for us. The second is a 300 x 300 x 300 Core XY machine with all high end components. That machine is still prototyping the mechanical aspects using RAMPS because I have a bunch. The Core XY will be self contained, WiFi, cam and I'd like to be able to slice from it but the Pi and BBB arent' quite there in regards to speed. My goal is to have that machine be able to operate as a full capability standalone, begining able to download models (not necessarily from the Internet but from a shared drive/cloud in a shop), fix them, slice them and print them. Much like a traditional CNC machine but instead of using a purpose built Windows box using ARM micro computers.

Quote
vegasloki
I'm speaking specifically to a Marlin port, not native Smoothieware. While I use two different kinds of machines in production I build and support 5 machine types with 7 variants total. There are already not enough hours in the day (which is why I'm in the shop right now...) so for me being able to port Marlin files directly is an advantage. Right now, it's mandatory as I've not the time to port the machines to Smoothie and test them. Even at that, except for the two higher end machine types they are value based machines for entry level builds where ease of use, wide community support and most important for the majority of the users I see for those machines, price. I use RAMPS in many cases not because it's the best technology but rather the best solution for that application.

I understand the problem.
Porting Marlin configs to Smoothie is quite easy though : because Marlin is the standard these days, we made Smoothie to be as compatible with it as possible. This means most config options mean the same things, have the same values/units, and do the same things.
One major difference is probably junction_deviation, which we use instead of jerk, and will require fine tuning. But that's only one.
And if you use delta, because we do some things more correctly, there are minor differences there too.
But other than that, it's just copy/pasting.

I understand if you have several machine types, the hour it takes to convert/test per machine can stack up quite fast.

Quote
vegasloki
There are two machines on which I'm working that will ultimately use Smoothie. The first is a Shapeoko being built to bring the CNC routing in house (we use Shopbots at Techshop and they won't let us cut ACP or AL) that should be in the shop cutting before the end of the year. That is just for us. The second is a 300 x 300 x 300 Core XY machine with all high end components. That machine is still prototyping the mechanical aspects using RAMPS because I have a bunch. The Core XY will be self contained, WiFi, cam and I'd like to be able to slice from it but the Pi and BBB arent' quite there in regards to speed. My goal is to have that machine be able to operate as a full capability standalone, begining able to download models (not necessarily from the Internet but from a shared drive/cloud in a shop), fix them, slice them and print them. Much like a traditional CNC machine but instead of using a purpose built Windows box using ARM micro computers.

Nice. I talk to quite a few folks that are working on those kinds of high-end corexy all-integrated-with-slicing machines, sounds like it's the way of the future Glad you'll be using Smoothie there.

Quick answer; They're all rubish. I'd like to see a board with 4 stepper drivers, 1 thermistor port, and an 8 bit processor. Something basic and minimalistic could serve a growing niche in the market. Plus, if you get rid of the HBP and bring your power limit down to 60 watts, it would be significantly more cost effective compared with the robust, but ultimately overpriced controllers on the market today.

Quote
DisruptiveTech
Quick answer; They're all rubish. I'd like to see a board with 4 stepper drivers, 1 thermistor port, and an 8 bit processor. Something basic and minimalistic could serve a growing niche in the market. Plus, if you get rid of the HBP and bring your power limit down to 60 watts, it would be significantly more cost effective compared with the robust, but ultimately overpriced controllers on the market today.

Actually, you only removed about $1 worth of parts from the board ( probably less than that ) ...
That's going to make a trivial difference only in the price of the cheapest boards.

What costs money is the mcu, stepper drivers, and manpower for proper board testing. And you can't remove any of that.

Quote
DisruptiveTech
Quick answer; They're all rubish. I'd like to see a board with 4 stepper drivers, 1 thermistor port, and an 8 bit processor. Something basic and minimalistic could serve a growing niche in the market. Plus, if you get rid of the HBP and bring your power limit down to 60 watts, it would be significantly more cost effective compared with the robust, but ultimately overpriced controllers on the market today.

As for cheap. I don't know how much cheaper you want this board for, but a mega/ramps/lcd/stepper package can be had for about 40usd shipped. IF you simply do not connect the rest of the crap, you can always whittle it down to about 60w... To me any lower price than that, and you are not saving much in the grand scheme of things. For me the biggest expenses were mechanical parts, not the electronics. In fasteners alone I spent about 20usd retail.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Oh I forgot. Get rid of the limit switches too. You can get rid of a lot of crap on these boards.