Smoothie.org

[mhackney]

The definitive site for Smoothieboard hardware and firmware is smoothie.org. Also note that Panucatt has released their smoothie-compliant board, the Azteeg X5. It runs smoothie firmware and there are pre-made configuration files for it in the software distribution.

[626Pilot]

I have to wonder about the future of the Smoothieboard, and the Azteeg X5. I was talking to wolfmanjm in IRC the other day and he said they're ALREADY running out of RAM and flash on these things! It's common for the firmware to run with only a few kilobytes free, out of 32 or 64 or whatever they have. They are looking to the Intel Edison platform as their next step. The Edison has a dual-core Atom processor at 500MHz and a single-core Quark microcontroller at 100MHz. It also has built-in wi-fi and Bluetooth. wolfmanjm tells me that Intel hasn't "unlocked" the Quark microcontroller yet, which seems a shame. The platform is pretty close to perfect if you want to develop a 3D printer controller. USB, O/S, display, etc. can run on the Atom and all the real-time motion control stuff can run on the Quark.

Also wondering why they didn't just move to a better ARM processor. An A-model Raspberry Pi is $30, and it doesn't have 64K. It has 256MB! How much could it possibly add to their BOM to use a faster ARM?

[bubbasnow]

come on over to linuxcnc

[teoman]

What is the preferred method to hookup IOs to linux cnc?

Can you configure ot run the Rambo etc so that it is only the slave? So linuxcnc transmits step numbers to the rambo which blindly applies them?

[Flateric]

I was just reading about this dual ARM controller board that has been developed for a kickstarter printer. Interesting anyways.

https://www.kickstarter.com/projects/am ... -3d-printe

[bdjohns1]

Figures...just when I get my Smoothieboard up and running.

[teoman]

I have a feeling one of the processors is consumed by the gui.

[Glacian22]

I have a feeling one of the processors is consumed by the gui.

They do actually say on their kickstarter page that one processor is dedicated to running the printer, while the other only does interface. So basically, yep. ^^

[bubbasnow]

What is the preferred method to hookup IOs to linux cnc?

Can you configure ot run the Rambo etc so that it is only the slave? So linuxcnc transmits step numbers to the rambo which blindly applies them?

I got linuxcnc running on my beaglebone black. The BBB plugs into a BEBOPR++, or CRAMPS. then it is wired just as RAMBo.

[626Pilot]

That dual-core printer is running the same kind of processor as the Smoothieboard. It is an interesting solution. I just don't get the idea behind using a processor with a fraction of a megabyte of RAM when other ARM processors come with exponentially more memory for not much more money. It would make sense if they were trying to conserve energy, but 3D printers with various heaters and motors are not the application to feel satisfied that you've cut half a watt of power consumption for the CPU. It just doesn't matter. Nor does saving a few bucks on a controller that already costs over $100. The Pololu drivers are what make these cost so much, not the CPU. The Azteeg X5 already makes the RAMBo look silly, shipping for fifty bucks cheaper and providing a much nicer processor.

The UI on their LCD is nice. It doesn't require much CPU or RAM to do something like that. I have seen LCD touchscreens at Adafruit that will do that, and come with Arduino libraries.

[DavidF]

So what's the reason for upgrading to these boards?? I haven't had any problems ( that I know of) with the rambo. Is it a speed thing or resolution or??? I know a lot of people have upgraded, I just have no idea why.

[Polygonhell]

The arduino based boards are running pretty close to the edge, you can see the result in occasional jerky motion, normally on long linear moves. You can also see it if you move to 0.9 degree steppers.
Smoothie is about 10x faster, an upgrade really isn't needed, but a lot of us like living on the bleeding edge.
I have 2 smoothie boards and neither are currently on my Max.

[DavidF]

ok, ill chose the "if its not broke dont fix it" route. But now ill be keeping an eye out on the motion more now. I have had my V1 for over a year now and havnt had any problems, but im beginning to think that upgrading the arms would be a good idea as I suspect they have a bit of flex as well as some wear. It looks like the v2 arms are alot beef ier ?? It would be interesting to know why specific changes were made from the rostock v1 and the v2. No complains, just curious...

[Polygonhell]

Can't comment on why changes were made for the V2, I have an original kickstarter V1.
I don't like the nylon arms on then max, if you move the platform to the out side of the build surface and give it a twist, there is too much play. Having said that I get great prints out of mine and I still run the original arms.

[DavidF]

Can't comment on why changes were made for the V2, I have an original kickstarter V1.
I don't like the nylon arms on then max, if you move the platform to the out side of the build surface and give it a twist, there is too much play. Having said that I get great prints out of mine and I still run the original arms.

Thats along the same thing I am seeing. I cant say for sure if it is effecting print quality at all, but i have started trying more difficult prints of figurines that have thin tiny arms and I wonder if upgrading to something like the trick laser arms would benefit any. Although most of the troubles im having are with the actual print flexing (or getting broken off from edge curl) Let me go find thingiverse file and see what you think...

[DavidF]

this is the file I have been trying to print http://www.thingiverse.com/search?q=fujoshi&sa=
Its a tough one IMHO and im getting close to getting it the way I want it to look, but not quite there yet. I having trouble with layer cooling and some sagging on the overhangs (sometimes) but I would really like to get a perfect print so I can cast her up in bronze. Ill try and get some pictures up of the print tomorrow, but I have been having trouble getting a good picture of it to show the details...

[DavidF]

[img]http://i1102.photobucket.com/albums/g446/davidflowers1/Mobile%20Uploads/utf-8BSU1BRzA4NDEuanBn.jpg[/img]

Ok here is print #4. I think I will try changing filament and re printing tomorrow to see what happens...

[626Pilot]

So what's the reason for upgrading to these boards?? I haven't had any problems ( that I know of) with the rambo. Is it a speed thing or resolution or??? I know a lot of people have upgraded, I just have no idea why.

I have been writing code for Smoothie for awhile and it's a lot easier to figure out than Marlin or Repetier. Less spaghetti, and the config is easier too. Text files rather than manually recompiling or editing EEPROM values. Right now I'm working on code to fix the Z=0 bug, where you get okay positional accuracy inside the triangle (the points of which are defined by the three towers) but outside the triangle it's awful. Either the nozzle is so high that it can't make layer 1 adhere, or so low that it blocks the nozzle.

SeeMe's V2 arms were supposed to do away with problems induced by sanding, but any sanding at all is stupid in a world where you can just get Trick Laser carbon fiber arms. Both V1 and V2 of SeeMe's arms gave me delta arm blues because of various sanding issues. Waste of time and money. TL arms work right the first time and require zero sanding. If I was SeeMe, I'd just do my own version of that and ditch the arms they use now. I wonder how many support requests they get on account of those arms alone. This is definitely one of those cases where saving money on parts costs you headaches later, which destroy the savings many times over due to the time spent puzzling over them.