Ramps adaption for Arduino Due

Hi,

currently i work on the small modifications of Ramps V1.4 to be compatible with the Arduino due board (mainly the 5V issues).

There are already experiences with a modified Ramps and the due ?

Best regards,

n3d

There's no firmware, and almost all of the firmware is using AVR tricks that don't work on the Due.

Might also be worth reading this thread, which covers other people (including myself) looking at a RAMPS-like board suitable for use on the Due.

[forums.reprap.org]

Dear Cefiar,

i used the search funktion with due as keyword but not found these discussions threads.

Thanks for that link.

I think to build the ramps interface board is the fastes way to get the hardware ready to work on porting the firmware

In fact, both myself and bobc have designed interface boards. Bobc's is here: [github.com] (it's called RIB.brd) and mine is here: [github.com]. Bob's was first, but I wanted a few extra things, so I replaced a few headers with through-hole (since the SM ones are expensive and harder to solder), and added a card reader, since the Due moved the SPI to the awkward 2x3 pin header right next to the mcu. I actually just received my PCBs from China yesterday, so hopefully soon I'll have it all soldered up and running. Seeed Studio has good prices and a fairly quick turnaround, and they now offer 4-layer PCBs if you want to get really fancy.

The reason you didn't find those other threads is because of a little "feature" of the forum search that drove me crazy for the first few weeks I visited here. By default, it will search through messages only if they're less than 30 days old. To search deeper, go into advanced options and change "Last 30 Days" to whatever search window you want.

Hope that helps!

EDIT: Stupid emoticons.

Edited 1 time(s). Last edit at 04/11/2013 12:46PM by TopherMan.

Hi guys,

i got a very important question to ask..
what is soo interesting about arduino due that you all want to build shields for?

why are you not using other - less expensive - and even -better- platforms such as STM32 series (also called stm32 discovery) or even more interestingly the TI STELLARIS series that are -even much more cheaper + low power ? (TI = Texas Instruments)

Thank you

badmanjoe Wrote:
-------------------------------------------------------
> i got a very important question to ask..
> what is soo interesting about arduino due that you
> all want to build shields for?

Not an unreasonable question, but not I would say "very important". An "internet perennial" question perhaps, since the same "why X and not Y?" challenge seems to arise every day since mankind was faced with multiple choices.

There are a whole bunch of reasons, some are obvious and some less obvious, but since I get weary of A vs B arguments, I will not bore you with my opinion, but ask, what do you think the reasons are?

Of course, if you think one of these other boards are a better platform, then go ahead and design a shield and write the firmware for it!

One idea from Adrian Bowyer I have come to appreciate, is that while it seems more efficient to channel efforts down one path, in practice with a distributed resource it is really hard to coordinate, and actually counterproductive. Let there be many paths, and allow natural selection to take its course.

With all the ARM chips around, there are dozens of vendors, and hundreds of different sizes from Cortex M0 to M4F, with really not much to choose between them. Inevitably any choice will face the "why not Y?" question.

The advantage to many isn't technical, it's practical. Arduino is a well-known brand, and readily available. For example, when I started my 3D printing journey, I had basically no experience with microcontrollers, and reading about the Arduino and playing around with the IDE was WAY less intimidating than looking at how to build a Gen7 or burning a bootloader or compiling a toolchain or whatever else. This may not make it the absolute best choice from any technical perspective, but considering how often it's sold out, I'd say it's a well-known and widely available product that could easily act as a gateway gadget to get more people 3D printing, or even just a convenient package for people who don't want to assemble all of their electronics from scratch.

Of course, the advantage of the open-source movement is that if you don't like it or don't want to use it, you don't have to. There are always alternatives. Even now if you're interested in ARM chips, there's already the Smoothieboard, 4pi, R2C2, Gen7 ARM, and I believe someone is using a Teensy 3.0. I don't think a Due-based electronics system is going to suddenly become the "main" electronics, I think it's just going to be another option in a wide field.

1. Availability. Long after the STM32 and TI Stellaris series get dropped by their manufacturers (usually for a "new and improved one" the Arduino will most likely be there, in some form or another. Also those suppliers can sometimes be hard to get systems from. Some even say that they're for prototyping/development only, and that you shouldn't build systems based on these boards, which basically means they can't handle the production volume if they need to start shipping 1000's of them.

2. Modularity. I'd like to be able to either re-use the Due if I disassemble my printer and make something else, or borrow a Due from another project if I manage to damage the Due in my printer and need to get something going. I've had a Mega get damaged with a RAMPS before (thermistor input got shorted to 12V), and having to replace a whole board because the micro is damaged seems stupid. While I myself could most likely replace the micro on a Smoothie, RAMBo or similar board, I wouldn't expect your average person who has a printer to be able to do it.

I understand your points.. but the thing that has been bugging me with the arduino due is simple.

arduino due costs around 70$
ti stellaris costs 13$
stm discovery series costs 20-30$

and we both realise that we are trying to lower the costs of 3d printers, and not make them higher. pay attention that the mega2560 boards that we've been using so far (mainly) cost 15-25$ so it was economical.. but moving to arduino due is not.

2nd reason: most people here just want to get their electronics and run the printers - they dont want to develop the firmwares themselves, so using the arduino ide because its simple is an irrelevant argument here.. or thats how i see it. And i beleive all the guys here that actually do spend the time to develop the firmwares dont have any problems developing firmwares outside the arduino ide.

badmanjoe Wrote:
-------------------------------------------------------
> I understand your points.. but the thing that has
> been bugging me with the arduino due is simple.
>
> arduino due costs around 70$
> ti stellaris costs 13$
> stm discovery series costs 20-30$
>
> and we both realise that we are trying to lower
> the costs of 3d printers, and not make them
> higher. pay attention that the mega2560 boards
> that we've been using so far (mainly) cost 15-25$
> so it was economical.. but moving to arduino due
> is not.

Sure, Due is not the cheapest, nor the best. But at some point cost becomes less important than all the other factors. That point seems to be around the $50-$100 mark. If a machine build costs $800, saving $30 on electronics is somewhat minor, especially if you then have a board no-one knows about and there is little support.

The Launchpad and Discovery are artificially priced, they are loss leader promotional tools for their manufacturers. They are also not very shield friendly, and not open source designs.

> 2nd reason: most people here just want to get
> their electronics and run the printers - they dont
> want to develop the firmwares themselves, so using
> the arduino ide because its simple is an
> irrelevant argument here.. or thats how i see it.

Actually no, because aspects of the printer config are hardcoded in include files, so in order to change config users need to compile and download firmware. You can argue that should not be necessary, but *someone* has to write the user-friendly config tool and support it. The Arduino is specifically designed for simplicity, so even novices can manage the process quite easily.

> And i beleive all the guys here that actually do
> spend the time to develop the firmwares dont have
> any problems developing firmwares outside the
> arduino ide.

That is true of some developers like myself who are experienced embedded guys, but many come from a PC apps or electronics background etc and and having a widely supported and easy to use IDE is a big plus.

I think every project develops an inertia because of legacy choices, both in terms of what hardware people physically own and also in "culture" i.e. what information people are familiar with and can share. This happens even in RepRap which is supposed to be all about innovation... Not much you can do about it apart from try to encourage incremental change, or start again with a new project, neither are easy. Trying to push the community in a certain direction will not work, one can only lead by example.

Atmel and SAM3X8E as on the Due would not be my preferred choice of ARM, there are better choices IMO. Obviously Arduino went with Atmel because of existing AVR relationship, but they could have chosen any manufacturer on technical grounds. There is nothing in common between AVR and Atmel ARM.

So we have to work from where we are now, not from where we would like to have been.

bobc Wrote:

> I think every project develops an inertia because
> of legacy choices
------------------------------


This is the argument that i mostly agree with you about. the rest is totally arguable.

badmanjoe Wrote:
-------------------------------------------------------
> I understand your points.. but the thing that has
> been bugging me with the arduino due is simple.
>
> arduino due costs around 70$
> ti stellaris costs 13$
> stm discovery series costs 20-30$

But for how long?

The Mega used to be $60 or more when it was first introduced (from memory) and only in the last few years has that price dropped.

TI have seemingly always had huge issues with under supply of dev boards. Just ask anyone who tried to get an MSP430 after the first few days it was available on TI's site, where in some cases it took literally 3-4 months from order time to get one, if the order actually went through (as their ordering site was notorious for losing orders).

TI and STM want dev's to use the chips in their products, NOT their dev boards. As mentioned, some dev boards out there have specific licenses on them that (supposedly) forbid you from embedding the dev board in a product for final sale. I say supposedly, because I don't think anyone has ever "tested" that. Personally I wouldn't want to either.

> and we both realise that we are trying to lower
> the costs of 3d printers, and not make them
> higher. pay attention that the mega2560 boards
> that we've been using so far (mainly) cost 15-25$
> so it was economical.. but moving to arduino due
> is not.

Comments as per the price quote above.

> 2nd reason: most people here just want to get
> their electronics and run the printers - they dont
> want to develop the firmwares themselves, so using
> the arduino ide because its simple is an
> irrelevant argument here.. or thats how i see it.
> And i beleive all the guys here that actually do
> spend the time to develop the firmwares dont have
> any problems developing firmwares outside the
> arduino ide.

They may not want to, but at least with most of the available firmwares you have to edit the config at least ONCE (to choose your electronics) and then recompile and upload it. Yes, there is at least one that supports having all these config options on an SD card, due to the ARM's lack of non-volatile storage for such options.

Well thank you for the comments! will definately take them into consideration.
i will also check the license thing that you mentioned.

I am by no means an expert on the various flavors of Arduino, but this looked like it could be an interesting piece of hardware for a 3D printer driver: [www.kickstarter.com] If only it had a shield to add the stepper drivers, that is The Kickstarter page says that it follows the Arduino R3 shield pinout, and runs at 3.3 volts. Hopefully a Due-compatible RAMPS would plug right in. . .

That looks impressive! I'd normally be wary of "2-in-1" type things being worth less than the sum of the parts, but here they seem to have combined the best of two worlds into one neat package. The requirement of PC+low level has a lot of applications, I can see this being huge if they execute right.

It brings the possibility of slicing and printing, and it should be really easy to combine the two onto this package. I think I will be tempted to pledge for a quad core...

I should really get on with the RAMPS for Due then...

That UDOO looks...delicious.

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

that udoo is what i would love to switch my ramps too, though i do gotta say i would almost wanna opto isolate the ramps from it

---

[mike-mack.blogspot.com]

But who will port firmware (f.e. marlin) to this board ? It's useless insteed.

Edited 1 time(s). Last edit at 04/18/2013 02:25PM by karabas.

There's been a lot of grumbling of porting a firmware to Due. I've recently started looking again at porting Smoothie to Due, and since this board is made to run exactly like a Due, it would be able to work. The other side could then be a server, or a slicer, or something else, all in one package, similar to how some people use a Raspberry Pi as a print server.

@ dissidence

I totally agree with you. There should be an interface shield between the UDOO and any IO board to isolate the two. A $130 coaster would suck.

---

- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]

ya its a sucky feeling i have done that though.

though i gotta wonder just how to do that, for most of the things a opto isolator would work not sure what to use on the thermistor imputs though.

---

[mike-mack.blogspot.com]

Regarding the thermistor inputs, it's just a matter of pegging the max voltage that you can get into the board.

I've been thinking a bit about this, and I'm wondering if we can use a Zener diode setup across the input (perhaps with a real fuse in line with the thermistor?) to make sure that the voltage to the thermistor input is limited to a max of 3.3v. If the voltage goes over it, the Zener conducts keeping the voltage at 3.3v. The higher the voltage, the more current passed through the Zener and the fuse till it eventually blows, protecting the board.

It's not opto isolation, but it's definitely a good way to avoid losing your board.

PS: Marlin and most AVR firmwares are probably not a good choice to port to the Due. The amount of AVR optimisations you'd need to reverse engineer is large, plus you'd then want to add in a bunch of ARM optimisations (that apart from being different, may need to be implemented in a completely different way. The Smoothieboard firmware looks like a good starting point, as it's already ARM based.

well i think i know what i am going to be starting to work on next week, i have also been thinking that i might try to work on a bigger ramps board too, something with a little more space for proper fuses and a few other things. i just think its time to try to make it a more stable build, too.

---

[mike-mack.blogspot.com]

It's certainly possible to isolate IO from adverse voltages, it depends on how much you want to spend. Our products at work are powered by 24V DC, so all I/O are protected to voltages +/- 24V DC. Customers do accidentally connect things wrong, and our pumps cost $1000s so we don't want them toasted easily. We also protect to static discharges of 6 kV or more.

Smoothie may be ARM based, but that doesn't help much, because the peripherals on different types of ARM are as different as they are on AVR, and it is porting access to those which takes time. Since Smoothie is based on mbed, you have to port mbed library to SAM3X8E for Due or create a compatible library, which is probably as much work as porting low-level peripheral access. It is not trivial, anyway. OTOH, Teacup is actually quite easy to port to an ARM CPU.

I have a dream where there is a low-level driver API for RepRap firmware which is easily ported to different CPUs, and every firmware use the same API and doesn't poke around in low-level registers...probably forever just a dream, I expect.

Hi !

There are several persons willing to work on porting Smoothie to the Due and 4pi, you should contact the smoothie community if you want to do so too, there is probably some collaboration to be done.

Cheers