RepRapFirmware running on the RADDS v1.5 shield (Arduino Due)

[Dan Newman]

In the spirit of some of the "what are you working on threads", here's a time lapse video of a Cthulhu model printed on a custom Core XY printer running the RepRapFirmware ported to the RADDS v1.5 shield for the Arduino Due.

https://youtu.be/IxniBllWmrY

• Time lapse video produced by Octoprint running on a Raspberry Pi 2 Model B
• RepRapFirmware is the dc42 fork ported to RADDS, https://github.com/dcnewman/RepRapFirmware.
• Cthulhu model by Cymon (Joseph Larson), Thingiverse Thing #17000 (http://www.thingiverse.com/thing:17000)
• See http://www.thingiverse.com/thing:612857 for information on this Core XY design which has evolved from previous designs by Jetguy, Ryan Carlyle, James Armstrong, myself, and others

[James Armstrong]

So that is what you look like. Nice to put a face to the name.

[Chris P]

I like your z axis ;-P

[Ryan Carlyle]

I'm just surprised Dan is taking some time away from Sailfish :-)

[Dan Newman]

On 09/10/2015 6:52 PM, James Armstrong wrote:
> So that is what you look like. Nice to put a face to the name.

Yeah that's in there. But hey, I have a little less grey hair in my
LinkedIn photo. (One of those sites I check maybe once per year and
remain unclear why even that often ;)

Dan

[Dan Newman]

On 09/10/2015 7:05 PM, Chris P wrote:
> I like your z axis ;-P

Yes, for folks who want the truly nice stability of a three leadscrew Z axis but
don't want the headache of sourcing all the parts for yourself and building one,
look into printers from Fusion 3

http://www.fusion3design.com/

Either that or become Jetguy's friend and have him build you one ;)

1. My first Core XY was built by Jetguy, possibly with some help from James Armstrong.
I needed one to develop and test support in Sailfish but didn't have time to put one
together myself.

2. Yes, three leadscrew platforms are really very stable and loads better than
cantilevered platforms or Mendel/Prusa/whatever styly platforms where you
are sliding all that mass back and forth. (I make an exception for the also
well designed Bukito from Whosawhatsis and Deezmaker since it's a small,
lightweight platform.) No problems from vibration and I basically level my platform
once and forget it. If I make a new build plate, then I relevel. Otherwise, I remove
and re-install the same one months on end with no need to ever relevel.

Dan

[TheZippyMan]

Nice work.
My CoreXY has 3 Leadscrews driven by 2 stepper drivers. How do I mirrow those drivers in the RepRap Firmware on the RADDS board?

[mikk]

It seems that this wont't work:

Currently, only one driver can be allocated to each axis or extruder.

As seen here:
http://reprap.org/wiki/G-code#M569:_Set_axis_direction_and_enable_values

Can't you put all motors on one driver?

[TheZippyMan]

No, because of the current. One other sulution is to connect two drivers to one port.

[Jetguy]

Stacking drivers could be an easy solution in this case.

A pair of A4988s stacked such that the output is the only pins that  are not connected between them. Can be done with any driver. DRV8825s, RAPS128s, etc..

Notice this is the output side and I put a low ESR cap on the Vmotor and ground lines to help stabilize since we don't want a destructive ring to happen on the rail (per the data sheets).

Notice I put heatsinks on both the IC and the board on the lower one since the board blocks more airflow to it. The upper one is more exposed and therefor runs cooler.

Notice that one motor is connected to the board output since the lower driver output pins feed through the board to that connector. The UPPER driver has it's output pins facing UP, and therefore the second motor plugs DIRECTLY into the driver.

[Jetguy]

Sorry, the pictures didn't go through, it's all documented in this post over in the clone group. Scroll the whole page and expand any hidden sections (sometimes Google groups shows them as quoted sections for reasons I still cannot fathom).

https://groups.google.com/d/msg/clone3d/5JxKWZr45xs/ts2UGAtm-MUJ

[mikk]

Sry no access to this group. Cant see your images :-(

[Jetguy]

Just go to my flicker and/or request access to the clone group https://www.flickr.com/photos/90025904@N04/albums/72157639644521393

[TheZippyMan]

That looks good. I might go ahead and solder a small PCB for it to keep everything clean. :)

Why did you put the small cap there? Is it really necessary?

Am Dienstag, 13. Oktober 2015 13:09:56 UTC+2 schrieb Jetguy:

[Jetguy]

Per the driver data sheet, you need low ESR capacitance on the MAIN power input feeding the driver. The socket on the board ONLY has enough capacitance for ONE driver.

You need direct, shortest path/ lowest resistance to ensure a STABLE power source for the driver when it's trying to accurately control the loop current through the motor.

Entire point of microstepping drivers.

[Dan Newman]

On 13/10/2015 2:33 AM, mikk wrote:
> It seems that this wont't work:
>
> Currently, only one driver can be allocated to each axis or extruder.
>
>
> As seen here:
> http://reprap.org/wiki/G-code#M569:_Set_axis_direction_and_enable_values
>
> Can't you put all motors on one driver?

http://forums.reprap.org/read.php?146,475752,562655#msg-562655

Dan

[William Cook]

Would you consider RepRapFirmware for RADDS to be stable at this point? What is the targeted hardware setup (RADDS + RADDS LCD? RADDS + PanelDue? RADDS + UDOO Quad + OctoPrint?)

Thanks. I am excited to try it.

[Dan Newman]

On 13/10/2015 10:13 AM, William Cook wrote:
> Would you consider RepRapFirmware for RADDS to be stable at this point?

Stable for testing, yes. I've been printing with it for about 6 days now and no
surprises. But not enough people have used it yet on RADDS to consider it stable.
That said, this is a port to new hardware of a very stable firmware. As such,
the underlying firmware itself is quite stable. In question is just how many
dumb things I did.... For example, the SD card support uses the higher level
FAT32 support in the RepRapFirmware (based on the Atmel ASF libraries). I pulled
in the Atmel ASF hardware SPI library (spi_master), the Atmel ASF MMC SD library
(mmc_sd_spi or some name like that), and then put it my own DMA support. Seems
to work.... But that's an area which needs more usage to establish confidence.

> What is the targeted hardware setup (RADDS + RADDS LCD? RADDS + PanelDue?
> RADDS + UDOO Quad + OctoPrint?)

First a comment on the UDOO: it is NOT 100% Arduino Due compatible despite
UDOO's claims. Namely, they didn't provide the MISO/MOSI/... connector for
shields. That means, if you choose to use an UDOO, you will need to wire
from their off-to-the-side hardware SPI connector over to the RADDS connector
which expects to mate with the matching connector on a real Due board.
You can use a wire-wrap tool to do that or tag solder or whatever. But, to
get the SD card working you have to do something. And the RepRapFirmware
likes to have an SD card (but can operate without one).

There is no native LCD support at present. I plan to add it in the very
near future but have a lot of demands on my time right now so no promises.

And, I happen to like Octoprint a lot and use it with mine. And there
is some effort to make Octoprint work more cleanly with the RepRapFirmware

1. Recently David Crocker made the common branch of the RepRapFirmware
default to behaving like Marlin as regards responses to serial commands.
(It is configurable over gcode; having it default to Marlin behavior is
one less thing to configure for use with Octoprint).

2. In the development branch of Octoprint, Foosel gave Mark Walker the go ahead
to make the Control Panel's "fan on" button default to sending "M106 S255"
instead of just "M106". While many firmwares allow the S255 to be omitted,
not all do. For example, Teacup and RepRapFirmware require "S255" to enable
the fan.

3. Mark Walker is working on having Octoprint not assume that a "M0" command
behaves the same for all firmwares. It doesn't. The RRF crowd uses M0 to
end their prints and for RRF it, among other things, turns heaters off.
For Marlin, for example, it's a pause with the heaters left on. Octoprint
presently intercepts the M0 and pauses. Not good for RRF. Work on solutions
are in progress. If you use Octoprint with RRF at present, just don't use
"M0" as your end gcode. Instead explicitly turn things off with individual
commands.

To answer your question though, at present the target is to support on RADDS

optional native LCD UI
optional Panel Due
optional Octoprint and similar

The second and third are supported today. Missing is native LCD UI support.

Dan

[Dan Newman]

P.S. some brief doc on wiring RADDS v1.5 for RepRapFirmware is at

https://github.com/dcnewman/RepRapFirmware/blob/dev/doc/RADDS-v1.5.md

Dan

[TheZippyMan]

Thanks! I am getting the power simple from another, unused stepper port.

[William Cook]

Thank you for the very detailed post, particularly the clarifications on the UDOO of which I was already a bit wary. Out of curiosity, is there anything that should stop this from working on a RADDS 1.2?

The next printer will of course be RADDS 1.5, but I already have a 1.2 in the stable.

[Dan Newman]

On 14/10/2015 7:50 AM, William Cook wrote:
> Thank you for the very detailed post, particularly the clarifications on
> the UDOO of which I was already a bit wary. Out of curiosity, is there
> anything that should stop this from working on a RADDS 1.2?

I myself have no idea: I've never seen the pinouts for a RADDS 1.2.

Dan

[JolietDelta]

As far as I can tell the pinouts of the 1.2 and 1.5 shields are the same.

[Dan Newman]

On 14/10/2015 5:14 PM, JolietDelta wrote:
> As far as I can tell the pinouts of the 1.2 and 1.5 shields are the same.

It may work or it may not. I myself have no way of telling. Other possible differences
would be the need for a pullup resistor being different between the two. I suppose
you can always try. If it doesn't work, put back whatever firmware you have working
on it now.

Dan

[JolietDelta]

Schematics for 1.2: http://www.dr-henschke.de/RADDS_23_sch.pdf
and for 1.5: http://www.dr-henschke.de/RADDS_26_sch.pdf

I have the exact same repetier (delta) setup running on both 1.2 and 1.5 shields.

[James Armstrong]

Thanks for the Schematic links.