Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 06, 2015 04:43PM
Title says it all. I'm using RAMPs right now. Thinking about doing an electronics overhall.
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 06, 2015 05:39PM
I upgraded my Mini Kossel from RAMPS to Duet electronics, as I described in my blog at [miscsolutions.wordpress.com]. The benefits I obtained were, in no particular order:

1. Built-in Ethernet port and web interface, including fast file upload to the built-in SD card. I find this much nicer to use than USB-based host programs, and it gives me real-time status of all the temperatures etc. I can also control the printer from a smartphone.

2. Fast, native USB port with built-in flow control (whereas RAMPS has USB-over-serial without flow control). This avoids problems with the host program failing to send data fast enough to keep up with the printer because of having to wait for an "OK" response after sending each gcode (even Octoprint has this problem sometimes according to recent posts on these forums). Although this point is somewhat irrelevant to me now, because I do all my printing via the web interface and I only use the USB interface for debugging new firmware.

3. Segmentation-free delta movement. I'm not sure whether this significantly improves print quality, but it is mathematically precise, reduces the load on the lookahead queue, and saves having to worry about configuring the right number of segments/second.

4. Fast auto calibration using a least-squares fit. Calibrates the endstop corrections, delta radius and tower positions in a single iteration of bed probing (or 2 or 3 iterations if your parameters are a long way off to start with).

5. Software controlled stepper motor currents. Much easier that twiddling pots. Also allows the current to be reduced when the machine is idle.

6. Better cooling of the stepper motor drivers. Modern stepper driver chips are designed to be cooled through the PCB, and plug-in drivers have far to little area to cool the chips properly.

7. Easier firmware configuration. No need to recompile and re-flash the firmware when you want to change something. The machine parameters are read from a configuration file on the SD card at startup. The homing, bed probing and pause/resume behaviour are also defined by gcode files on the SD card.

8. Colour touch screen control panel.

9. Faster travel speeds, especially when using 0.9deg/step motors (which need double the step rate).

10. Compatibility with 24V power, which is better for driving large heated beds and 0.9deg/step motors.

11. The 5V regulator on Arduino Mega is notorious for overheating and failing if you power anything else from the 5V rail, such as a graphical LCD. This in turn can lead to failure of the heated bed mosfet due to not getting enough gate voltage.

Some of the above are specific to the Duet, others are provided by other 32-bit boards too (and even some 8-bit boards).

To anyone building a new printer - especially electronics novices - I strongly recommend choosing a 32-bit board. The software controlled stepper motor currents, simpler firmware configuration and fast auto calibration alone make commissioning the printer much easier.

Whether an upgrade would be worthwhile to you will depend on how well your existing Arduino/RAMPS setup is working, and if it is working well enough, whether any of these additional features would be of value to you.

HTH David

Edited 4 time(s). Last edit at 10/06/2015 05:41PM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 04:23AM
A well calibrated delta can achieve high print speed at reasonable quality, the bottlenecks are CPU power and/or extruder feedrate. In my case an upgrade to 32bit wouldn´t change much, as I´m using a printed direct drive Bowden setup. The drive gear grinds into the PETG filament. So I have to upgrade to something tougher before I spend money on more CPU power.
-Olaf
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 05:17AM
Is there a duet like board which does not need an external power regulator (since the on-board one generates too much EMI)?
Is there a duet like board which supports 1/32 micro-stepping?
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 05:43AM
Quote
hercek
Is there a duet like board which does not need an external power regulator (since the on-board one generates too much EMI)?
Is there a duet like board which supports 1/32 micro-stepping?

All switching regulators generate EMI and need to be shielded if low EMI is important. The on-board switching regulator of the Duet 0.8.6 generates less EMI than the Duet 0.6 board. Even with the Duet 0.6, I was only able to detect interference on an FM radio when it was placed just a few inches from an unshielded Duet and tuned to a weak station. The EMI issue on the Duet only came up because RepRapPro wanted their completed printer kits to be CE-certifiable. Is any other manufacturer offering kits that are certifiable to CE or FCC standards? Other electronics with on-board switching regulators might generate just as much EMI as the Duet 0.6, or even more.

RepRapFirmware has now been ported to the Alligator board, which has DRV8825 drivers on board, so 1/32 microstepping. Unfortunately it uses SPI to access the SD card, so the file upload speed over Ethernet might be poor. I have yet to see any measurements of it. The Duet typically achieves 200kbytes/sec upload speed, and one user has reported 400kbytes/sec using a carefully-chosen SDHC card.

For a delta printer, I prefer using 0.9deg/step motors with 1/16 microstepping to 1.8deg/step motors and 1/32 microstepping. You could use both, but I think it would only make much difference to the noise level at low speeds, for example when homing (which is when my delta is noisiest). The reason is that with 1/16 microstepping and 0.9deg motors, you have 160 or 200 steps/mm. So at 50mm/sec the step rate is around 10kHz for Z moves, and around 20kHz or more on at least two of the motors for XY moves. As the chopper frequency is only about 30kHz, the drivers can't really react to individual microsteps at these or higher speeds.

Edited 2 time(s). Last edit at 10/07/2015 05:46AM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 07:40AM
Thanks.
I already need 0.9° steppers just for the basic micro-step size of about 0.015 mm with 1/16 micro-stepping. I'm playing with an idea of building a rather robust delta and that may require a small step size. E.g. one option uses Ø 1.5 mm steel cables instead of spectra and that also means that I would need rather big drums and therefore a small single-step angle. Of course I do not want to actually use 1/32 micro-stepping for precision since holding torque will be too small there anyway. So the 1/32 or 1/128 micro-stepping is optional but nice (at the end, it would increase precision at low speeds).
As well as I do not care that much about low EMI :-D
These together with maximum phase current of 2 A are just small disadvantages of duet board.

Looks like my options are duet, alligator, or radds. I did not consider smoothieboard or beaglebone based options yet (mostly because they do not run repetier which I'm familiar and satisfied with).
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 09:10AM
hercek, if you want more than 2A phase current, you can connect up to 4 external stepper drivers to the expansion connector of the Duet. RepRapFirmware supports an M-code which allows you to remap the axis and extruders motors to different driver channels, and to invert the Enable signal if necessary (thanks to Dan Newman for adding that facility).



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 10:59AM
I'll say yes, but for absolutely none of the reasons dc42 listed. I have a 12v system with a fairly large (Onyx) heated bed, and I got tired of smoking the cheaply built ramps cards (most of them have undersized connectors for the 11A terminal, and lack a heatsink on the Q3 MOSFET).

Switched to Arduino Due board (knockoff) and a RADDS board with Repetier Due firmware, and haven't had a problem since.

It also means in the future I can try out the RepRapFirmware, as a port is in progress.

Note also, RRF will not (currently) drive your LCD, if you have one. RADDS might, or it might not (custom cabling may be required).
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 11:14AM
I just discovered this thread and I was wondering the same, I'm currently thinking about building my first 3D Printer, probably a Prusa i3.

I wanted to start by deciding which board to use for the printer, could anyone say if it is wise to choose a board like the Replicape as a beginner, or do you need advanced knowledge to be able to use it? From what I understand the capabilities of a 32bit controller are greater, so I wonder if there is any use in using an older Arduino based board?

Thank you for your help!

Edited 1 time(s). Last edit at 10/07/2015 11:14AM by Boaschtel.
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
October 07, 2015 01:35PM
Boaschtel:

If you want to build a cartesian printer then an 8 bit board will be enough (e.g. arduino mega2560 with ramps 1.4). A lot of people use this so it is easy to find support information. It is good to buy good quality boards so that you do not finish like grat.

If you want to build a delta printer then get a 32 bit board
rp1
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 02, 2016 11:54PM
Thanks for interesting information in this thread,
I am planning to build Prusa i3 variant with 32 bit electronics to explore new technologies. Can anybody suggest me most feature rich 32 bit hardware / software platform available today?
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 03, 2016 06:18PM
Quote
rp1
Thanks for interesting information in this thread,
I am planning to build Prusa i3 variant with 32 bit electronics to explore new technologies. Can anybody suggest me most feature rich 32 bit hardware / software platform available today?

Take a look at the Duet WiFi:

* ARM Cortex M4 processor with hardware floating point
* TMC2660 drivers capable of more than 2A stepper motor current and up to 256x microstepping
* Built-in web interface over WiFi, including fast file upload to the SD card
* Optional daughter boards to support thermocouples and PT100 temperature sensors

Currently you can still order it at the pre-order price (see [www.duet3d.com]) - but that won't last much longer because manufacturing is rapidly catching up on the pre-orders.

Edited 1 time(s). Last edit at 09/03/2016 06:19PM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 17, 2016 08:13AM
For the last 18 months I've been printing and modding/building machines I've been quite against the 32 bit approach due mainly to cost. Also I'd not had any real problems with ramps/arduino.

My newest machine has duet wifi and its a massive improvement. It's a bit like driving around in a Ford fiesta saying "this car is enough and gets me from a to b" then getting a tesla model S and realising there is a big difference.

That'll learn me.

Model s next winking smiley

(seriously though if cost is a big factor and you're building a printer on a shoestring get the least dodgy ramps board you can from a supplier in your country, who can swap it if it breaks. If you're onto your 2nd, 3rd or 4th machine and quality, reliability and ease of use matter more than cost then then it's a no brainer).

Edited 1 time(s). Last edit at 09/17/2016 08:14AM by DjDemonD.


Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 25, 2016 09:32AM
I have a fairly large Delta printer (320mm diameter, 460mm height). I'm running it off of a Rumba that has been serving me quite well (after I got all of the kinks worked out). When printing large models that have large curves/arcs/circles, the 8 bit CPU shows its ugly head. Even if your model had a million triangles making a perfect curve, the CPU is only capable of processing so much. You will notice facets in what should be perfect curves. I will be upgrading to 32 bit electronics eventually. If you have a small Delta, it probably isn't that big of a deal (except for the other features the newer boards give you).
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 27, 2016 10:15PM
Quote
MrBaz
I have a fairly large Delta printer (320mm diameter, 460mm height). I'm running it off of a Rumba that has been serving me quite well (after I got all of the kinks worked out). When printing large models that have large curves/arcs/circles, the 8 bit CPU shows its ugly head. Even if your model had a million triangles making a perfect curve, the CPU is only capable of processing so much. You will notice facets in what should be perfect curves. I will be upgrading to 32 bit electronics eventually. If you have a small Delta, it probably isn't that big of a deal (except for the other features the newer boards give you).

What firmware are you using? I have the same problem when I was using Marlin. When I switched to repetier firmware, the problem disappeared for me.
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 12:56AM
Quote
MrBaz
I have a fairly large Delta printer (320mm diameter, 460mm height). I'm running it off of a Rumba that has been serving me quite well (after I got all of the kinks worked out). When printing large models that have large curves/arcs/circles, the 8 bit CPU shows its ugly head. Even if your model had a million triangles making a perfect curve, the CPU is only capable of processing so much. You will notice facets in what should be perfect curves. I will be upgrading to 32 bit electronics eventually. If you have a small Delta, it probably isn't that big of a deal (except for the other features the newer boards give you).

Can you post a pic of the issue?

If you're printing at 50 mm/s, and using 120 segments/sec (the normal numbers i print at), then each segment will be less than half a mm long... I can't see them. I can, however, see the segments in the models that I print. You can tell the difference because the segments in the model line up with the layer below, while the segments in the firmware won't line up with the layer above.
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 07:12AM
I would like to see the pictures too. Because when segment length is about 0.5 mm long then the maximum interpolation error from the segmentation is below 0.2 µm. (*) Far far far less than the micro step size. I do not thing this can be visible at all. You are probably seeing something else ... e.g. an artifact of the micro-stepping itself ... or artifact from conversion from precise solid (step file) to imprecise stl file format.

(*) This can depend on printer geometry and precise XY position but it is roughly in that range if you are using older Marlin which limits maximum tower speed (not sure whether they changed this in newer versions). It probably does not differ much with Repetier.

I doubt anybody can see any artifact from delta segmentation when segment length is below 3 mm. This is what you get when segments per second (SPS) is about 90 and printing speed about 120 mm/s. This is also the location where maximal error is about 6 µm (about half the common micro-step size). Fear of delta segmentation error is overrated. If you see some artifacts then it is probably because of something else. If you already have an 8-bit controller on your delta then you probably can keep it (just decrease SPS to about 80-100 range). If you are buying a new controller board for a delta then get a 32 bit one so that you do not need to fiddle with SPS (and can keep it at the overly pessimistic default of 200).
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 12:29PM
Quote
hercek
I would like to see the pictures too. Because when segment length is about 0.5 mm long then the maximum interpolation error from the segmentation is below 0.2 µm. (*) Far far far less than the micro step size. I do not thing this can be visible at all.

There is at least one other factor that affects the segmentation error. The head position at the boundary between each pair of segments has to be rounded to the nearest microstep for each motor. So the segmentation error is not 0.2um, it is half a microstep movement of all 3 motors translated to head position. Unless there is another factor I haven't thought of that makes it even higher.

Half a 1/16 microstep error per motor using 1.8deg motors at 80 steps/mm is definitely visible (just), because the moire pattern reduced when I changed to 0.9deg/step motors. See the photos I published at [miscsolutions.wordpress.com].

When I get time, I'll repeat that print using the 1/256 microstepping provided by the Duet WiFi that is now controlling that printer.

Edited 2 time(s). Last edit at 09/28/2016 12:35PM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 02:12PM
Quote
dc42
There is at least one other factor that affects the segmentation error. The head position at the boundary between each pair of segments has to be rounded to the nearest microstep for each motor. So the segmentation error is not 0.2um, it is half a microstep movement of all 3 motors translated to head position. Unless there is another factor I haven't thought of that makes it even higher.

Yes the moire difference is well visible. Thanks for posting it. But that is microstepping error and not a segmentation error. You are always rounding carriage positions to the nearest microstep. This is independent from segmentation (segment length). What I want to say is that if you do not change the microstep length then you will have hard time to see a difference between segment size 0.3 mm and 3 mm. At these segment lengths the segmentation error is at most 1 mirostep ... and it is spread over the distance of the segment (e.g. 3 mm) if it is there at all. It is biggest in the segment center. So that is the difference of about 0.0125 mm in the center of 3 mm segment (precise at the ends). Well as precise as the microstepping itself allows. So the question is can one see 0.0125 mm difference with period of 3 mm on a straight wall of a cube? Since it is spread over the whole segment it will be harder to see as moire. And with shorter segments it is less probably that it will be there at all.

Anyway one gets "bigger" segmentation errors only at high speeds. And at high speeds things like belt elasticity will be much more significant than segmentation errors. Well, at least if one is not printing only smooth "round" objects (like e.g. vases only).

If one is interested in precise parts for some technical stuff then things like extruder calibration, belt and frame elasticity will be a bigger problem. Well and if one wants pretty smooth surfaces for artistic stuff then it is simple to paint ABS parts with acetone or PLA parts with tetrahydrofuran (though THF dulls colors).

For the old Marlin firmware (and maybe even for a new one), the segmentation error is biggest around the middle distance from bed center to a tower. It is so because old Marlin limits carriage speed and not hotend speed. I did not analyze the situation for firmwares which limit the hotend speed (like e.g. Repetier).
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 02:56PM
Hercek, you seem to be concerned only about the error at the centre of the segment. What I tried to point out is that there is a much bigger error at the boundaries between segments. Let's suppose you print a cube, and the perimeters print in the same direction on each layer. If you use a 3mm segment length, then every 3mm there will be an error at the segment boundary. So at intervals of 3mm I would expect to see a very slight convex or concave angle between the segments, or (where you are lucky and the error is tiny) nothing.

If you reduce the segment length to 0.5mm then you will get the same thing with a repeat interval of 0.5mm. However, 0.5mm is comparable with the extrusion width, so I suspect that the pattern will get smoothed out somewhat and be harder to see. I no longer have a delta printer that uses segmentation-based delta movement, so I don't know how visible this effect is. When my printer prints a cube, simultaneous rounding of the position to the nearest microstep for all three motors only happens at the corners. Between the corners, the microsteps for each motor are generated independently at precisely (to within a few microseconds) the right times.

Edited 1 time(s). Last edit at 09/28/2016 02:57PM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 03:42PM
At least with the old Marlin the segment ends are precise to ±0.5*microstep. I.e. rounding to the nearest microstep is the only error at segment ends. In the center of a segment, the error is at most additional 6 µm for 3 mm segments. This additional 6 µm can cause that it will be rounded to a different microstep than if there would be no segmentation error at all. Therefore I said at most 1 microstep error in the center.

Quote
dc42
When my printer prints a cube, simultaneous rounding of the position to the nearest microstep for all three motors only happens at the corners. Between the corners, the microsteps for each motor are generated independently at precisely (to within a few microseconds) the right times.

How this is possible? I know you do not do segmentation therefore each carriage position is precise to ±0.5*microstep. You cannot easily achieve more precision because our stepper motor drivers are discrete. With the discrete stepper drivers, the steppers either move by a microstep or do not move at all. There is nothing in between. This is definitely true for small speeds. With higher speeds, you could use the fact that static stepper motor moment is finite (and rather low for bigger microstepping) and that the rotor/carriage/platform/hotend have some inertia and motor coils have some inductance and belts are rather elastic. Therefore the actuall speed of head movement will vary with something like an offset sinusoid and not constantly (or according to the curve from delta-cartesian transformation). The more quickly you print the more "flat" the sinusoid will be. And the more precise intermediate positions. Provided that your stepper inductance is not too high for given speed and power source voltage. This is a bit better than the ideal discrete movement (which would happen if there would not be any inertia in the system). Still, the intermediate head positions between cube corners will not be perfect. Of course they are not perfect even in your pictures, because if they would be perfect then we could not see the moire in the pictures.

For precise intermediate positions, you need analog current control to the stepper motor phases. Without analog current control to the motors you will get rounding to the nearest microstep, at each position, regardless whether it is at the cube corners or in the middle of the wall. This is more true the slower you print. If it would not be true you would not see any difference on the cube walls between these two pictures. You would see a difference only at the cube corners.


Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 05:06PM
I agree that at slow speeds the motors will move in individual microstep-sized jerks, if the friction is low enough. That is presumably why my cube print still shows some moire. Perhaps I should try printing the cube a little faster. But with the segmentation-free algorithm, at least the microstep-sized jerks happen at the optimum places, instead of being forced to all happen together at segment boundaries.

It would be interesting to model the effect of segmentation, and to see some prints done with different segment lengths on the same printer.

One other thing that I forgot is that Marlin uses the Bresenham algorithm, so the microsteps within each segment don't occur at regular intervals anyway, except perhaps for the fastest-moving motor.

Edited 2 time(s). Last edit at 09/28/2016 05:08PM by dc42.



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].
Re: Is it worth upgrarding to a 32bit controller like a Smoothie or Replicape?
September 28, 2016 06:06PM
I cannot easily do the test with Repetier since I do not have model for delta segmentation error of Repetier. That is the reason I was interested in the pictures. To see if it looks like what MrBas suspects. In other words I wanted to see something without effort smiling smiley

And when I changed on Marlin from SPS 200 to SPS 90 I did not notice a quality change. But I was not really looking for it either. Somebody else reported that he did not notice a quality change when lowering SPS from 200 down too. Maybe if we would look hard enough we could notice it. My computations showed that the difference between head position with SPS 200 and SPS 90 should be at most one microstep for my printer geometry and my print speed. And I decided that one microstep difference does not matter since other errors are much bigger at the speeds/accelerations I print. I think my biggest problem is the belt elasticity. When looking at the objects, I guess it contributes about 0.1 mm error. It is clearly visible around sharp edges of the printed objects. One does not see this error on round objects since accelerations are low when movement changes smoothly. I guess this is the reason people like to print vases with deltas. It is the most easy object to print nicely. It is smooth (therefore elasticity problems are not visible) and it does not require a lot of filament retracts (therefore plastic deposition errors due to bowden will not appear).

Another problem (which dwarfs segmentation errors) is different shrinking of plastics as it cools depending on the horizontal area size. Heat chamber should help here. But I do not have it.

My point is that compared to these errors, one microstep additional error because of delta segmentation is not even on my radar. I could be on my radar if I would print slowly and with low accelerations. But that is not the direction I want to go.
Sorry, only registered users may post in this forum.

Click here to login