Maple Native as student learning tool?

LeafLabs Garden » Community » Help Us Help You

Maple Native as student learning tool?

(8 posts) (4 voices)

Started 4 years ago by brucemellen
Latest reply from poslathian

brucemellen
Member

I’m looking for a relatively economical (<$100 for students) ARM Cortex M3 base development board w/an IDE that will get new students who don’t yet know C, up and running relatively quickly (a semester goes by very quickly)...

As for a board, Maple Native looks intriguing in that it seems to give access to all pins (albeit many without a reusable header) for some types of projects. QUESTION: And with the headers along two sides, might it soon support ‘leaflabs ARMed shields’ that could bring FPGA, newer touch displays, robotics, vision/imaging, LED/speaker/navigation/button/keypad UI, a std JTAG debug hardware tool, and other add-on shields to a project/course – similar to the Arduino approach, but with a more complete set of pins?

Toward quick student productivity, I’m looking at possibly using LabVIEW’s GUI to generate C for students [in the beginning – until they know C better], especially when multiple threads may be appropriate. In that the LabVIEW Embedded Module for ARM controllers ties to the Keil RealView Microcontroller Development Kit/uVision development C toolchain & RTX OS, and Keil knows of the STM32F103ZE on the MAPLE Native…
QUESTION: Has anyone out here tried to setup the MAPLE Native under the above LabVIEW/Keil environment to have it run more like what LabVIEW considers a tier 1 device rather than tier2 (which doesn’t support communication external to the chip)?
The alternative seems to be using LabVIEW’s blessed Luminary Micro/TI Stellaris LM3S8962-based board.

Posted 4 years ago #
robodude666
Member

brucemellen,

LeafLabs' Maple products make for great learning tools. Their low cost, open nature puts them as a perfect candidate for an educational environment. I certainly feel they're a better option than official boards released by Keil, etc. as those easily cost in the $200-500 range and force you into expensive software environments.

I personally have no experience with Keil or LabVIEW, so I cannot comment on getting the Maple working with either of these development environments. There are other people on the forums who claim to have their Maple working with Keil/LabVIEW, and they might be able to provide some insight.

I think it depends on what the students are being taught, personally: If they are being taught how to use these industry tools (Keil, LabVIEW, etc.) then it's a good idea to use the environments they'll be working with in their careers. However, if the purpose is to teach the students C/C++, I feel going with libmaple & make would be an easier route. I understand many schools are against using open source software, as it clearly can't be better than a $5,000 IDE, but in reality there's no need for Keil to learn C. Sure, the Keil compiler optimizes slightly better than CodeSourcery, but for learning purposes it doesn't matter.

As far as shields goes, I don't believe LeafLabs has any plans on producing any shields. The Maple Native doesn't follow the Arduino's form factor, but the classic Maple does and therefore many existing shields will work with minor library tweaks. However, because the Maple Native is completely open, the schematic and board files are available and can be used to engineer your own shields (possibly a good student project?).

-robodude666

Posted 4 years ago #
gbulmer
Moderator

brucemellen - IMHO the main advantages of LeafLabs' Maple boards are the simple to use IDE, and very simple to get started libraries with examples. This is free to download, so you can see for yourself if it seems suitable without getting a Maple.

As robodude666 has explained, Maple has an arrangement of header pins similar enough to Arduino that some Arduino shields work. Maple Native is a different arrangement of pins from Maple. IIRC LeafLabs and at least one member have mentioned daughterboards / shields on the forum but I can't remember any definite announcement.

If cost is a big issue, look at the range of ST Micro discovery boards, e.g.:
http://www.st.com/internet/evalboard/product/250863.jsp
http://www.st.com/internet/evalboard/product/252419.jsp
They are typically under $20, and include interesting peripherals like accelerometers and headphone amplifiers.

If it matters (though it might not for beginners), the ST micro discovery boards have hardware debug support onboard, unlike current Maples. To get at that support, you could use free code limited versions of Keil. ST Micro mention other products including Atollic which I believe has a free 'nag-ware' version that pops up ads for the product. I believe that Rowley Crossworks 'Non-profit Edition' works properly for the STM32F4 based board.
Maple IDE does not currently support hardware debugging of Maple even when you have the extra debug hardware (JTAG).

Work has been done to get the STM32VLDISCOVERY working with LeafLabs software, and you can find threads using the search.

There are web sites with complete instructions of how to set up Eclipse to use Cortex-M3 (e.g. STM32F processors), using the same free CodeSourcery toolchain as Maple, but IMHO Eclipse is not the ideal place to start for people new to programming. OTOH, if the students go on to industry and use Eclipse, it might be worth working on that path.

Olimex makes what looks like a clone of Maple, but several people have posted on this forum, asking for help, so I assume it isn't a perfect clone.

Other (non-SRM32F) option include the NXP LPC Cortex microcontrollers.
There is the LPC17xx-based mbed, which has a free cloud-based compiler. That is a non-starter for me because I can't always rely on an Internet connection when I need to do work.
There is also NXP's own LPCXpresso boards, which are pretty low-cost (I think less than $25 for an LPC13xx based board), and they have a free Eclipse-based development environment. A Maple's STM32F103 has a lot more resources, than an LPC13xx, but that might not matter for beginners. I believe the LPCXpresso libraries are the industrially-orientated CMSIS, which are rather harder to use than Maple's, but again, if the students are going out to jobs using that sort of stuff, it may be worth considering.

(Full disclosure: I am not a member of LeafLabs staff.)

Posted 4 years ago #
poslathian
LeafLabs

brucemellen - unfortunately I dont have the experience to comment on the feasibility of labview+kiel. However, Maple has been used in high school curriculum before, with great success.

If you are trying to provide a non-C programming environment to the Maple, check out juliandasilva's excellent project Minibloq - which supports Maple. http://blog.minibloq.org/

You can read more about minibloq on our own forms: http://forums.leaflabs.com/topic.php?id=734#post-4303

Its possible that you could also get Modkit - another visual programming environment - to work for Maple, I am not sure if this has been done.

If you are interested in using Maple in an educational environment, please email us!
info at leaflabs dot com

Posted 4 years ago #
brucemellen
Member

Robodude666, gbulmer, poslathian,
To elaborate a bit more on our environment…
This is a large community college that feeds several 4yr schools, such as RIT. There are a few courses that deal with microprocessors. We are committed to expose our students to LabVIEW in multiple ways and are looking for an updated common student platform for the microprocessor, robotics, and FPGA. Only some of these students will be exposed to C/C++ before they go on to other schools. One beginning digital course involves the teaching of logic and, using simple code, wiring basic microprocessor I/O, logic chips, & maybe a little PWM/ADC. We have robotics courses. A course does MCU assembly programming and optionally some C. Another course plans on getting heavily into using LabVIEW to do FPGA through the same platform, and to do some vision/image processing.

At this point there is discussion about possibly using a $200 National Instruments student board along the lines of the NI sbBIO-9606 with a Coldfire MCU, and the Xilinx Spartan-6 LX45 FPGA mezzanine card connector. Another possibility is a Keil NXP-based ARM7 board. BOTH are fully defined within a multivendor LabVIEW IDE that has been put together by NI.

Based upon what little I see from the outside (picture) of the Native and very short OAK FPGA description, you may be on a good hardware track, but there is a desire to go to the next level in program design using graphical tools such as LabVIEW. NI has a short list of ‘tier 1’ ARM boards that fully integrate. And the processor on the Native and FPGA on the OAK appears similar and supported by the Keil parts of the NI IDE. NI has published the method to bring boards with these devices fully to the LabVIEW environment, and that’s why I’m wondering if anyone has attempted or made headway on getting the Native and/or OAK to be fully supported under LabVIEW. I, personally, would prefer to see students use/purchase a less expensive and more current technology base board similar to the Native, with smartly chosen headers, including a complete stackable right header. And with these headers, build up in a redone/improved (yes, not compatible) Arduino style, add-on boards to do FPGA, robotics, GUI… And between your Native & OAK based upon Native boards, the Maple IDE, and some knowledge of C, studentsstudentshobbyists wouldn’t have to subsequently depend upon expensive licenses from NI & Keil to continue to use their boards when they leave here.

…if no work has been completed on getting Native to the LabVIEW environment, then perhaps this thread just needs to be relocated into your development forum.

…Bruce

Posted 4 years ago #
gbulmer
Moderator

Bruce - thank you very much for sharing your experience and views - you've given me lots of food for thought.

Do you know of a viable, preferably Open Source, alternative to LabVIEW?

(Full disclosure: I am not a member of LeafLabs staff.)

Posted 4 years ago #
brucemellen
Member

Gbulmer,
In collaborative, effectively, pre-industry environments such as, in effect, this, as was noted above, often times open source solutions are often overlooked… unless truly better than those well established. I have not personally heard of a LabVIEW alternative; it would not likely be entertained. Just Google’ LabVIEW jobs’ and see demand. A lesser known open source solution would not be as attractive to an employer because of the time needed to get up and productive on the new app.
Boards, on the other hand, seem to be everywhere and change often. if one works under the big software players like NI & Keil, and offers a less expensive and expandable enough alternative for student purchases, it may be entertained. We have made use of the DRAGON12P in our assembly language and another course. The Basic Stamp 2 is still being used in another course. The Arduino was talked about for a while as a possibility for a common platform, until LabVIEW was considered as a means of getting around a course prerequisite knowledge of C. If the Native was positioned as expandable (FPGA, etc.) like the Arduino and could to be demonstrated to run well (including FPGA, USB communication under program control to/from PC, ready for incorporation into robots) under LabVIEW, it would have a shot at adoption here. Unfortunately, the decision on direction is being made now for use in the Spring semester. And NI has the team resources to come in and demo their environment. I also hear the big players send, during Christmas, lots of token stuff to decision makers in companies too. (I have distaste for that business practice) So without an advocate for open-source (I’m one), open source solutions can be at a disadvantage in many environments. On the other hand, though, if you could get your foot in the door in a set of well known schools, it could then snowball by word of mouth, which would be very good for open source.
Best of luck for leaflabs…
…Bruce

Posted 4 years ago #
poslathian
LeafLabs

There are quite a few alternatives for visual programming in general, but nothing with the hardware support that Labview has. For dataflow you can go to:

require a host PC:
pd + firmata
max/msp + firmata
grasshoper

targets hardware:
matlab simulink
flowstone

I have seen several starts (but no finishes) of PD mods that compile and target hardware. I would LOVE to see this running on our gear, particularly once the FPGAs get involved. But this is a very serious development project, and most likely wont happen with the strong support from a company or research grant.

Posted 4 years ago #

RSS feed for this topic

Reply

You must log in to post.