eLua as intelligent expansion?
Locked
 
|
Patrick Mc(avery |
eLua as intelligent expansion?
|
Hi Everyone
I normally write horrible and long emails but to try to avoid this... in a nutshell I really wish I could have been more active with eLua over the past year. I am spread very thin between my children's needs and my day time work. I have a project that involves controlling scientific instruments and crunching data afterwards. In another nutshell I have had a hard time finding the right solution to the problems I am confronted with and have looked into baremetal embedded, RTOS-like embedded and desktop solutions. It now looks like all three together will be needed. So to get to the point(sorry), SBC like RPI and Olimex's IMX233 are very nice but they don't have enough GPIO pins or peripherals for what I need. Baremetal is nice but sometimes it's easier to have an OS, especially when interacting with the user. I was wondering if anyone has tried to link a SBC together with an eLua microcontroller? So for instance Rpi's 8 GPIO pins could be used for sending data back and forth to an eLua MCU such as a Luminary development board and Lua could do all the low level stuff. Both the luminary board and Rpi can communicate over Ethernet and be linked this way but in my case it would be nice to have the eLua Ethernet port dedicated to instrument control and the SBC port used to communicate with the outside and possible hostile world. If this is a logical thing to do, is there a protocol you would recommend for communicate from a limited number of GPIO pins on a SBC and an eLua MCU ? Hope everyone is doing great-Patrick _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
Tom Freund |
Re: eLua as intelligent expansion?
|
|
Not knowing any more details about your configuration do you have the ability to use SPI or I2C ?
On Tue, Feb 12, 2013 at 11:51 PM, Patrick <[hidden email]> wrote: Hi Everyone Tom Freund Dig.y.SoL (TM) "Systems overseeing public and private infrastructure"
Voice - 860-232-1614 Skype ID - digysol _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
In reply to this post by Patrick Mc(avery
Hi Patrick,
On Wed, Feb 13, 2013 at 6:51 AM, Patrick <[hidden email]> wrote: > Hi Everyone > > I normally write horrible and long emails but to try to avoid this... in a > nutshell I really wish I could have been more active with eLua over the past > year. I am spread very thin between my children's needs and my day time > work. > > I have a project that involves controlling scientific instruments and > crunching data afterwards. In another nutshell I have had a hard time > finding the right solution to the problems I am confronted with and have > looked into baremetal embedded, RTOS-like embedded and desktop solutions. It > now looks like all three together will be needed. > > So to get to the point(sorry), SBC like RPI and Olimex's IMX233 are very > nice but they don't have enough GPIO pins or peripherals for what I need. > Baremetal is nice but sometimes it's easier to have an OS, especially when > interacting with the user. > > I was wondering if anyone has tried to link a SBC together with an eLua > microcontroller? So for instance Rpi's 8 GPIO pins could be used for sending > data back and forth to an eLua MCU such as a Luminary development board and > Lua could do all the low level stuff. > > Both the luminary board and Rpi can communicate over Ethernet and be linked > this way but in my case it would be nice to have the eLua Ethernet port > dedicated to instrument control and the SBC port used to communicate with > the outside and possible hostile world. If this is a logical thing to do, is > there a protocol you would recommend for communicate from a limited number > of GPIO pins on a SBC and an eLua MCU ? Some serial protocol is probably what you need, something like SPI/I2C/U(S)ART, even if you have to bit-bang the GPIO pins. It's hard to give you a better answer without knowing the actual requirements of your applications. Best, Bogdan > > Hope everyone is doing great-Patrick > _______________________________________________ > eLua-dev mailing list > [hidden email] > https://lists.berlios.de/mailman/listinfo/elua-dev _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
Patrick Mc(avery |
Re: eLua as intelligent expansion?
|
|
In reply to this post by Tom Freund
Hi Tom, Hi Bogdan !
I guess I kept it too brief this time :( So most scientific instruments these days use ethernet. However GPIB and serial were universal not very long ago and there are many instruments that are otherwise perfect that do not use ethernet. Some instruments also use one-off RS232 solutions with special breakout boxes and so on. It would be nice to skip all this junk and control each from a single serial port. It would also be nice to mix instruments from different brands that use different communication methods. So I would like to create a board that had GPIB and about 3-6 serial ports. I might also like to set up A/D but I would need 20-24 bit resolution and it might be too expensive. Thanks again guys _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
In reply to this post by Patrick Mc(avery
Hi Patrick,
On Tue, Feb 12, 2013 at 8:51 PM, Patrick <[hidden email]> wrote: Hi Everyone What are your goals? If you need need to get the project up and running quickly, and money is available, then it's worth considering commercial solutions (such as GPIB to Ethernet interfaces ($200 at Sparkfun) or GPIB to USB, serial to Ethernet converters (boxes are available from many sources, and modules from companies such as Lantronix and Wiznet), etc. If you want to do protocol conversion, or have fun learning, then an eLua solution makes a lot of sense. Although the BeagleBone has a substantial amount of I/O, I've still seen people link them to Arduinos, so you're not the only one thinking about combining various boards. BTW, all the TI Cortex-M3 chips are marked as NRND (not recommended for new design); they're still available for ordering. TI should've handled this better; normally chips are marked NRND after the recommended replacements are available. _______________________________________________ _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
Quentin Barry-2 |
Re: [SPAM] Re: eLua as intelligent expansion?
|
|
Hello Tony
This is interesting...where on the TI website did you find that the parts are NRND? Where on the TI website can you still order them? If the LM3S6965 is still available for purchase directly from TI then I will order some more. However, the last time I looked under their Cortex offerings no mention was made of the M3 family at all.
Furthermore, all the example code to develop a serial to Ethernet gateway is just missing. I had the TI rep here, and he also could not find anything online. I was also informed that a local company here had designed one of the Stellaris M3 parts into one of their products. They are now struggling to get the components and the lead times are excessive.
Luckily I downloaded all the project files etc. a while back and I have 30 pieces of the LM3S6965 in stock. If I can get my hands on another 50 pieces at a REASONABLE price then I will complete my design and wait for the M4 which is going to be released later this year with a MAC and PHY on board.
As I understand things, TI bought out Luminary Micro. Rumour has it that the parts were buggy and people had to base their designs around certain revisions of the silicon or use the Errata sheets to overcome problems. TI thus scrapped the range and their in house engineers designed the Stellaris M4 from the ground up. So it is a completely new part. Personally I think they purchased Luminary Micro to get their hands onto a few patents.
Looking forward to your response as I have purchased all the necessary development kits and tools in order to complete my design. Now it looks as if it is wasted money. Maybe you can guide me in the right direction.
Quentin
On 15 February 2013 03:10, Tony <[hidden email]> wrote: Hi Patrick, _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
In reply to this post by Tony-12
Hello Tony
This is interesting...where on the TI website did you find that the parts are NRND?
Where on the TI website can you still order them?
If the LM3S6965 is still available for purchase directly from TI then I will order some more. However, the last time I looked under their Cortex offerings no mention was made of the M3 family at all.
Furthermore, all the example code to develop a serial to Ethernet gateway is just missing. I had the TI rep here, and he also could not find anything online. I was also informed that a local company here had designed one of the Stellaris M3 parts into one of their products. They are now struggling to get the components and the lead times are excessive.
Luckily I downloaded all the project files etc. a while back and I have 30 pieces of the LM3S6965 in stock. If I can get my hands on another 50 pieces at a REASONABLE price then I will complete my design and wait for the M4 which is going to be released later this year with a MAC and PHY on board.
As I understand things, TI bought out Luminary Micro. Rumour has it that the parts were buggy and people had to base their designs around certain revisions of the silicon or use the Errata sheets to overcome problems. TI thus scrapped the range and their in house engineers designed the Stellaris M4 from the ground up. So it is a completely new part. Personally I think they purchased Luminary Micro to get their hands onto a few patents.
Looking forward to your response as I have purchased all the necessary development kits and tools in order to complete my design. Now it looks as if it is wasted money.
Maybe you can guide me in the right direction.
Regards
Quentin
________________________________________________________________________________________________________________________________________________________________________________________
Hi Patrick,
On Tue, Feb 12, 2013 at 8:51 PM, Patrick <[hidden email]> wrote:
Hi Everyone I normally write horrible and long emails but to try to avoid this... in a nutshell I really wish I could have been more active with eLua over the past year. I am spread very thin between my children's needs and my day time work. I have a project that involves controlling scientific instruments and crunching data afterwards. In another nutshell I have had a hard time finding the right solution to the problems I am confronted with and have looked into baremetal embedded, RTOS-like embedded and desktop solutions. It now looks like all three together will be needed. So to get to the point(sorry), SBC like RPI and Olimex's IMX233 are very nice but they don't have enough GPIO pins or peripherals for what I need. Baremetal is nice but sometimes it's easier to have an OS, especially when interacting with the user. I was wondering if anyone has tried to link a SBC together with an eLua microcontroller? So for instance Rpi's 8 GPIO pins could be used for sending data back and forth to an eLua MCU such as a Luminary development board and Lua could do all the low level stuff. Both the luminary board and Rpi can communicate over Ethernet and be linked this way but in my case it would be nice to have the eLua Ethernet port dedicated to instrument control and the SBC port used to communicate with the outside and possible hostile world. If this is a logical thing to do, is there a protocol you would recommend for communicate from a limited number of GPIO pins on a SBC and an eLua MCU ? Hope everyone is doing great-Patrick What are your goals? If you need need to get the project up and running quickly, and money is available, then it's worth considering commercial solutions (such as GPIB to Ethernet interfaces ($200 at Sparkfun) or GPIB to USB, serial to Ethernet converters (boxes are available from many sources, and modules from companies such as Lantronix and Wiznet), etc. If you want to do protocol conversion, or have fun learning, then an eLua solution makes a lot of sense. Although the BeagleBone has a substantial amount of I/O, I've still seen people link them to Arduinos, so you're not the only one thinking about combining various boards. BTW, all the TI Cortex-M3 chips are marked as NRND (not recommended for new design); they're still available for ordering. TI should've handled this better; normally chips are marked NRND after the recommended replacements are available. ______________________________ _________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/ mailman/listinfo/elua-dev _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
Patrick Mc(avery |
Re: eLua as intelligent expansion?
|
|
In reply to this post by Tony-12
Hi Tony
Thanks for responding to my post! My project requires a lot of explaining and could bore people to death but here are the basic goals and an outline of the problem: Goals 1)To enable scientists to control their instrumentation, to collect data and to process that data 2)To make a revenue stream out of this 3)to use technologies I like 4)To use technologies I am competent with. Outline of the Problem The normal scenario for this is: 1)A company writes windows based software and uses a proprietary interface board(usually GPIB) 2)A company does the same but uses an off the shelf board and suffers with a lot of piracy 3)The same as above but they use serial The market is now moving towards ethernet but there is still millions, if not billions of dollars worth of equipment that uses GPIB/RS232. Ethernet based vendors are moving towards remote authentication, like Microsoft. I don't want to employ this. I switched from Windows to Linux about 10 years ago. Windows does not seem stable for anyone and I would have a hard time providing technical support on this platform. For the longest time, I thought about doing the same as these other companies but doing it on Linux. However I can't see a revenue model. There are other options but I would prefer the GPIB card option and that requires linking code to the kernel and that means GPL which is the worst thing for me as I will obey it but I cannot enforce it afterwards due to the locations of my customers. I have thought about an all eLua device but I can't see how the scientists could crunch their data afterwards. I have thought about SBC computers running Linux or BSD but the I/O is a problem. Now I am thinking of writing a desktop application to crunch the data and a SBC + eLua solution for instrument control and data collection, splitting the problem in half. I could link the SBC and eLua via I2C or SPI as it would ship a neat tidy product. However I am concerned that the GPIB portion will overflow the data abilities of these protocols. Your idea of mixing already existing devices could work, I think I would just have to work out the packaging of it. Creating an eLua based board would also help the revenue model though as there would be one more thing that would have to be re-engineered and not just purchased off the shelf, and that could slow down competitors. Thanks again P.S I would like to mention that Quentin sent me an incredibly detailed overview of a variety of devices. I hope to get back to the tutorial series I was not able get done. I am armed with much more information this time around. _______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
|
|
In reply to this post by Quentin Barry
Hi Quentin,
I don't have any secret line to TI, but if you're looking for the LM3S6965-IQC50 you're in luck, since both Mouser and Digikey have >2K in stock (but they're a little pricey; other versions aren't as available). I found the NRND discussion at TI's forums. In that discussion, TI claims that the lead times issues aren't related to the NRND status; I should note that we have a custom board using a TI 320C6201 DSP (introduced in 1997) and are still able to get it without problems. TI should have at least have some example code, since they sell the Stellaris Serial To Ethernet reference design kit, and it seems to come with some demo software... --Tony On Thu, Feb 14, 2013 at 11:46 PM, Quentin Barry <[hidden email]> wrote:
_______________________________________________ eLua-dev mailing list [hidden email] https://lists.berlios.de/mailman/listinfo/elua-dev |
«
Return to eLua Development
|
1 view|%1 views
| Free forum by Nabble | Edit this page |