One Wire?
#1
Posted 08 April 2012 - 03:24 PM
I read in the introduction thread http://forums.netdui...ng-netduino-go/ that one-wire will be available via the shield base.
I wondered will 1-Wire be officially natively supported with the go and its shield base module?
Also are there any plans for a dedicated 1-Wire bus module?
I was thinking that someone a little more technical than I may already have some plans in the pipeline for this, but I also had some ideas for such a module...
Could it be possible to make a one wire module with perhaps 8 connectors (all joined on the same 1-wire bus) but perhaps with a shift register (which I believe each module will have as standard now?) that detects when a device is plugged in to 1 of the 8 connectors in order to automatically identify and register a one wire device to a specific port (maybe with the same led idea as the main GO board). This would make it easy if you add/remove/change connected devices to easily identify the correct device in your code. Also as it would all be on the same bus you would still be able to address devices simultaneously (such as DS18B20 temp probes).
Netduino GO looks amazing, now I just need to work out how to fit my aquarium project in with this, oh and buy a GO which is going straight to the top of my shopping list!
Thanks for the hard work on the GO everyone, can't wait to get one!
Andy
#2
Posted 08 April 2012 - 05:10 PM
My .NETMF projects: .NETMF Toolbox / Gadgeteer Light / Some PCB designs
#3
Posted 08 April 2012 - 06:58 PM
#4
Posted 09 October 2012 - 02:47 PM
#5
Posted 09 October 2012 - 04:02 PM
You could do all the processing on the device as Stefan mentioned, and then return the data via a virtual AnalogInput. The easiest solution will be to hook up the OneWire device to the on-module MCU, virtualize the pin, and then write your code in C# in your mainboard driver.
What do you actually mean by virtualizing the pin, as I understand it, you "cheat" the go mainboard to communicate with a pin, that is not a pin on the mcu, but a pin on another module emulated in software, and materialized on the go module's mcu's pin.
As of digital and analog pins I can see how that can be done, as they are read from or written to, for PWM, expect the hardware PWM function on the go modules mcu is used, right ?
As for serial, spi and i2c, that must be the hardware on the other end too, right ?
Now here comes the question, what does a go module do when the data clocks in on the spi, i2c or one wire pin on the local mcu, and the go mainboard is doing something else ?
Answer is probably buffering them, but if the logic is implemented on the go mainboard and the pin is only virtual, 0 and 1 just clocks away, or am I completely stupid?
- Ulrik
#6
Posted 09 October 2012 - 04:08 PM
--
Asbjørn
#7
Posted 09 October 2012 - 06:10 PM
Would be interresting to get more insight on this yes.
But I guess, that the cpu on the module/shieldbase need to have enough capability to run the full protocol in hardware, that, spi, i2c etc, and the data is sendt back using spi to the Go mainboard, in chunks. (I think I read about frames somewhere here?)
I would assume that it has to be the full protocol on the shieldbase or otherwise we would have I2C (as well as everything else) available today.
#8
Posted 09 October 2012 - 06:14 PM
--
Asbjørn
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users