Divorcing the arduino and ARM routers from the P2
Patter
Posts: 13
I get that for now, the xbee 802.15.4 and other wireless communication "modules" are necessary because that's the only way we have, for now, to get wireless communication working on the various wireless protocols. I understand the convenience but, am I the only one that cringes when the solution is to connect the P2 via SPI to an arduino or other Microcontroller to route the SPI frames over 802.15.4 and have the that microcontroller route it for the P2 because we either can't or just haven't coded the 802.15.4 object? I know, it would be a big project. People say that now the P2 has the xbee module it can now transmit via 802.15.4. I'm not convinced, I see it as sending the traffic via SPI to the other computer that routes the traffic over 802.15.4 for the P2. I'm just thinking about all those idle cogs.
I would like to hear other peoples thoughts or speculations on this state and hear responses to the following questions.
Can protocols like 802.15.4, bluetooth, LoRa for example, be implemented on one or two of the cogs in the P2?
If it can, how much Transceiver hardware would be necessary at that point?
Would it be significantly more energy efficient to implement the processing in the p2 with the necessary hardware than communicating over SPI and powering that external module? Would the juice be worth the squeeze for battery powered devices?
I have not heard anyone say when talking about communication modules with other microcontrollers in them that this is just a quick way to get this done until we implement it in one of the cogs on the P2. Maybe there is a reason for this. I think it would be interesting to have the P2 able to do the processing for bluetooth, wifi, xbee etc. with only minimal necessary transciever hardware to keep either energy cost down for battery powered devices, keep costs down for hardware or just the bragging rights being independent of other microcontrollers. I get that getting other microcontrollers involved simplifies the coding to SPI and awareness of how to interface with the module but in the long term do we really need to get other microcontrollers involved?
Someone please shoot this idea down, I don't know how.
Comments
I don't get it. XBee, ESP8266, ESP32 have nothing to do with Arduino. They are products produced by vendors that are trying to solve a communication problem.
You can't just build a WiFi or Bluetooth radio and call it good. These units have to be tested and certified or you can't sell them.
Mike
...
Since you asked for it..
It's dead simple - cost/benefit.
Too much effort in doing it for too little gain, if any. As you rightly said, a hardware RF layer will still be needed anyway. And ready made, ultra cheap and certified modules (I am speaking in high volume terms) are widely available so your idea doesn't seem to be economically feasible to me. I estimate the development and certification costs would be in tens (if not hundreds) of thousand of US dollars and that has to be at least regained somehow in the end.
In the case of LoRa, which I'm now looking at for multiple projects, I see no good reason why you can't/shouldn't attach a module with an SX1262, SX1276, etc. transceiver and its requisite support circuitry directly to a P2 to achieve your wireless. The interface (data and commands sent via UART or SPI) isn't all that complicated from what I've seen.
thanks
SDR hardware is a lot more expensive than dedicated hardware. Could we run a local quadrature oscillator through a quadratic mixer from a couple pins in DAC mode to transmit a digital signal? Sure. Could we take the same local oscillator and the same quadratic mixer, and run it as a receiver into a couple pins in ADC mode? Sure. But, that local oscillator and quadratic mixer are a lot more expensive than a dedicated 802.15.4, 802.11, or BLE chip.
Honestly, it's probably much easier to program a bridge firmware into an integrated application processor/radio than to try to implement those protocols directly on the P2. I've probably spent more time than is healthy looking into that.
I think the closest thing to what you're asking for would be something like the SX1262 RobertF mentioned, or a DecaWave DW1000 radio, since that has no microcontroller, just a set of SPI-accessible registers for the radio.
Looking at other Semtech radio parts, the less integration, the more expensive the chip. For example, the SX1257 is a straight I/Q modulator/demodulator that would be very nicely driven by a quad of smartpins. It's 2% cheaper than the SX1262 which has a whole protocol/packet engine.
double thanks. I really appreciate your guys inputs. I'm a newbie so I don't realize the problems or cost vs rewards my ideas run into.