Jee Labs Talk

This forum has moved to forum.jeelabs.net - this read-only archive is for reference only!

1. mars
   

   JC,

   Since the JeeNodes might function very well within a wireless DOmotica environment, have you ever considered to use the RFM31B receiver and RFM43B transmitter as the subjects for new 'Plugs' ?

   

   They can be programmed at 240-930Mhz, support both ASK/OOK and FSK modulation at very high bitrates (upto 256kbit) and seem to be very sensitive. Size-wise they are the same as the RFM12B, so that shouldn't be problem.

   Assuming the RFM12B can be combined with these receivers/transmitters on a single JeeNode using the 'Plugs' system and since you already have so much experience with the RFM12B chip, it might be a nice addition to your shop and to the ever growing functionality of the Jeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee family!

   Posted 17/05/2010 18:48:33

2. jcw
   

   Interesting idea. What's the interface? I'll look into it. Thx for the tip!

   Posted 17/05/2010 20:12:53

3. mars
   

   As far as I understand it, it has a SPI interface for settings, and a serial data interface for... the data! So seems about the same to me as the RFM12B.

   This might be to much for connecting to a standard port (due to the SPI interface), but I might be wrong. Maybe the Carrier Board is required to hook these things up?

   The Conrad receivers/transmitters have just the serial data input/output, so have no problem to be connected to a standard port.

   BTW: My quadrature decoder seems to work, at least in a simulated envirnonment using port change interrupts on digital outputs which are set HIGH/LOW in the main loop simulating CW/CCW behaviour of the electricity disc.

   Next step is to make it more generic and standardize on packet format send to be usefull at the other side... I might use the EMxx format and simulate the EM-1000-S electricity (and EM-1000-GZ gas) units...

   Posted 18/05/2010 07:11:45

4. mars
   

   Hmmmmmmmm, can't edit my post anymore...

   According to the EM20 Protocol negative numbers don't seem possible...

   Must be designed by ppl with fossile heads never heard of solar systems...

   Posted 18/05/2010 08:19:36

5. jcw
   

   Emulate two units? One sends positive values, other one sends negative values? I.e. send max(v,0) and max(-v,0) ?

   Posted 18/05/2010 10:16:49

6. mars
   

   Emulating two units is a possibility: 1 for consumption, 1 for production. Both are positive values in that way. However, I also want the total consumption, and that one an still be negative.

   Maybe its better to invent JLSensors (JLPower, JLGas, JLWater, JLCounter, JLTemperature, etc.) analogue to the RFXSensors and apply the concept of virtual sensors based on physical sensors to get 5-minute data, hourly data, min/max or high/low, on-demand, triggered by change etc.

   I got a bunch of 1-Wire DS18B20 temperature sensors coming in also from Thailand, and it would be nice if connecting them to a JeeNode would be "just connect them" and everything else is automatic...

   A well, first things first: lets send the data based on a change and see where that leads me ;-) analoguethe total consumption could still be negative

   Posted 18/05/2010 12:04:46

7. epokh
   

   Speaking of which. My current project is developing a standard sw interface for doing the things you are saying. There are many brands/sensors with different protocols. Our XML interface enforce the data format for exchange. I can post only more info when we have something to show, at the moment is all secret. :-p

   Posted 18/05/2010 13:28:58

8. mars
   

   Nice! But secret? Send me the data, pretend you got hacked, and we all won't be bothered by secrecy anymore and can all enjoy the uniform format ;-)

   But for an XML solution I would simply use xPl so I could use al the xPL related back-office software. For a more line-friendly (but less readable by humans) approach, I would simply use some custom compact but structured binary (or ascii) variable length format as used by many other sensor builders.

   Addressing the sensors should be unique: nodeId (from the JeeNode) and sensor# would be the easiest way, and a minimal requirement to avoid collisions while storing them in a central database or simple excel sheet (which also works very, very well!!).

   This weel seems to be invented again and again and again, but I still can't find an 'Arduino standard' for sending sensordata in an interchangeable format...

   Posted 18/05/2010 14:08:03

9. jcw
   

   > "The nice thing about standards is that you have so many to choose from." - Andrew S. Tanenbaum

   I'd say, let the tools you want to use dictate which standards you follow.

   Posted 18/05/2010 14:31:52

10. mars
    

    Exactly! @work I always use the standard that works most for me, so I have time to do nice things.

    And on-topic once more: I will of course use the Conrad 433/868Mhz recievers/transmitters AND still think the HopeRF units would be a very nice addition to the shop & the world of JeeNodes...

    Posted 19/05/2010 09:52:25