Efergy e2 power meter

[korttoma]

I have now received the RFM12b module, unfortunately I will not have time to set it up for the next 2 weeks but until then.
Could you clarify how I need to wire it?
Is it really enough to have RFM12B Data pin routed to D3?
Do I need to somehow reprogram it using ISP to work in OOK mode?

Did you have a chance to try any cheep 433MHz receiver yet?

[gtortone]

@korttoma

Hi,
you need to connect RFM12b module to Arduino following this schematic (you can avoid to connect IRQ line...)

http://jeelabs.net/attachments/download/331/jlpcb-128.pdf

after this cabling you need to connect RFM12b DATA pin to Arduino D3 pin (you can select a different pin but modify
code accordingly...)

only after this you can use the Mysensors-Efergy-e2 sketch (https://github.com/gtortone/MySensors-Efergy-e2) to get
RF data;

I tested my code with cheap RF modules (I received them some days ago) but the performance of these modules are very
poor... I will try to do other work on cheap RF modules in the future...

Thanks !

[korttoma]

@gtortone thanks for the clarification on the wiring of the RFM12b.

In what way was the performance of cheap RF radio poor? Is the range just bad or does it miss transmissions? The first cheap radios I got also had poor reception, this is why I upgraded to the RXB6 from my earlier post. Can you share the code you used with the cheap RF radios maybe my RXB6 module performs better?

[gtortone]

@korttoma

Hi, I have these modules:
http://www.banggood.com/433Mhz-RF-Transmitter-With-Receiver-Kit-For-Arduino-ARM-MCU-Wireless-p-74102.html?rmmds=search

the range is only few centimeters... I will try to do some antenna tuning but I don't expect any significative improvement.

If you want to try with your RXB6 module the code is always the same, you have only to remove the configuration code lines related to RFM12b configuration;

[korttoma]

@gtortone that is the same useless junk I got from e-bay at first, range is terrible.

This gives me some hope to get this working with the RXB6. I will test this as son as I can (might be a few weeks until I can).

[korttoma]

Hi @gtortone ,
I finally found some time to hock up the RFM12B. Seems to be almost working. I just get the value "inf" for current. Do I have some datatype problem or maybe it should be calculated differently my Efergy engage hub?
I'm using a Arduino Pro Mini 3.3V

Attached the serial print.0_1484211281663_efergyLog.txt

I should see values around 100W most of the time and some smaller down to 30W.

Edit: It seems like it might be payload[6] that is equal to 255 at 100W and 254 at under 40W that is causing problems with the calculation.

Edit2: Took 3 samples at around 30W, 60W and 100W. 0_1484220204136_efergyLog2.txt I need to find something so that I could test higher values to so I could figure out the formula to use.

[gtortone]

Hi @korttoma,
happy to read that you are getting "reasonable" values from RFM12B !!

but I agree that in your case the decoding protocol needs some refinements...
just a question: your line voltage is 110 or 240 ?

take into account that from Efergy current sensor you get the value of "current" and
not "power" than for 30W expected value from data is (V/30) [Ampere]...

[gtortone]

Hi @korttoma,
take a look to this decoding based on your collected data for 30W, 60W and 100W;

everything seems to be ok !!

Regards,
Gennaro

p.s. I checked in your profile that you live in Finland and power voltage is 220/240 V :-)

• bytes #1: 0x09 fixed value
• bytes #2: 0X50 fixed value
• bytes #3: 0x2E fixed value
• bytes #4: 0x40 fixed value
• bytes #5: current MSB
• bytes #6: current LSB
• bytes #7: exponent
• bytes #8: CRC
• current detected = ((MSB << 8) + LSB / 65535) / pow(2, (255-exponent))

30 W = 0.13 A (@220V)

9 - 0x09
80 - 0x50
46 - 0x2E
64 - 0x40

71 - 0x47
172 - 0xAC
254 - 0xFE
184 - 0xB8

0x47AC = 18348

current = 18348 / 65535 / pow(2, 1) = 0.139 A

60 W = 0.26 A (@220V)

9 - 0x09
80 - 0x50
46 - 0x2E
64 - 0x40

69 - 0x45
29 - 0x1D
255 - 0xFF
40 - 0x28

0x451D = 17693

current = 17693 / 65535 / pow(2, 0) = 0.269 A

100 W = 0.45 A (@220V)

9 - 0x09
80 - 0x50
46 - 0x2E
64 - 0x40

112 - 0x70
163 - 0xA3
255 - 0xFF
217 - 0xD9

0x70A3 = 28835 (10)

current = 28835 / 65535 / pow(2, 0) = 0.439 A

[korttoma]

Thanks @gtortone your formula seems to work. I just need to find a way to test higher values.

I almost had the same formula but I did not have the "65535". Could you explain where you got this value from?

-- S --
9
80
46
64
76
204
254
221
-- E --
abc_value = 19660current = 0.15A - power = 34.50 W
23701 TSF:MSG:SEND,9-9-0-0,s=9,c=1,t=17,pt=2,l=2,sg=0,ft=0,st=OK:34
OK = 4 - short pk = 0 - long pk = 0 - crc bad = 0
-- S --
9
80
46
64
71
173
255
186
-- E --
abc_value = 18349current = 0.28A - power = 64.40 W
29693 TSF:MSG:SEND,9-9-0-0,s=9,c=1,t=17,pt=2,l=2,sg=0,ft=0,st=OK:64
OK = 5 - short pk = 0 - long pk = 0 - crc bad = 0
-- S --
9
80
46
64
115
51
255
108
-- E --
abc_value = 29491current = 0.45A - power = 103.50 W
35688 TSF:MSG:SEND,9-9-0-0,s=9,c=1,t=17,pt=2,l=2,sg=0,ft=0,st=OK:103
OK = 6 - short pk = 0 - long pk = 0 - crc bad = 0

[gtortone]

Hi @korttoma,

the value 65535 or 32768 often is used as multiplier from sensor that send metric;
usually when you have to send a "small" float value (temperature, current) you can send it as integer
multiplied by a big constant instead of sending it with a float value converted in binary...

I just refined my formula with this that is more close to formula I found for my Efergy using a small
constant:

current = 18348 / 32768 / pow(2, 255-254 + 1) = 0.139 A
current = 17693 / 32768 / pow(2, 255-255 + 1) = 0.269 A
current = 28835 / 32768 / pow(2, 255-255 + 1) = 0.439 A

[korttoma]

Hi @gtortone I'm working on finalizing this node and I'm having some trouble with stability. The node stays alive from a few minutes to 30min and then stops communicating.

What kind of decoupling capacitors are you using in your setup?
I now have 2 pc 0.1uF, 2 pc 47uF and one 0.33uF.

I use a HLK-PM03, a Pro Mini 3.3, and the NRF24L01+

Debugging output don't give me much to go in since it just stops responding.

[gtortone]

@korttoma

Hi,
at the moment I'm defining some things in order to use RFM69 modules...
in the past, on my NRF24L01+ I was using 10 uF capacitor but I had same problems of "stop responding"...

[korttoma]

@gtortone My node has now been running for more then 40min since I added the 0.1uF and 0.33uF tantalum caps. I'm hoping this will be the solution for me since I have no tools to debug any disturbance like this.

EDIT: It failed again :(

[korttoma]

Tried exchanging the NRF24L01+ with an RFM69 but it does not seem to work at all with this setup. Might be because the RFM69 is using the same frequency as the RFM12B or maybe I can not have the 2 in the same node because they are related.

[korttoma]

@gtortone did you have any chance to look at the "stop responding" issue of this device?

I tried to add some more debug prints to your code today but if I add them to the wrong places the efergy receive function stops working and receives only 0 value. I guess it messes with the timing.

[gtortone]

@korttoma Hi, I just returned to play with Efergy and Arduino. I realized that debug messages on UART reset the micro. Now I'm skipping (with a #define DEBUG and #ifdef) all debug messages and it seems to work fine. But take into account that now I'm not using any radio module... just to test the firmware.

[korttoma]

:D What a coincidence. I returned to this project today also. Now trying to set this up as an Ethernet GW with local sensor without NRF radio. I still have to share the SPI bus with the Ethernet module though. I'll let you know how it goes.

I think I did try removing the serial prints earlier but I'm not sure..

[korttoma]

The ethernet GW approach performs even worse than your original solution. I get one or 2 measurements and then nothing. Still responds to ping but no communication with the controller.
I don't know how to debug this since it just stops working. Is there by any chance anything in the code that consumes memory?

[ailiev]

Hi All,

First of all, great work man!

I was wondering do you know if we need to have established link between the wireless sensor and the receiver?

I only have the wireless sensor so I was wondering if I can use this without the receiver?

Any thoughts?

[korttoma]

@ailiev it "works" without the receiver but for me the node fails after about 1h so I gave up. I could try again now with the new MySensors 2.2.0 lib to see if it makes any difference (when I find some time).