@ddos Yes, that would work - if VIN1 is pin 5, VOUT1 is pin 6, VIN2 is pin 2 and VOUT2 is pin 3, that is.

It's a bad habbit to have "double pins" in the schematic. You should really make a pin for each ground connection - just my 2 bits.

BR Jonas

It works in both directions. The board is literally a rip-off of this board (even the description): https://www.sparkfun.com/products/12009 . Read the hook-up guide there :-)

BR Jonas

@LastSamurai Yes, thats true. My order is always above the "free shipping" limit :-)
I'll give the datasheets a look, of the devices you mentioned.

BR Jonas

@LastSamurai said:

AT25DF512C

Well not in EU, but you can order from DigiKey or Mouser if everything else fails - thats where I order my parts.

EDIT: And I live in DK - maybe worth mentioning.

BR Jonas

Yes. The two power supplies need a common reference, so the grounds need to be connected.

BR Jonas

@mfalkvidd Yes, sure!

Thank you very much. I think that I will order a couple of sub-GHz modules to do some testing with, but stay with the nRF modules for now - as you say, they are cheap.

BR Jonas

I have updated the schematic to rev. 1.1.
The following has been changed:

• Si1133 has been replaced by Si1145
• The mosfets are removed from the radio power path
• Radio pins shuffled around a bit
• FTDI connector removed from PCB and instead I made a programming jig
• All I2C sensors are powered from the same pin of the uC, as well as the I2C pull-ups (haven't decided which pin yet, still routing the PCB)

Comments are more than welcome.

BR Jonas

@scalz Yes, the Vishay one is more precies indeed, but my main reason for putting the Silabs one on the board, was to detect the ambient light level.

Reg. the mosfet, yes I could have used a single fet. I think the reason why I did it like this, was that in a previous project, I was controlling higher voltages, hence the need for the n-fet.
Anyway, I'm not sure if I will keep it or if I will just let the radio be powered on, all the time, in the next rev.

BR Jonas

Hi all,

what are peoples experiences. On a longer timescale, is the recommended radio the HopeRF one, or could it be Nordic as well? Because of the lower frequency of the HopeRF radios, range/penetration could better, compared to 2.4GHz, but are there any tests that prove this?
What is the general opinion about radio selection?

BR Jonas

I use the N-fet to drive the P-fet, which powers the radio module - again, I'm not sure if its worth it, but when the power comes from a coincell, every nA counts.

/Jonas

Yep, thanks. Found out the hard way :-/ I'm not sure if I really want to power the sensors from the micro. I have to turn on every sensor to be able to use the I2C bus, so I cant turn them on selectively, as I had hoped.

BR Jonas

I have uploaded the Rev. 1.0 schematic. I'm currently making some small changes, to be able to use the RF69 module as well as the nRF24L01+.

Firmware upload doesn't seem to work :-/

BR Jonas

Ok, I tried to breadboard the sensebender micro and use the pre-build MYSBootloader - it works.
That indicates that I didn't manage to build the bootloader myself :-/

I will look into it later. Thanks for your replies.

BR Jonas

And the log from MYSController around the restart:
"12-03-2016 11:14:34 TX 4;255;4;0;3;C800010046077E2B9A29B329A529F6290D0A006E616E
12-03-2016 11:14:34 RX 4;255;4;0;2;C80001004607
12-03-2016 11:14:35 TX 4;255;4;0;3;C8000100470700696E66006F7666002E0000FFFFFFFF
12-03-2016 11:14:35 RX 4;255;4;0;2;C80001004707
12-03-2016 11:14:36 DEBUG Undefined firmware/type for node=4
12-03-2016 11:14:36 INFO BL version=65535
12-03-2016 11:14:36 INFO Send FW info to node 4: type=C8, version=1, blocks=0x0748, CRC=0x5FFE
12-03-2016 11:14:36 TX 4;0;4;0;1;C80001004807FE5F
12-03-2016 11:14:36 RX 4;255;4;0;0;FFFFFFFF
12-03-2016 11:14:36 DEBUG FW update started, node id = 4
12-03-2016 11:14:36 TX 4;255;4;0;3;C800010000000C9467030C94A6270C94CD270C948F03
12-03-2016 11:14:36 RX 4;255;4;0;2;C80001000000
"

BR Jonas

This is the last lines:
"<- 4,4,0,50,196,2,255,10,0,1,0,70,0,
-> 0,0,4,186,196,3,255,10,0,1,0,70,0,44,2,14,148,112,0,192,224,208,224,14,148,139,0,32,151,160,
<- 4,4,0,50,196,2,255,10,0,1,0,71,0,
-> 0,0,4,186,196,3,255,10,0,1,0,71,0,225,243,14,148,0,0,249,207,8,149,248,148,255,207,255,255,160,
fw nok
Go"

Is there any special requirements to the fusebits of the micro?

BR Jonas

The sensor is ~2m from the gateway so a bad link is not impossible, but highly unlikely I guess.
Is any form of checksum of each individual packet performed?

BR Jonas

Hi all,

I have flashed my atmega328 with the MYSbootloader. The device shows up in MYSController v. 0.1.2.282, and I can assign a new firmware to the device.
It starts transfering the firmware, but when the transfer is complete, it just restarts the transfer.
Anyone has a clue on what is wrong?
I have compiled the bootloader myself, because the CSN-pin of my radio module is connected to PB0 instead of PB2. Also I'm running on the internal 8MHz oscillator.

Thanks in advance.

BR Jonas

@jenbaker Hi,
well, it could be possible to embed it in a RFID reader, but it isn't really the purpose of the device.
It is intended to be a stand-alone environment sensor.

BR Jonas