Re: 96boards Nitrogen


Steve Brown
 

Hi Vanayak,

I've got some more information.

I converted the SDK's radio_test to run on both the nano2 and the
nitrogen.

The nano2 is transmitting continuous carrier (test "c") on 2402 and the
nitrogen on 2426. TX power is -4 dBm. The data rate is 1M.

The attached sample was from ubertooth-specan-ui. Lots of jitter at
2426 with both nitrogen boards.

Steve

On Mon, 2017-09-18 at 05:04 +0200, Vinayak Kariappa wrote:
Hi Steve,

Good. You have clarified that indeed radio h/w is live and kicking.

May be something around the antenna matching is bad that only some
peer see the transmissions.

I can be suspicious that the Factory Configuration Information is
probably erased (like a full erase done using nrfjprog). You can
check if the identity address printed on  "init" in shell application
remains same after power resets, if so, then I have no more ideas.

Few of us are available at #zephyrproject on freenode.net, in case
you prefer to IRC.

- Vinayak

On Fri, Sep 15, 2017 at 7:27 PM, Steve Brown <sbrown@cortland.com>
wrote:
More

On Thu, 2017-09-14 at 19:54 +0000, Chettimada, Vinayak Kariappa
wrote:

Hello and blinky work. BT does not.
Ok, so the CPU and UART works :-)
And, the 32KHz clock domain works, hence the blinks.

Now, CPU can run from internal High Frequency RC or External
crystal.
If your 32MHz crystal on the board has dry solder (or what ever
has
happened), then most probably internal RC remains to clock the
SoC.
But 2.4GHz radio will not be 2.4GHz.

Zephyr v1.9 has Radio Direct Test Mode implementation, hci_uart
sample can be used with DTM HCI commands to test the radio. The
DTM
HCI commands will fail to respond if the HF clock cannot source
from
a stable crystal.


And, yes nitrogen_beacon.hex from http://builds.96boards.org/re
leas
es/n
itrogen/zephyr-1.8/ flashed to the nano2 advertises. There's no
console
output as I explained previously.

I'm now more convinced that the devices somehow got damaged in
transit
from Seeed. 
Very unlikely (and likely based on conditions), I use nRF51
dongles
over 5 years now, carry it around in bare hands, worst I have
done is
to have my 32KHz crystal go out of its spec. but BLE operational
using the 32KHz RC.
 

I'll report when I get the ones I ordered from Digikey. I
expect
them
to work just fine.

If you are curious about what happened to the radios, I'd be
glad
to
send you the 2 I got from Seeed.
I think Seeed need to contact our support (I am just a software
guy).


The shipment didn't seem to have been opened. I'm pretty ESD
conscious
at my end.

Steve
These tests were run using tests/bluetooth/shell.

I found and fired up an ubertooth. The 3 advertising freqs are spot
on.

Whatever is on the air isn't decoded by ubertooth-btle or bluez w/
a
CSR (0a12:0001) dongle.

So, I tried your iOS nRF connect app. It picks up the
advertisements. I
can also connect from my iPhone.

All 3 applications function with the nano2.

Next I tested with a rpi3. It sees the advertisements and can also
connect to the nitrogen. Curiously, ubertooth picks up the SCAN_REQ
from the rpi3, but not the responses from the nitrogen. All the
devices
are within 1 meter of each other.

So, I have 2 apps (nRF connect & rpi3) that see what the nitrogen
puts
on the air and 2 apps (ubertooth & x64 w/ bluez & CSR dongle) that
do
not.

Versions

ubertooth:
  Firmware version: 2017-03-R2 (API:1.02)
  ubertooth 2017-03-R2 (dominicgs@hydrogen) Mon Mar 13 16:09:02 MDT
2017

x64 w/ bluez: V5.47

nRF Connect: V1.7.5

rpi3: raspbian Linux raspberrypi 4.9.43-v7+ w/ bluez V5.43

Can you think of any possible explanation?

Steve





_______________________________________________
Zephyr-users mailing list
Zephyr-users@lists.zephyrproject.org
https://lists.zephyrproject.org/mailman/listinfo/zephyr-users

Join users@lists.zephyrproject.org to automatically receive all group messages.