RTOS/CC1310: 15.4-stack Long range mode SDK 2.20.00.38 not working

Hi ON,

How many sensors do you have in the network and what mode are you running (non-beacon, beacon or FH)?

M-W

In the simplest use case I cannot get ONE sensor to join ONE coordinator.
I use non-beacon mode.

Some more info is that when trying the packet sniffer in 5 kb/s mode, we never see the beacon request from the sensor.
So it seem that the "end device" is not handling this OK.

However on the spectrum analyser we did see the packets.
So if seem that with new SDK, that some parameters of the LRM is changed in device mode.

I notices that the change log says "EZLINKPROP-344 Enable whitening for SLR mode"
Is SLR mode = simplelink long range mode?
This could be realted to the issue we see. (at least explain why we do not see beacon request in packet sniffer)
But when communication between two devices, both made from new SDK, then I do not see how this is should be an issue...

ON

Hi ON,

The LRM mode joining times can become quite long (>10 minutes) depending on what PHY you are using LRM for, what frequency band are you targeting?

In some cases this is simply due to the extended radio time of the LRM conflicting with the back-of mechanisms.

Can you try to go into the config.h file and bump the CONFIG_MIN_BE and CONFIG_MAX_BE exponents to 5 and 8 respectively?

M-W,

I have investigated meanwhile.

With the new SDK, LRM @ch 0 on 868 MHz, I see that any attempt to send (either association request from slave or a broadcast from the coordinator) leaves gives the failure code 0xE1. According to documentation this means that MAC_CHANNEL_ACCESS_FAILURE =The operation or data request failed because of activity on the channel.
So when association fails with this message it seem to not be timeout message. (I also use only one channel in my channel mask, so joining should be quicker than...)

Could there be the issue that background noise is above LBT threshold, and this is change in latest SDK?

I will look into the CONFIG_MIN_BE and...MAX....

ØN

M-W,

I am using the co-processor project and config. h is not part of that...
Anyway, CONFIG_MIN_BE and ...MAX... was not changed from SDK 2.10... to SDK 2.20...

My simplest test case is then: Start coordinator on ch 0, 868 MHz 5 kb/s LRM. Send a broadcast message:
SDK 2.10.... Packet seen on sniffer- Status = 0x00 = OK
SDK 2.20.... No packet seen on sniffer. Status = 0xE1= Not OK

Even tested without antenna and terminated in 50 ohm to verify that external noise is not generating problem. Did not help...

-ØN

Hi ON,

I will look into this and see if I can find any changes that should impact LRM between these two SDKs. Could you verify what max and min exponents are set per default in your case by reading out the MAX PIB Attributes values for MAC_MAX_BE and MAC_MIN_BE?

M-W

MAC_MIN_BE = 5
MAC_MAX_BE = 8

BR
ØN

Hi ON,

That should be ok, I will need to look more into this and get back to you with an answer.

Also, would it be possible to share the script you are using for testing this out?

Hi ON,

We tried this out by running a 2.20 SDK sensor and connecting to a 2.20 SDK co-processor collector using 868 LRM non-beacon mode and we can not see any problem. Is this similar to what you are testing?

If this is true, could you share a UART log dump of the co-processor when seeing this issue?