XIAO nRF52840 & Wio-SX1262 Kit failures

Products & Technology XIAO
, ,
21

Hi there,

Awesome, Yes he is very attentive to the LIB, I wish more were like that :grin:
Nice, narrowing it down. I’m confident you’ll get to the root, and Heck with all the experience and work. You’ll know the solution already. :+1:
Meanwhile , can’t hurt.
RadioLib relies on:

  • DIO1 interrupt → sets flag → main loop processes it

If:

  • ISR fires
  • BUT semaphore / flag handling fails (you mentioned this!)

Then the CPU:

  • never blocks
  • spins in a loop
  • never returns to low-power wait
    :backhand_index_pointing_right: That alone can explain mA-level current

I’m feeling like maybe SX1262 not fully transitioned to sleep/standby

  • BUSY line might still be active
  • Or radio maybe left in RX/TX standby instead of sleep
    RadioLib usually handles this but I recall a post where it was in standby not sleep because of timing, it was an edge case.

also only does it on Battery Right, not when USB is plugged in?
What happens if you Force-clear the interrupt explicitly
After TX complete:

radio.clearIrqStatus();

Even if RadioLib should do it — do it anyway.

or just slam it into a known state for sure before hand.:grimacing:

radio.sleep();
delay(2);  // give it time

Also maybe, Check BUSY pin before sleeping MCU as a gate.

while(digitalRead(BUSY_PIN) == HIGH);

if BUSY never drops → bingo, that’s your smoking gun.

Good Stuff.
GL :slight_smile: PJ :v:

22

When sleeping, DIO1 goes low and BUSY goes high, which matches what I read on the pins. I have tried everything.

I have tried pulling the +3.3V supply to the SX1262 when I measure high current consumption, and this has no effect. Performing reset() has no effect, nor does finishTransmit().

I suspect that I have several faulty boards. The problem is that there aren’t any alternatives to the kit at such a good price. Heltec and RAKWireless are several times the cost.

23

i dont believe the boards are faulty… i linked this thread to seeed.. but they are hard to deal with

24

Hi there,

Here, Here… You have done everything IMO. Good on you for the effort. :wink: :oncoming_fist:
I agree there are no Cheaper alternatives, the RAK is good but too expensive IMO.

I’m with you though , In thinking you may have some marginal boards.
You got a picture of them? or the setup?

Pulling the power, more less seals it for me… that should Drop everything, sleep or NOT. so puzzling to say the least.! :grimacing:

Maybe we can get them to setup the same test , with your code and see if it shakes out :thinking:

HTH
GL :slight_smile: PJ :v:

2 Likes
25

I have excalated this to the Seeeders Only area… hopefully someone will come to work this week.. before Friday in China… what da?

If this is a hardware problem… and I am assuming it is based on all the work that is going on… it must be something on the chip, so seeed will probably not be able to “FIX IT” other than moving on to the newest chip. I know it does not solve the question… but moving on to live with the current available or a software solution. I am not an expert… but i cant think of a hardware fault that would cause this absent a bad transistor that will not switch fully…

one last thought is internal pullups?

If you can manually in the code activate and deactivate the individual pins/gpio

2 Likes
26

It has been very time consuming and frustrating. Jan Gromes over at Radiolib has been quite responsive but hasn’t been able to suggest anything.

I have ordered a cheap ProMicro NRF52840 board and will test that tomorrow.

2 Likes