Skip to content

T1000-E and W12 sleep checks

Checked on Mercerwood Pi on 12 November 2026 using the installed applications:

Board Radio Firmware
Seeed Tracker T1000-E LR1110 v1.18.0-dev-db54a6ad, built 7 November
Meshnology W12 LR2021 v1.18.0.2-halo-keymind-cascade-dev-feb7c43d, built 8 November

These chips do not use the SX1262-specific 0x08B5 workaround. No SX1262 register writes were added to their drivers. Their native gain-restoration and sleep commands remain separate.

Observations

Both radios stayed on their original 910.525 MHz, BW62.5, SF7, CR5 channel, with boosted RX gain and CPU power saving on. No deliberate RF transmission, firmware flash or recovery-service interruption was performed.

Check T1000-E W12
New AGC, 16-second setting, 70-second observation 3 successful refreshes; 17 packets received 1 successful refresh in its active multi-SF mode; 17 packets received
AGC failure counter 0 before and after 0 before and after
Ordinary RXPS, 30 seconds after effective=armed Level 8 / preamble 16; 7 packets received Level 6 / preamble 16; 7 packets received
RXPS arming failure counter Pre-existing value 1, unchanged 0, unchanged
Soft/hard watchdog recoveries 0/0, unchanged 0/0, unchanged

W12 multi-SF reception was temporarily disabled for the ordinary RXPS check; its normal multi-SF mode holds RXPS in continuous receive. It was restored to on 5-8 radio1; single active. T1000-E's AGC setting was restored to 0 and W12's to 16. Original radio settings, RXPS levels/timings and gain settings were verified afterward.

These are sleep/reception observations using ambient traffic, not a controlled packet-loss comparison, RF blocker test or direct measurement of analog gain. The gain setting readbacks report the configured setting. They do not prove that either chip is immune to every gain-related hardware fault.

Private evidence: /home/mikec/hwtest/lr-sleep-check-20261112 on the Pi.