How could I debug a strange issue with my Mi AX3000t (rd03) model.
I recently tested nearly all different builds (snapshots, our "local" eko builds, xwrt builds, immortal etc) - and there is a problem with transfer rates after resuming from sleep on my laptop device (AX210 latest drivers).
When I first connect to AX3000T - AP configured device - and test ex. Openspeedtest to my local server - I get near 1000/1000 Mbps rates. After some time (~2h or more), I resume work with my computer, but the AX3000t log is full of this messages:
Thu May 30 17:43:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:43:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:44:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:44:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:45:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:45:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:46:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:46:44 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:47:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:47:40 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:48:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:48:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:49:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
Thu May 30 17:49:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic
Thu May 30 17:50:13 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
and transfers are 100/800 or 60/200.
If I connect with a new device (like smartphone) - there is no problem - the transfers are normally good.
The problem is maybe with SMPS AX Intel driver and sleep/resume negotiation - but on screen everything is ok and rates are high. But real transfer is horrible.
I tested all firmwares for wifi module (newest from 05. 2024) or original. WED/no WED etc.
When I revert to original Xiaomi firmware, based on mtk sdk - there is nothing similar like this - sleep/resume works ok. I know, that the firmware is much different, but it works, and the solution of changing SMPS in properties of Windows driver is not elegant - and for no all users.
Also I applied ht_capab "SMPS-STATIC' on AP side - no effect.
Thanks for tips how to debug more this behavior.
How could I debug a strange issue with my Mi AX3000t (rd03) model. I recently tested nearly all different builds (snapshots, our "local" eko builds, xwrt builds, immortal etc) - and there is a problem with transfer rates after resuming from sleep on my laptop device (AX210 latest drivers). When I first connect to AX3000T - AP configured device - and test ex. Openspeedtest to my local server - I get near 1000/1000 Mbps rates. After some time (~2h or more), I resume work with my computer, but the AX3000t log is full of this messages:
Thu May 30 17:43:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:43:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:44:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:44:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:45:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:45:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:46:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:46:44 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:47:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:47:40 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:48:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:48:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:49:38 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off Thu May 30 17:49:39 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 dynamic Thu May 30 17:50:13 2024 daemon.notice hostapd: phy1-ap0: STA-OPMODE-SMPS-MODE-CHANGED 64:49:7d:e5:3b:57 off
and transfers are 100/800 or 60/200. If I connect with a new device (like smartphone) - there is no problem - the transfers are normally good. The problem is maybe with SMPS AX Intel driver and sleep/resume negotiation - but on screen everything is ok and rates are high. But real transfer is horrible. I tested all firmwares for wifi module (newest from 05. 2024) or original. WED/no WED etc. When I revert to original Xiaomi firmware, based on mtk sdk - there is nothing similar like this - sleep/resume works ok. I know, that the firmware is much different, but it works, and the solution of changing SMPS in properties of Windows driver is not elegant - and for no all users. Also I applied ht_capab "SMPS-STATIC' on AP side - no effect. Thanks for tips how to debug more this behavior.