PSM Current Consumption Issue on HL7812 After Wake-Up

[jyijyi]

a bit confused, too many different factors affecting the test result…
can resending AT+CPSMS during wake up works as a workaround?

[zsen]

@jyijyi

can resending AT+CPSMS during wake up works as a workaround?

seem the issue is here tau timer, isnt it?

it didnt work tau: 15 mins, and it worked with tau: 24 hours.

I hope I am not confused about what TAU timer means. as far as I know it means: Maximum duration the module can remain attached to the network without transmitting (i.e. in a data
session or TAU procedure).

[jyijyi]

didn’t you say 24 hours TAU still has problem on current consumption?

PSM Current Consumption Issue on HL7812 After Wake-Up

Yes, exactly. That time I didn’t observe this behavior. But that time test case was a little bit different than what I have now. In that case, I had tcp comms but now I dont have it. @jyijyi OK I tested again with the same cpsms config: AT+CPSMS=1,“10101111”,“00001111” that I used at that time. exactly same issue current consumption still stays around 25 uA even though I recevied psmev: 1 URCs [image] [2025-06-19 17:07:40.000] +CEREG: 4,,,,,,"00001111","00111000" [2025-06-19 17:11:08.19…

[zsen]

@jyijyi

didn’t you say 24 hours TAU still has problem on current consumption?

The 24-hour TAU setting still causes issues with consecutive PSM entries, although the first PSM entry works fine. After setting TAU to 24 hours, the modem successfully entered hibernate mode with a current draw of ~4 µA. However, after waking it up and waiting for the next PSM cycle, it never returned to hibernate.

But in this case:
For reference, in this post:

PSM Current Consumption Issue on HL7812 After Wake-Up

This is exactly what I tested here, these logs belong to that scenario. And also see the attached full log for more detail, please. correct, it is still consumes 20uA after psmev: 1 and VPIO is low AT+KSLEEP=1,2,0 I have also attached a full log that show all the modem comms from booting up to 2 consecutive PSMs. serial-terminal-16062025_132404.txt (192.3 KB)

I tested resetting the AT+CPSMS with a 15-minute TAU, but this is the case resetting cpsms didnt work.

now is it clear?
basically,

• AT+CPSMS=1,,,"10101111","00001111" (15-minute TAU) does not work, even for the first PSM entry after sending the command.
• AT+CPSMS=1,,,00111000,00001111 (24-hour TAU) works for the first PSM entry, but fails for subsequent entries.

[jyijyi]

@zsen

I tested resetting the AT+CPSMS with a 15-minute TAU, but this is the case resetting cpsms didnt work.

does this mean that resetting cpsms to 24 hours work?
i mean what if every time wake up and resetting +CPSMS to 24 hours?

[zsen]

• AT+CPSMS=1,,,"10101111","00001111" (15-minute TAU) does not work, even for the first PSM entry after sending the command.

I still have the same AT+CPSMS=1,,,"10101111","00001111" (15-minute TAU) configuration set on the modem, and with that configuration, the device eventually entered hibernate mode—but it took around 17 minutes to do so.

Observe the red line—there’s a steady 25 µA current draw from the moment the wakeup pin is released until the modem enters PSM hibernate. You’ll notice a brief current spike before it settles at 4 µA, ( purple line )indicating successful transition into hibernate mode.

[zsen]

@jyijyi

i mean what if every time wake up and resetting +CPSMS to 24 hours?

Ok, this is insane. I just reset the cpsms again with a 24-hour tau timer, and it didnt work this time. still consumes 25uA in the psm.

Totally random behavior.

[jyijyi]

//Periodic-TAU:15 minutes
//active timer: 30 seconds
[2025-06-19 17:11:10.016] AT+CPSMS=1,,,"10101111","00001111"

[2025-06-19 17:11:12.931] OK

[2025-06-19 17:11:12.931]

[2025-06-19 17:11:12.931] +PSMEV: 0

[2025-06-19 17:11:13.395]

[2025-06-19 17:11:13.395] +CEREG: 5,"2332","02E9F81F",9,,,"00001111","00000111"

[2025-06-19 17:11:43.523]


//active timer: 30 seconds
// Periodic-TAU:70 minutes

[2025-06-19 17:11:43.523] +CEREG: 4,,,,,,"00001111","00000111"

Just checked your log, actually you cannot set Periodic-TAU to15 minutes as eventually your network will set to 70 minutes.

[zsen]

@jyijyi

Just checked your log, actually you cannot set Periodic-TAU to15 minutes as eventually your network will set to 70 minutes.

Yes, that’s another important point. Regardless of the values we configure with AT+CPSMS, the network parameters ultimately override them. In any case, the transition into PSM hibernate appears completely unpredictable. I’m honestly running out of ideas at this point and not sure what else to try.

[jyijyi]

this is the network setting
Some network does not allow periodic-TAU to be a small value…

[zsen]

@jyijyi

Some network does not allow periodic-TAU to be a small value…

thats fine. In any case, I set it again to 24 hours, but it didnt work this time.

[jyijyi]

does this mean that sometimes it works by resending AT+CPSMS with perioidic-TAU to 24 hours?

[zsen]

Yes, correct. But I dont know what the case is that psm works after setting it. See I tried resetting it every time. Behavior is quite different here.

[2025-06-19 18:28:57.434] +CEREG: 5,"2332","02E9F81F",9,,,"00001111","00111000"

[2025-06-19 18:28:57.434]

[2025-06-19 18:28:57.434] +KCELLMEAS: -92.0,121.0,,,6.0,1,1

[2025-06-19 18:29:07.458] at+cpsms?

[2025-06-19 18:29:13.510] +CPSMS: 1,,,"00111000","00001111"

[2025-06-19 18:29:13.510]

[2025-06-19 18:29:13.510] OK

[2025-06-19 18:29:26.142] at+cpsms=1,,,00011000,00001111

[2025-06-19 18:29:28.816] OK

[2025-06-19 18:29:28.816]

[2025-06-19 18:29:28.816] +PSMEV: 0

[2025-06-19 18:29:30.497]

[2025-06-19 18:29:30.498] +CEREG: 5,"2332","02E9F81F",9,,,"00001111","00011000"

[2025-06-19 18:30:00.653]

[2025-06-19 18:30:00.653] +CEREG: 4,,,,,,"00001111","00011000"

[2025-06-19 18:30:00.653]

[2025-06-19 18:30:00.653] +PSMEV: 1

[2025-06-19 18:30:24.455]

[2025-06-19 18:30:24.455] +CEREG: 5,"2332","02E9F81F",9,,,"00001111","00011000"

[2025-06-19 18:30:24.477]

[2025-06-19 18:30:24.477] +KCELLMEAS: -92.0,119.0,,,9.0,1,1

[2025-06-19 18:30:31.819] at+cpsms=1,,,00011000,00001111

[2025-06-19 18:30:32.717] OK

[2025-06-19 18:30:57.329]

[2025-06-19 18:30:57.329] +CEREG: 4,,,,,,"00001111","00011000"

[2025-06-19 18:31:31.032]

[2025-06-19 18:31:31.032] +CEREG: 5,"2332","02E9F81F",9,,,"00001111","00011000"

[2025-06-19 18:31:31.032]

[2025-06-19 18:31:31.032] +KCELLMEAS: -92.0,119.0,,,7.0,1,1

[2025-06-19 18:31:33.445] at+cpsms=1,,,00011000,00001111

[2025-06-19 18:31:34.984] OK

[2025-06-19 18:31:40.376] at+cpsms=1,,,00011000,00001111

[2025-06-19 18:31:41.002] OK

[2025-06-19 18:31:44.251]

[2025-06-19 18:31:44.251] +CEREG: 4,,,,,,"00001111","00011000"

here [2025-06-19 18:30:00.653] +PSMEV: 1 the device went into psm hibernate with a current draw of 4 uA but not after that, even after resetting it here and further from it [2025-06-19 18:30:31.819] at+cpsms=1,00011000,00001111, I never received psmev:0 and 1

[jyijyi]

i remember you have talked about this before, not sure if this is still related:

Unexpected +KCELLMEAS URC prevents PSM from functioning properly

And the +kcellmeas URC comes 10 seconds after +PSMEV: 1, which seems matches to what you set in the timeout parameter of +kcellmeas. If you set the timeout to longer period of time, will the modules sleep longer? I believe this is not using URC you can do this: disable +KCELLNMEAS URC go to PSM dormant mode module should be able to sleep when you need to send data, assert wake up pin module should wake up now enable +KCELLNMEAS URC go to PSM dormant mode according to you test, module sho…

[zsen]

I think this could be the reason why power consumption stays around 20 uA

here is the explanation in the 2174229 HL78xx Low Power Modes Application Note r7.pdf