Bug? ESP8266 Timer stops after 75 minutes

[Bert Oudshoorn]

The led stops blinking after about 75 minutes, whether 500 or 1000 msec is specified.
Serial and Log statements have been removed temporarily for testing and the physical
tx/rx wires not attached. The usb only attached for power.
Might it be a memory leak? Or a counter limit in C? Or processor setting?
The processor was not dead / in sleep. A button and 2nd led in an other try-out was
still working.

'// ESP8266-01 blinking try-out via a Timer
#Region Project Attributes
#AutoFlushLogs: True
#CheckArrayBounds: True
#StackBufferSize: 300
#End Region

Sub Process_Globals
Private led2 As Pin
Private led2_state As Boolean = False
Private timer1 As Timer
End Sub

Private Sub AppStart
led2.Initialize(2, led2.MODE_OUTPUT) '// on GPIO2
timer1.Initialize("timer1_Tick", 500) '// or 1000
timer1.Enabled = True
End Sub

 

[Erel]

Please use [code]code here...[/code] tags when posting code.

Can you post the full code?

I ran the clock program for several days and it ran properly.

 

[Bert Oudshoorn]

Erel, thank you. You give me hope What I listed is the full code (without indenting, unfortunately). Weird, what can be the difference? Two libraries attached: rCore 1.10 and rESP8266 1.00 I run on a ESP8266-01 with 512K64, Generic ESP8266 module. The UTC+2 cannot be the problem.

 

[Erel]

Where is Timer1_Tick?

 

[Bert Oudshoorn]

Oeps, that was missing, I have some problems with embedding/attaching things (Wondows 10 Edge?)
SubTimer1_Tick
led2_state = Not(led2_state)
led2.DigitalWrite(led2_state)
EndSub

 

[Erel]

Try it with the logs and see whether the timer continues to run.

You can check the available memory with:

Log(AvailableRAM)
Log(StackBufferUsage)

Both should be constant.

 

[Bert Oudshoorn]

Thank you. The AvailableRAM is more or less constant (swaps between 2 close values), the StackBufferUsage stayes on 0. However, the Timer does not stop, if a Log is specified. It does not matter what is logged. Even each minute a Log(counter) "solves" the problem. Would it be a sleep mode?

 

[Erel]

Would it be a sleep mode?

No. It never sleeps.

My guess is that the problem was somewhere else.

 

[Cableguy]

could it be a brown-out situation? if the board was running in battery mode, a low voltage could have that effect

 

[Bert Oudshoorn]

Thank you, no I tested it even with a 2,4 A Apple power supply and use a LM1117 for the 3,3 V.
But I tried it now with WeMos D1 mini, the first example from
https://www.b4x.com/android/forum/threads/timers-loopers-and-callsubplus.65989/
Did not change anything. Pin13 = D7, a led with a 220 ohm resistor.
It also stops blinking (without a Log statement in the Timer event).

And this can not be the problem, I guess
#Region Project Attributes
#AutoFlushLogs: True
#CheckArrayBounds: True
#StackBufferSize: 300
#End Region

 

[Erel]

I was able to reproduce this issue. It is not yet clear what causes it. I haven't measured the time so I don't know when exactly it has stopped working.

With a log every minute it seems to work properly. It now runs for two hours here.

For now it is not clear why it happens.

 

[woniol]

Hi,
I have the same problem with my NODE MCU module.
Any idea how to solve this?

My code:

Private Sub AppStart
   Serial1.Initialize(115200)
   Log("AppStart")
   If wifi.Connect2("xxx","yyy") Then 'change to your network SSID (use Connect2 if a password is required).
     Log("Connected to wireless network.")
   Else
     Log("Failed to connect.")
     Return
   End If
   ScanNetworks
   pp.Initialize(p2.D0, pp.MODE_OUTPUT)
   pp.DigitalWrite(True)
   timer1.Initialize("timer1_Tick", 50)
   timer1.Enabled = True
End Sub

Sub Timer1_Tick
    pp.DigitalWrite(Not(pp.DigitalRead))
End Sub

 

[woniol]

I can also confirm that adding Log to the Timer_Tick Sub solves the problem.