another rare bug - page 4 - General Help and Assistance on using CHDK stable releases - CHDK Forum

another rare bug

  • 48 Replies
  • 18411 Views
*

Offline srsa_4c

  • ******
  • 4451
Re: another rare bug
« Reply #30 on: 17 / March / 2016, 13:41:48 »
Advertisements
When I live my car at the parking lot during a hot day it shutdown automatically.
Does anybody know is it programmed to do so?
Never experienced that myself, but AFAIK Canon EOS cameras do shut down when overheated. It's possible that Powershots have similar protection.

Quote
Is there any way to bypass it and make it work?
I have no experience in this area, so don't know.
Quote
As I understand camera has 3 different temperatures.
I can make a study about it by printing temperatures out just before shutdown.
That would be useful, so please do (you could also log the measured values like once per minute).

*

Offline reyalp

  • ******
  • 14080
Re: another rare bug
« Reply #31 on: 17 / March / 2016, 13:44:14 »
Does anybody know is it programmed to do so?
It wouldn't be surprising if the camera had a high temp shutdown. I have never noticed code that looked like it was for this, but that doesn't mean it isn't there. I though my A540 did this once, but I never investigated further.

Quote
I can make a study about it by printing temperatures out just before shutdown.
You can log the temperatures periodically. If there is a fixed shutdown value it should be pretty obvious. To keep the log files from getting too big, you could only log it when it changes more than a certain amount.
See http://chdk.wikia.com/wiki/Script_commands#get_temperature
I would expect the sensor temperature to be the one that triggers shutdown, but I guess battery could too on cameras that use lithium batteries.
Don't forget what the H stands for.

*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #32 on: 22 / July / 2016, 11:50:17 »
I made preliminary investigation of temperature issues.

Conclusion: the camera shuts down when get_temperature(1) reaches 80. I investigated A1400 camera.

More details:
I had a script that call the function

Code: [Select]
fout = io.open("A/out.txt", "a")
fout:write(tostring(get_temperature(0)),"    ",tostring(get_temperature(1)),"    ",tostring(get_temperature(2)),"   ",os.time(),"\n") 
fout:close()

approximately every five minutes. I attached the out.txt file.
I run my camera this summer in my car and periodically I had many shutdowns during few months. Particularly I had a lot of them on a parking lot when airconditioner is off and car became red hot inside of it.

If you look at the file you will see that os.time() normally has a time step of 300 seconds (about 5 minutes) between measurements.
Every time when get_temperature(1) gets 79 it shuts down and it takes a long time before I start it again. The time step is much bigger.
There are few cases when temperature reaches 79 but after that goes down to 78 e.t.c. That means that camera still works when temperature is 79 and it shuts down immediately when temperature gets 80.

Somehow  for my other commercial project, elph 115 camera worked continuously during whole summer without any shutdown issues but I need to do extra measurements to understand why it works and I will publish it. I just know that A1400 has two electrodes (two AAA batteries) and elph115 has power supply with three electrodes and different battery. There must be something different about it.




*

Offline reyalp

  • ******
  • 14080
Re: another rare bug
« Reply #33 on: 25 / July / 2016, 22:23:15 »
Conclusion: the camera shuts down when get_temperature(1) reaches 80. I investigated A1400 camera.
Nice work.

get_temperature(1) is sensor temperature. 80 c is quite hot, so I'm not surprised it triggers a shutdown.
Don't forget what the H stands for.


*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #34 on: 26 / July / 2016, 10:06:45 »
I would appreciate if somebody would give me idea (not precise instruction) how to bypass this feature or how to to overload or block temperature measurement.
In my project I also try to use external fan but still don't have an answer how it cools down the sensor temperature. I also don't know if there is any variation from camera to camera.

Re: another rare bug
« Reply #35 on: 26 / July / 2016, 14:16:19 »
There will almost certainly be variations been camera models. The ability of the sensor to dissipate heat will be affected by many things in the camera's design.  Metal body vs plastic body for example.  If you try to bypass the temperature checks, you risk turning your camera into a brick.  Also, i doubt the type of battery matters at all relative to the sensor temperate.
Ported :   A1200    SD940   G10    Powershot N    G16

*

Offline reyalp

  • ******
  • 14080
Re: another rare bug
« Reply #36 on: 26 / July / 2016, 20:59:49 »
I would appreciate if somebody would give me idea (not precise instruction) how to bypass this feature or how to to overload or block temperature measurement.
Figuring out how to bypass it would require reverse engineering. I'd start by searching for references to GetCCDTemperature, but I would expect the camera to fail (hang / crash) fairly close to the shutdown threshold. If that's the case, bypassing the check would have limited value.

Quote
In my project I also try to use external fan but still don't have an answer how it cools down the sensor temperature. I also don't know if there is any variation from camera to camera.
If it works with your project, I'd suggest switching the sensor off to allow it to cool down between shots. Playback mode will do this for sure, and on some switching the display off may also. For cameras with an optical viewfinder, the disp button display off should work. Some others have a power saving screen off mode that can either be assigned to a key, or triggered after a short time by power saving settings.

From script, you could trigger this based on the sensor temperature before it hits the shutdown limit.
Don't forget what the H stands for.

*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #37 on: 28 / July / 2016, 10:32:18 »
Thanks, it's a good idea to cool it down periodically, for example when temperature reaches 75C  bring it to playback mode for 5 minutes and start again. Unfortunately in a hot days camera will heat up quickly again and I will loose more than 50% of availability and it would be a bad thing for my customers.
I know that LCD screen produces enough heat when it's on so I decided to keep the screen brightness as low as possible. Is there any CHDK parameter for LCD brightness that would setup it even smaller than minimum firmware value?


*

Offline reyalp

  • ******
  • 14080
Re: another rare bug
« Reply #38 on: 28 / July / 2016, 13:13:05 »
Unfortunately in a hot days camera will heat up quickly again and I will loose more than 50% of availability and it would be a bad thing for my customers.
IMO it's quite unlikely that bypassing the thermal shutdown limit would have a better outcome. The limit is most likely there to prevent either crashes or hardware damage.

It sounds like you need active cooling for your application.
Quote
Is there any CHDK parameter for LCD brightness that would setup it even smaller than minimum firmware value?
Yes, use set_backlight(0) to turn of the backlight, or set_lcd_display(0) to turn off the display entirely.
Don't forget what the H stands for.

*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #39 on: 28 / July / 2016, 13:30:28 »
Yes, use set_backlight(0) to turn of the backlight, or set_lcd_display(0) to turn off the display entirely.

cool :) I didn't know about this feature. What is difference between this two functions electronically and physically in terms of power consumption/dissipation?
Will it affect CHDK motion detection function in any way?

 

Related Topics