I've posted the initial findings of the overnight-freezer cooled body here. More analysis of the data is forthcoming.
Here's the graph from that post, showing the EXIF-reported temp over time between yesterday's experiment (15 minutes before thermal cutoff) and today's experiment (69 minutes before cutoff). Note from the post that there are other differences as well, including which CFE card was used, along with interesting details on how the different CFE card yielded a different precision of available-time indicators.
how did he deep cool the camera? what was the resultant interior ambient temperature of the camera after deep cooling? what was the camera temp of the 1st experiment?
ooopppsss..... posted this before you re-posted new data..
InnomnateViem wrote:
how did he deep cool the camera? what was the resultant interior ambient temperature of the camera after deep cooling? what was the camera temp of the 1st experiment?
Click the link to see the thread over at DPreview.
what i see is that cryogenic cooling the R5 camera frame will work to extend the 8K record time, as stated in some of my previous post. albeit, one has to modify the camera to do so efficiently.
disclaimer: i have no intent of modifying an R5.
however, the EOS-R is a different story. i am going to modify the EOS-R to accept a Peltier cooling module to hold the sensor (interior ambient operating temperature) @ 0°C. this will be a dedicated full colour astrophotography camera with improved sensor noise and dark current response as a result.
i've been following your post on DPR.
based upon your preliminary results it looks very promising.....
Very interesting stuff, Snapsy. I'll be following your thread on DPR now. The card swap was interesting in that it extended the time. Premature for us to decide how that works out due to the refrigerated body.
I'm going to guess that there's both a hardware measured temperature facet and a time-vs-throughput timer facet to the temperature warning/shutdown. Like it's estimating core temperatures or something based on time, data rate, and actual measured temperature at some particular point in the camera. Maybe...
I just read through probably too much of this. But I've concluded that Canon can easily address the problem with a simple firmware update that reads the temperature directly off the main board and provides a more accurate cool down recovery time and thermal limitation. I legitimately can't believe the lengths people are going through to figure out what's going on with these camera...and destroying them in the process.
So I’m helping snapsy with a test on cool down times... started to cool down around 1pm to see when it would give me my full time back... it’s now 9:34 pm.... still not back to full 15 minutes of 8k time limit. It’s been turned off, only turned on to check the status every now and again
i don't think that the individual conducting the experiment had any say in letting the camera ramp up to 69°C/156°F. that would be a function of the camera's combined SW/FW/HW.
what is interesting is that the camera ramped up to 69°C the operating environment specification is 0°C - 40°C.
Ziffl3 wrote:
sounds high like a high temp. but this is actually like 70 degree C.
i would suspect the sensor can at least handle 150 degree C.
just saying.
Just as a point of reference, modern cpus(and gpus) have a temp max of around 100c, after which they throttle down or just shutdown to prevent damage. 70C is warm for a processor that isn't under a sustained and strenuous load. I sincerely doubt that a camera's sensor would be rated for higher than that(and probably less than that tbh), 150c is very hot, in a small device with a battery.
Also Andrew reid has a serious chip on his shoulder when it comes to canon anything
Yeah, I have a hard time thinking that about any kind of consumer electronics would work after sustained stints in my oven heated to the temperature of boiling water. Even my car keeps its coolant temps below that temperature and that is primarily mechanical.
Even at 70C, I have a hard time thinking that electronics constantly heated to temps that cook meat to over well done shoe leather consistency and gets rid of about any harmful pathogen in our food is good either.
I am not sure there would be much distinction between consumer and military when it comes to cpu operational temp limits.
I fairly recently used to work at a little place, TI. I worked in the DLP group.
These are cmos based chips.
Also low temp silicone like camera sensors.
Can’t say much more other than while 70 c are the public specs.... they can and do operate at higher temps.
This doesn’t make sense, people have had these R series cameras going for over 45 minutes by removing the SD card and using an external recorder, so why wouldn’t the firmware based issue kick in during that time?
My guess this is a some sort of old software rule from the old rules of no recordings past 29min.
Canon probably put this there to honor old EU regulations. If you have cards in, there is a limit, including whatever goes out on HDMI. If you take the cards out, no limit. The laws were likely worded in a way that in order to enforce them, this was the workaround.
InnomnateViem wrote:
what is interesting is that the camera ramped up to 69°C the operating environment specification is 0°C - 40°C.
Operating environment = external environment, i.e., the local temperature where the camera is used.
I believe that the 69 C is being measured or provided as an internal temperature. I think the original DPR post describes the methodology used, whether they're using an infrared thermometer to measure the sensor temperature or some internal readout from the camera.
TeamSpeed wrote:
My guess this is a some sort of old software rule from the old rules of no recordings past 29min.
Canon probably put this there to honor old EU regulations. If you have cards in, there is a limit, including whatever goes out on HDMI. If you take the cards out, no limit. The laws were likely worded in a way that in order to enforce them, this was the workaround.
That all needs to recoded/revamped.
Yeah, I think it's 29:59 as a limit based on EU laws/regulations. But I'm not sure that if you stop and immediately restart, that doesn't restart the timer on that regulatory restriction.
And I believe I've read that using an external recorder DOES work around that regulation.
Working around the heat issue is something else, and we're seeing external recording as a workaround for that too - maybe.
Snapsy, et.al., is doing great discovery work on that issue, and its much appreciated.
But with the R5, you have to take the cards out to remove that time restriction to the external device. Seems like an old software regulatory rule that should be removed. I suspect that to meet regulations years back, they had to show that with cards in the camera, the video output was restricted per the rules, with the cards out, no regulations had to be met.
