Canon EOS R5 teardown suggests overheating time limits hard-coded into firmware
Aug 11, 2020
Share:
After seeing numerous reports of the lengths to which people had tried and failed to cool down the Canon EOS R5 to get longer video shooting times out of it, I was starting to suspect that the “overheating” time limits might actually be coded into the camera’s firmware and it actually has nothing to do with the temperature of the camera or the environment.
Now it seems those suspicions might have some evidence after a Chinese EOS R5 owner tears his camera down to try and fix the problem himself. He posted images of the process to Baidu and we can see exactly how the EOS R5 is laid out inside, which seems designed not to aid cooling, but stifle it. A video of the teardown was also posted.
The video and images show that the hardware is laid out in such a way that it actually seems to trap heat around the CPU rather than take it away. Strangely, while there are thermal pads covering the RAM chips, much of the CPU remains uncovered. In fact, the small sections of the CPU that are covered actually share the thermal pads from the RAM.
On removing the thermal pads from the RAM, we can see exactly how big the CPU inside the EOS R5 is, and how little of it is covered. The RAM is likely to heat up quite substantially and transfer heat to the edges of the CPU, which would otherwise generally be lower in temperature than the unexposed centre.
There’s also a whole second board that sits on top of this one, further trapping air and preventing the CPU from venting heat out through the magnesium alloy chassis. A chassis that, as EOSHD points out, Canon claims mitigates overheating issues. On reassembling, the user added thermal paste over the CPU to help allow increased heat transfer away from the chip, and after measuring the temperatures with a thermal image camera, the CPU does seem to be the main source of heat within the camera, and not the sensor (which didn’t go over 40°C).
According to EOSHD, though, the camera still shut down after 20 minutes, even with more heat taken away from the CPU. Combined with the fact that people seemed to get the same overheating time limits regardless of whether they were in arctic conditions or the Arizona desert, this suggests that the overheating limits are actually hard-coded into the EOS R5’s firmware and that it’s not actually monitoring the temperatures inside the camera at all.
This wouldn’t be without precedent. Flash manufacturers have often done the same thing to prevent their strobes from overheating, with clearly defined tables showing how many flashes you can do at different power levels in a given time frame before the overheating prevention kicks in automatically – regardless of the conditions in which you’re shooting. And they usually hit those numbers pretty spot on (and pulling the battery out and reinserting it usually resets it – suggesting that they’re not actually monitoring the temperatures, either).
The good news, though, is that if it is just a firmware limit, then Canon may choose to remove it or alter the code in such a way that it allows for longer recording times. Whether or not they’ll do it, however, is another matter entirely. If they don’t, then maybe we’ll see Magic Lantern make something of a comeback.
[via EOSHD/Canon Watch]
John Aldred
John Aldred is a photographer with over 25 years of experience in the portrait and commercial worlds. He is based in Scotland and has been an early adopter – and occasional beta tester – of almost every digital imaging technology in that time. As well as his creative visual work, John uses 3D printing, electronics and programming to create his own photography and filmmaking tools and consults for a number of brands across the industry.
Join the Discussion
DIYP Comment Policy
Be nice, be on-topic, no personal information or flames.
9 responses to “Canon EOS R5 teardown suggests overheating time limits hard-coded into firmware”
If there is no heat monitoring, it will be impossible to remove this timed limitation.
You need a sensor to monitor the temperature, read the sensor data, and advice the user and shut down if it gets out of control.
But I never saw any processor without temperature reading, so I guess they can implement something to fix it.
Ideally this should have a cooler but the question is where it would exhaust. The termal pad covering the cpu and less of the RAM could help mitigate the issue. While video processing is a heavy operation, probably a lot of hardware decisions was pushed by keeping a smaller factor design. All speculation since only Canon knows why this is happening really.
I believe the overheating issue is not due to CPU or RAM, rather the CFExpress card reader and the microchips responsible for the writing to the card.
As we can se in the following article from CanonRumors, one guy managed to record for 4 hours straight externally and no writing to any card.
If so, the overheating issue is a huge flaw and Canon is taking every step to not burn those expensive CFExpress cards and not just destroy the ability to write to the card but also make it nearly impossible to recover any data from it.
https://www.canonrumors.com/canon-eos-r5-records-4-hours-of-4khq-30p-to-an-external-recorder-with-a-couple-of-simple-tweaks/
Solución: don’t buy canon
The reason external recording removes the problem is that most of the processing of the video data (compression etc.) is removed from the camera’s main processor and to the external recorder, and so inside the camera the processor heat is reduced and internal card can be removed so it won’t generate heat, either. Finally, using an external recorder allows the camera to avoid the tariff related to video camera that is able to record >= 30 min of footage, since it’s not technically recording anything.
I see many blame the CFE-B cards. Canon has used it in 1DXIII and they know the characteristics of CFE-B card. So, there is no reason to put the blame on the card and how it works.
the same card is used in different models and brands… so it’s Canon inadequacies or ignorance to be blamed.
Jesse Evans on fredmiranda reports that 25 min in the freezer restored the full recording time of the R5.
“On reassembling, the user added thermal paste over the CPU to help allow increased heat transfer away from the chip, and after measuring the temperatures with a thermal image camera, the CPU does seem to be the main source of heat within the camera, and not the sensor (which didn’t go over 40°C).”
What if the limit is because of possibly overheating the memory cards and not so much the camera components themselves?
How hot does the memory card get when shooting in 8K?