Canon EOS R5 Overheating Contradictions

[skul]

I've been reading the articles posted by Andrew about the R5 overheating just because I'm interested as anyone about what's going. After reading the last few articles, I've trying wrap my head around some conflicting information.

In articles which describes the teardown that was done on the R5 it says that Canon didn't put in thermal pads over the processor and in fact placed a PCB over it, indicating that they were intentionally trapping heat over the processor.

In articles where EXIF data is used to track the temperature of the the sensor, the camera shows no signs of overheating.

The contradiction that I'm trying to wrap my head around is given the data given by EXIF, are the thermal pads even needed since based on the data the Camera doesn't seem to overheat.

The only way I see of resolving this contradiction is either the EXIF temperature isn't measuring the processor temperature or Canon had found some way of keeping their processor cool using poor external thermal dissipation.

I'd appreciate hearing anyone's thoughts or explanations regarding this. Am I misinterpreting something?

 

[Avenger 2.0]

I think that the EXIF data is indeed the temperature of the sensor and not the CPU.
I guess they didn't add a heatsink or thermal pads to the CPU because they know they can limit heat by software design (overheating protection).
The real question is, why does it take 2 hours to cool down (and doesn't ambient temperature play a role) and why can you do 4 hours while recording external without the overheating protection triggering.
To me it looks like they deliberately designed the camera hardware (no heatsink) and software wise (overheating timers) to limit the 4k and 8k recordings.

[Andrew Reid]

It is not documented anywhere how the EXIF temp is measured. The reason it is still useful is that it can give an indication of the camera operating temperature, inside the case, and can be correlated with the 'supposably' temperature related inhibitors reported by @BTM_Pix's app over Wifi.

In a 25C room the internal temp reported 46C during stills and 64C during prolonged 8K video.

So it does give an indication of the real-time workload of the processor.

In the fridge at 4C it reported a steady 34C. It also gives a reliable indication of whether the camera is heating up or cooling down.

I don't see why a camera that isn't heating up and is far away from 64C internally can refuse to record even a single frame of 8K at 34C internally.

This would indicate that the main CPU is thermally throttled because it is working so hard in live-view doing one JPEG per minute, it has reached a critical 85C or over.

I can't see how would be the case.

If live-view pushes the image processor that hard then there is something seriously wrong with the LSI.

Also it is important to bear in mind all the other weird discoveries...

Like the fact the image processor is happy encoding 4K H.265 10bit 4:2:2 from pixel binned sensor data with no heat limits, but not from the oversampled sensor data, even though to the image processor the resolution, bit depth and colour sampling is identical in both modes... But that the sensor is happy doing the oversampled 8K sensor readout, and image processor is happy to receive this and produce 4K HQ from it, sending it to the HDMI port... for FOUR hours to an external recorder! So clearly the sensor isn't heat limited doing 8K and the image processor isn't heat limited doing 4K H.265 10bit 444, so where's the overheating issue?

[BTM_Pix]

7 minutes ago, Andrew Reid said:

It is not documented anywhere how the EXIF temp is measured. The reason it is still useful is that it can give an indication of the camera operating temperature, inside the case, and can be correlated with the 'supposably' temperature related inhibitors reported by @BTM_Pix's app over Wifi.

In a 25C room the internal temp reported 46C during stills and 64C during prolonged 8K video.

So it does give an indication of the real-time workload of the processor.

In the fridge at 4C it reported a steady 34C. It also gives a reliable indication of whether the camera is heating up or cooling down.

Yep, the EXIF temperature data is just the relative reference gauge to what is happening as its the calculation that triggers and then resets the inhibitors that is the key issue.

Clearly, it is not being sourced from an external area as it would match the ambient temperature of the room so it is relevant as a partial indicator of the internal temperature.

I'm old enough to remember when the crutch that was being used for this camera was that it was OK as long as you didn't shoot long video clips and even if you did it was only supposed to overheat in very hot ambient temperatures.

If nothing else, 60 jpegs in a fridge has totally removed that crutch and the defence of it is all getting a bit Black Knight from the Holy Grail now.