I bought a Canon EOS R5 - potential overheating solutions

[Video Hummus]

13 minutes ago, Andrew Reid said:

1. Why is a circuit board sitting between the main CPU and back casing, blocking the heat from spreading away into the chassis

Cost cutting on R&D.

13 minutes ago, Andrew Reid said:

2. Of course, why is there no thermal conductive material on the CPU?

Cost cutting on materials and assembly.

13 minutes ago, Andrew Reid said:

3. And why does the RAM thermal pad overlap onto the CPU, but not entirely cover it? (It seems to spread the heat from the RAM onto the CPU which is never a good idea).

 

Cost cutting on materials and assembly.

13 minutes ago, Andrew Reid said:

4. Why does ice not cool the camera and speed up recovery time? The firmware recovery countdown timer is so slow to go back up and always the same.

Because absolute temperature of circuit board has very little to do with recover times.

The ultimate test of this is to open the damn thing up in a walk-in ice cold freezer with leaf blower on it and then try and record. I bet you the record limits won’t change.

[noone]

Maybe just blame Sony since the camera seems to be actually made using at least one Sony part!

This is really a horror story of a great camera for some almost useless for others and it was sold as being for both groups.

[Cliff Totten]

6 hours ago, Andrew Reid said:

Thanks @androidlad

This is the kind of post we need more of.

Andrew....I have said this over and over and over again. The heat condition of the R5 is not an engineering "mistake" on Cannot's part. It is literally doing exactly as it was designed to do from day one. Cannot wants the R5 to be used only in specific ways and this does not coincide with the way "we" want to use it.

We have a need that we want the R5 to fill. However, Cannot wants to sell us a different camera for our need.  Yes, the firmware is deliberately highly restrictive. They are actively "steering" their customers the best way the know how with these crippling tricks.

As customers, we need to stop thinking the R5 was an engineering mistake and simply let accept it as a casual camera that has zero overlap with any of their true camcorders.

Cannot will never stab their Cine EOS team in the back....no matter how bad we want them to. If anybody thinks the Cine EOS team doesnt care about the R5 and was not in many of the R5 devopment meetings and planning discussions....you are crazy!

[visionrouge]

This is very interesting indeed.
It's look like no thermal management was planned from start.
Something like:
We are going tot stop recording after 20mn anyways by software, so there is no need to worry about.
We are also gonna write this very clearly in the user manual so we will not be sued for such.

Finding that the Wifi menu is enough to trigger it will only help the release of a firmware which stop the overheat counter in this mode to stick to their user manual.

Now, it's also quite interesting cause it means there are room from improvement.
Either with simple thermal pad or better, a redesign of the plate by some smart second part.
Ans, of course a software hack to get rid of the time limitation.

The fact that the camera have a huge potential may help the community to create something better out of it.
When Canon released their first 5D2 with video mode, they never though it will become such a trend. The public push and they answered it at one point.

Let see what hack will bring to this one.


 

[User]

1 hour ago, Cliff Totten said:

Andrew....I have said this over and over and over again. The heat condition of the R5 is not an engineering "mistake" on Cannot's part. It is literally doing exactly as it was designed to do from day one. Cannot wants the R5 to be used only in specific ways and this does not coincide with the way "we" want to use it.

We have a need that we want the R5 to fill. However, Cannot wants to sell us a different camera for our need.  Yes, the firmware is deliberately highly restrictive. They are actively "steering" their customers the best way the know how with these crippling tricks.

As customers, we need to stop thinking the R5 was an engineering mistake and simply let accept it as a casual camera that has zero overlap with any of their true camcorders.

Cannot will never stab their Cine EOS team in the back....no matter how bad we want them to. If anybody thinks the Cine EOS team doesnt care about the R5 and was not in many of the R5 devopment meetings and planning discussions....you are crazy!

Amen.

It also needs to be said that forums like this sustain themselves (and actually thrive) on exactly this type camera release with irrate champions pushing on both sides.

At the end of the day this camera is 'capable' of recording 8k... just not for very long. Better to drop the expectations and actually choose the right tool for the job when the hype-smoke clears.

 

[kaylee]

once people dig deep enough into the firmware we're gonna find cripplehammer.exe

[crazyrunner33]

14 minutes ago, kaylee said:

once people dig deep enough into the firmware we're gonna find cripplehammer.exe

Well played. lol.gif 

[horshack]

"Why does the Canon API documentation claim you are able to poll the temperature, but the camera only returns a text status like “Normal” rather than a real temperature in Celsius? Why is this all-important temperature readout obscured?"

The precise internal temperature is available in the EXIF metadata of R5 images. If you'd like to experiment on thermal shutdown temperature thresholds/cooldown periods you can take a photo before/during/after the test and use the EXIF of those photos to monitor the temperatures over the period of the test. For example, here's an exiftool invocation to get the internal camera temperature for all R5 .CR3 raw files in a directory:

exiftool.exe -CameraTemperature "c:\mypics\*.cr3"

 

 

 

[horshack]

The best way to determine if the time limit is firmware-related is to attempt to defeat the firmware's mechanism to record the time/states associated with the thermal shutdown. All camera settings and firmware state variables are stored in NVRAM. The settings are typically only written out when the camera does an orderly shutdown (optimization). You can verify this yourself by changing one of the shooting parameters (aperture, shutter speed, ISO) and then pulling out the battery while the camera is on. When you plug the battery back in you'll likely see the camera didn't save that change. The state variables associated with any potential video timer/limit/thermal threshold might be defeated using this same technique.

[jpfilmz]

Crippleware at its finest.  I mean seriously...you can shoot raw on the 5DMk3 and below.  IN DIFFERENT RESOLUTIONS. It is laughable at how purposefully limited the R5 is by Canon.  I believe this camera can shoot 8k,6k,4k,& 1080 raw with no issues but Canon wont program it to do so.  I don’t believe any excuse Canon makes nor canon apologist make regarding this  silly segmentation game they are playing.

[EphraimP]

5 hours ago, Adrian Bacon said:

 I’m not saying that having 4K HQ or 8K isn’t desirable. It is. It’s just that those Modes are best reserved for special shots, and getting them should be planned for accordingly..  as it is, shooting anything at 4K at any quality with the bit rates canon has been doing is really expensive on the rest of the video chain, especially if you shoot a lot of content. It’s just not viable to shoot a lot of 4K unless you have a big budget, and if you have a big budget, then why not get an appropriate camera?

That's complete bullshit. As others have stated, a 4K workflow, even for 1080 delivery, is standard now. I've got about 20 Terabytes of storage sitting on my desk top right now, in mixed SSDs and spinning disks. Cost a couple hundred bucks. 3 Terabytes worth of SSDs in my editing machine, cost a couple hundred more. The 1 Terabyte drives I use for my two Ninja 5 recorders cost a bit over a hundred a piece. Really big budget.

6 hours ago, Andrew Reid said:

Adrian here is the ultimate proof that people are blinkered idiots, that live under rocks. They miss long term trends and they even miss how millions of people use cameras differently to how they do.

As for professional camera for video... I refer you to the famous Canon EOS R5 marketing and PR statements!

https://www.eoshd.com/news/opinion-on-dpreviews-canon-eos-r5-overheating-test-in-4k/

Even Richard Butler at DPReview didn't absorb or register the Canon statements about professional video and cinematography use.

It's unbelievable it really is.

Why is the world so utterly full of chinless wonders who wander around the planet with their eyes half shut??

Why are we feeding the troll? I know, I feel like doing it too. But it's pointless. He either doesn't actually know much about video production, or he's a Canon shill, or, most likely, he's just here to troll. Lame.

 He doesn't believe in the cripple hammer, which is unfortunately very real in this case, hit him with the ban hammer.

[John Matthews]

Has anyone considered sabotage or conspiracy as a possibility? Was it an engineer? Canon itself? A fan manufacturer? Sony? So many players... this could be a film. 😀

[Geoff_L]

So, shouldn't money be raised and gave to magiclantern, in order to reform the past team and hack this to give us something usable 😄 ?
Concerning the "it is not a professional video camera" argument (or excuse), something is worth mentioning : at forums, especially dpreview, everybody was complaining that canon gave the camera to videographers/ youtubers to review, thus the first infos we had were almost only on the video side of the camera. They were desperate to find a proper photography oriented review. In France, one of the biggest video youtubers came back to canon, from sony, and was given the r5 to make videos, review, long lives etc.
So yeah, "not a professional video camera", but, as Andrew said, a heavily marketed "video camera", gave to videographers by canon to review.

 

[Stronz]

It feels like the industry as a whole has some kind of collective Stockholm Syndrome with Canon. Why do we want them to be better? Why do we want them to fix their clearly intentional choice? Why do we want to pay them money when they clearly don't care to earn it?

Sony and Panasonic have been pushing the limits of hybrids for over a decade and yet many people are still hoping and praying Canon releases a patch or a fix for the R5 & R6. If you put your faith in Canon, prepare to be disappointed. 

[MrSMW]

1 hour ago, Stronz said:

It feels like the industry as a whole has some kind of collective Stockholm Syndrome with Canon. Why do we want them to be better?

For me, it's not Canon per se, but all of the manufacturers because it seems not one of them can manage to produce a hybrid without at least one serious flaw.

Re. Canon in general, I suppose it's because more people have had or have Canon than any other brand and if you want a pure stills camera, they have always made some of the best and if you want a cine camera, ditto. They just refuse (it seems) to want to fuse the two which is one of the biggest demands in this day & age.

Personally, I am brand agnostic and though I do favour Fuji, I probably would have bought a pair of R6's had they not been crippled (for my needs) and seriously considering the new Sony for 2021. I'd have jumped into bed with the Panasonic S1 this year if the AF had been up to snuff and the Z6 last year if it had had 4k 60p internal.

If they can fix some of the issues with a flick of a switch, then in an ideal world, folks will vote with their wallets, ie, not purchase on principal. But back in the real world, if a 'thing' can be made to work, a lot of the folks who were adamant they would not purchase, will find a reason to do so.

And there are the serial camera collectors who will buy it just because it says 8k on the tin. They will never use the 8k and never actually use the camera for anything other than some snaps of; their kids, their cat, their garden and maybe a brick wall. A very long lens will be attached and it will hang around their necks as a badge of honour. If they are at a wedding and you are there to officially capture it, they will wish to let you know early on in regard to the camera jewellery they are transporting around. And if it's got the potential for 8k, they will feel especially smug. Those people will buy the camera in droves.

But I digress... All the talk of how to fix the thing outside of Canon recalling and doing it themselves is madness.

If they don't wish to do that, then fair enough and they will lose a lot of the pro market, maybe just for a time, maybe for ever and that may or may not hurt them, - it depends how much of their market is pro. I have no idea what it is but suspect it's not enough to break them.

In principle, I'd like to see it working though just because any product brought to market should work and for the overall good of our industry, but maybe there is a positive from all this...

Maybe...and I hope it's the case, those companies working on new products such as Nikon, Fuji and maybe Panasonic, who will all be aware of what has been unfolding with Canon over the last few weeks, will ensure that their own products are not open to such criticism and actually end up being better than they otherwise might have been (without the Canon debacle)?

[Torben]

A qualified guess would be, that it is the CCD sensor that overheats - not the CPU. 

As an electrical engineer, I am surprised to see how much attention is drawn to the heating of the CPU. The CPU is appartement not the problem - it is not heating up, is has no thermal paste, and it is still not heating up.

Most of the power ends up in the sensor chip. It has to move a lot of information in a very short time, where the CPU merely has to move data to RAM and memory cards. The CPU process is much less power demanding than moving analog pixel information around the sensor and converting them to digital. So my guess is, that it is the sensor that needs cooling. 

It would be interesting to see measurements of the surface temperature of the sensor chip, and to see if there is a relation between that, and the "cool down timing" in the camera. 

BTW: Silicon is a heat isolator (as well as en electrical insulator). Dispersing heat from silicon is a delicate task, and even more so in a sensor chip, that only have thermal contact from one side. 

[flipside]

1 hour ago, Stronz said:

It feels like the industry as a whole has some kind of collective Stockholm Syndrome with Canon. Why do we want them to be better? Why do we want them to fix their clearly intentional choice?

For competition. If Canon continues to produce duds then Sony et al.  will have less incentive to improve their products.  I'm a Sony user, but I would be very happy if Canon released a fix that made the R5 a better video than the A7s iii, even though I would almost certainly not buy it.

[UncleBobsPhotography]

I tried normalizing some cut outs from the video, flipping the back panel to make it easier to compare. 

There are 2 air gaps and one PCB board between the heat conductor covering the chips in picture 3 before it gets to the back panel. I am not sure what the white and black material in the first picture is. If the white material is a soft heat pad, then it would provide fair conductivity to the PCBs and eliminate one air gap. It could also just be the magnesium chassis. The black rectangle is placed right above the hot parts, could be a thermal pad. There seems to be one copper part on the top left under the wiring. This could mean the small white chip on the top left in the second picture also generate heat. 

The heat conductor in picture 4 seems to only distribute the heat around the internals of the camera, which means the heat will slowly reach the chassis after heating up the internals and conducting through the air gaps. This conductor has a thermal pad to the back of the PCB with the main chips (CPU?), but I would guess that the main purpose would be to redistribute heat from the image sensor since there is a large surface facing the sensor. However, because of the IBIS there is an air gap between the sensor and this heat conductor.

Everything that generates heats except for the memory card is covered by PCBs, which makes it a lot more difficult to make any improvements. The easiest and maybe most promising DIY improvement would be to cover the parts with electrical insulating thermal paste. Silicon thermal paste is much worse than metal thermal paste from a thermal perspective, but should be safe to add anywhere as long as it doesn't get near the IBIS. It would make it a mess if you ever need to have the camera repaired.

 

 

[stv]

I know this is probably crazy talk but I wouldn't be surprised if the R5 isn't overheating at all and Canon have just implemented a time based system to purposely make the HQ modes completely unusable.

Armando Ferreira tested the camera and he was getting the overheating warning despite the camera and battery being cool to the touch (that part of his review can be seen here https://youtu.be/E0851VJfPmg?t=629).

Canon have long been criticised for crippling the specs of their cameras so they've decided to take a different approach with the R5 and crippled the performance instead. That way they can say "Hey! We listened! We've given you a camera with amazing specs. We tried our best!".

If sales figures for the R5 are extremely poor and they suddenly release a firmware that "magically" solves the overheating issues then you know something is up.

[UncleBobsPhotography]

With how poor the heat design of this camera is, it should be pretty easy for them to create a fantastic cinema camera with the same sensor and internals but with a proper cooling system. From how the R5 looks it shouldn't even be necessary with fans as long as they use proper heat pipes and heat sinks. However, it doesn't seem like any of the new cameras, R200, R300, XF505 or XC20 is going for this approach, which makes me wonder what they are trying to protect.

It should be noted that I was so impressed by their new technology (and the 40% drop in their stock price) that I bought Canon stocks a few days ago, which makes me a bit biased. Canon's problem has been that they haven't had the technology to keep up. Now they have a great photo camera (I believe the R5 will be good for photo but crippled for video), and the technology to release quality products the next few years.