snapsy wrote:
It doesn't quite indicate the CFE card is the problem. By bypassing CFE the camera also bypasses the entire internal recording path, which includes bypassing the video compression path and logic through DIGIC. That is more likely to be the issue than the CFE card.
Exactly. It is worth to point out that the DIGIC processor, the RAM and the CFE card is very close to each other and heat from one or two of these will influence the other parts. It is quite interesting that the thermal pads are fully covering the RAM units above and below the DIGIC, but not the DIGIC itself, giving the clue that they may actually run hotter.
I’m starting to wonder whether the overheating is not so much the sensor and DIGIC processor (which I first thought) but rather the RAM and CFE writing circuits. But we don’t know which of the temperature sensors Canon hold as the reference to shut down the camera. It could be RAM, CFE circuits or the DIGIC processor being influenced by the hot RAM blocks.
You might wonder whether is was a good design to put the CPU+RAM so close to the CFE card, but having them removed from each other might be problematic too, adding electronic noise that might influence the sensor performance. These things are always a compromise.
I’m not sure about this artificial firmware crippling theory, but limiting the video recording time after having used the camera for other stuff that is not processor intensive might imply that their thermal management system is being over conservative or even bugged. It is quite simply to early to tell, IMO. But if you need to hate, you need to hate.
The bigger takeaway from the teardown is that the cpu is insulated by an additional PCB that prevents cooling the body externally. Without internal modifications that Tilta cooler is not going to do anything.
imho, it's highly unlikely for canon to release a 2nd gen ff milc and sabotage their effort to be competitive in a shrinking market.
and with all due respect, i sensed that nearly all photographers suddenly became experts in camera design, i.e. trade-offs and compromises, when the overheating issue became the headline..🤷♂️
for me, i'm still waiting for a definite conclusion on the r5's af ability to track erratic bif? 🤔
Jesse Evans wrote:
The bigger takeaway from the teardown is that the cpu is insulated by an additional PCB that prevents cooling the body externally.
That is a very good point.
It would still be nice if somebody could just use a visual IR thermometer to map an infrared heat map of the camera during heavy use. I would pay particulary attention to the heat map inside the CFE slot (taking the CFE card out and measuring right after the heat warning). Could be telling.
It would still be nice if somebody could just use a visual IR thermometer to map an infrared heat map of the camera during heavy use. I would pay particulary attention to the heat map inside the CFE slot (taking the CFE card out and measuring right after the heat warning). Could be telling.
It will probably be 'telling', but not in the way you think. IR thermometers handle reflective surfaces badly.
Jesse Evans wrote:
The bigger takeaway from the teardown is that the cpu is insulated by an additional PCB that prevents cooling the body externally. Without internal modifications that Tilta cooler is not going to do anything.
I could see the peltier plate producing temps 40-50 deg below ambient right on that back panel, and that in turn would have to have some sort of effect on the air inside at least. It won't transfer well to the main board however, due to how "insulated" that is with all the other boards around it.
I bet it reduces the temps about 10%, maybe 20%, but will that be enough, especially for the bulk and extra power needed?
Andrew J wrote:
I have a sick feeling this is so bad, lensrentals won't touch an R5 teardown with a 10 foot pole.
More likely is that Dr Cicala is making money renting out the R5 bodies he has in inventory since they are in high demand right now and cannot sacrifice having one out of inventory instead of having it out on a rental.
LR doesn’t make money tearing cameras down. They do it once one breaks to figure out if they can save money fixing the problem themselves. I’m sure we’ll get a very awesome tear down and analysis from him at some point.
I found these two posts interesting. I read the EOSHD rant and all 5 pages of link to the Chinese thread. The person cooled it down with thermal silicone but claims it didn't overheat in the first place if I understood that correctly. I think there have been other thermal tests from independent reviewers that showed it does. I might be wrong. Either way who is right and who is wrong. The topic of how is going to manage heat in that size of body came up long before the specs came out.
It would be worth trying to change the time of the camera by 2 hours to see if the recovery time can be fast forward it that is firmware related. Just a thought.
kezeka wrote:
More likely is that Dr Cicala is making money renting out the R5 bodies he has in inventory since they are in high demand right now and cannot sacrifice having one out of inventory instead of having it out on a rental.
LR doesn’t make money tearing cameras down. They do it once one breaks to figure out if they can save money fixing the problem themselves. I’m sure we’ll get a very awesome tear down and analysis from him at some point.
That's funny he did an R mount lens tear down saying it was the best built lens ever? I don't recall any mention of it breaking.
I am sure it's been tried, it has certainly been mentioned. There is a temp sensor on the main board, it would be a whole lot easier to just check temps periodically instead of putting in a timer based on current time once a threshold was reached.
What a huge, embarrasing, self-inflicted wound for Canon. It didn't have to be this way. They could have crushed the entire market at this price point and stood toe to toe with the A7SIII. As is, just another slap in the face for Canon video shooters.
deepbluejh wrote:
What a huge, embarrasing, self-inflicted wound for Canon. It didn't have to be this way. They could have crushed the entire market at this price point and stood toe to toe with the A7SIII. As is, just another slap in the face for Canon video shooters.
Has the artificial FW time limit been proven? Reading the EOSHD write up was nothing but speculation based on a test by someone unknown in China. I read all 5 pages of the the Chinese thread. Mostly dumb comments and Emojis. Maybe Canon did but I want proof from a known source. EOSHD. Drama = hits and hits = $$$$$.
The video files will be quite a bit smaller, which means I/O times are reduced, the CF card will work less, etc. This may have a slight impact on record time, coupled with an increased time limit. They didn't add this compressed raw 8K option because it has been reserved for their cinema cameras. However, I suspect they have no choice now to add it here.
However none of this fixes the heat up that occurs with just normal stills shooting, which then just continues to reduce overall video time.
Zenon Char wrote:
Has the artificial FW time limit been proven? Reading the EOSHD write up was nothing but speculation based on a test by someone unknown in China. I read all 5 pages of the the Chinese thread. Mostly dumb comments and Emojis. Maybe Canon did but I want proof from a known source. EOSHD. Drama = hits and hits = $$$$$.
There is no way to definitively prove the artificial time limit. There just is no other compelling explanation for how the camera behaves with regard to “cool down” after reaching the heat limit. There is virtually NO explanation for why the camera acts the way it does.
Leaving the camera in front of an air conditioner seems to have no impact on how quickly it cools down. This defies the laws of thermodynamics. The mass of the EOS R5 is not large enough and not nearly insulated enough to defy all of the efforts that have been made to cool it down.
You don’t really need to look any further than the fact you have to remove the cards to get the camera to let you record externally for longer amounts of time. While many pointed to the “CFExpress overheating” being the reason for this, that has been shown to be completely wrong, because leaving only an sdcard in the slot ALSO triggers the overheat warning prematurely.
Why are the cards even being considered at all when recording externally? It makes no sense.
The most obvious explanation for all of this bizarre behavior is in the software, which has had a bug triggered when there is no card present, the artificial software limitations have been bypassed.
Is it proven? Of course not. There’s nothing other than confirmation from Canon or looking at the camera source code which could give us definitive proof.
Zenon Char wrote:
Has the artificial FW time limit been proven?
The testing by the guy in China all but proves it, yes..
Cliffnotes for those who didn't bother to read the article.
1. The CPU does not have proper cooling. In fact, it doesn't even have its own heatsink.
2. The CPU, while its the hottest part of the camera, does not even get that hot before it enters "shutdown" mode.
3. Improving internal cooling had no effect on recording time limits
4. External ambient temperatures had no effect on recording time limits
5. Recording without memory cards to an external recorder exploits a loophole and allows the user to bypass artificial recording limits imposed by the firmware.
Bottom line, the recording limits of this camera are in no way a result of the thermal limitations of the camera. It is software based.
But then why does the back of the camera get quite warm? Where is that heat coming from if it isn't coming from the main board? What are the temps this person is seeing on the processor vs what others are seeing from the body underneath the rear LCD?
I won't trust what I see in that video for testing until it is confirmed by others. Like how was the temps recorded? Where is the thermometer in the camera that is seeing temps? So many questions with that video.
The cards having to removed to do hdmi to a recorder for long periods is indeed puzzling however, but the lack of heat in that mode is understandable. I bet Canon plugs that card slot hole for us with a firmware update.
