Not a Happy Camper with the A139

Well, the fan should be only for emergency purposes while parked in case of overheating, not as a permanent and all the time solution. Not to mention you will probably be irritated by the 65+°C more than the noise of small fan. :p
So currently it shuts down when the cabin air temperature reaches 65°C, you add a fan, the camera will then shut down when the cabin air temperature reaches 65°C. What effect has the fan had?
 
So currently it shuts down when the cabin air temperature reaches 65°C, you add a fan, the camera will then shut down when the cabin air temperature reaches 65°C. What effect has the fan had?
Lets say it shuts down when air temperature reaches 65°C, what si the temperature of the Novatek's silicon? 100°C? 90°C? 80°C? (and this temperature si distributed to the heatsink) Without fan the 80-100°C air between fins diffuses from the heatsink by it's own speed (even worse when the heatsink is basically upside down as it is in A139), due to low speed it accumulates air with higher temperatures near the heatsink, which decreases the temperature difference, which lowers the diffusion.

By increasing flow of the "hot" 65°C air you will provide "fresh" medium with higher temp difference which will increase the diffusion speed and also actively push the air outside the dashcam. That will lead into lowering the temperature of the silicon (closer to the ambient) and thus the dashcam will run just fine even at higher ambient temperatures (because of lower ΔT).
 
By increasing flow of the "hot" 65°C air you will provide "fresh" medium with higher temp difference which will increase the diffusion speed and also actively push the air outside the dashcam. That will lead into lowering the temperature of the silicon (closer to the ambient) and thus the dashcam will run just fine even at higher ambient temperatures (because of lower ΔT).
That may reduce the temperature of the processor, but it will increase the temperature of the super capacitors, memory card, and sensor, which are the three most critical elements in the dashcam... so resulting in failure...

Whereas an external fan will simply cool the camera, resulting in success, maybe temporary success, but no reliability issues.
 
  • Haha
Reactions: EGS
The same inaccurate description as @Nigel's, it depends on what you consider as hot. It will blow "hot" air over "hotter" heatsink to make the heatsink only "hot"...

As long as there is a temperature difference, you can make the heat dissipate. And since "quite accurately" 30-40°C difference was measured, there is a room for improvements (larger area/faster flow).

Well it "depends". In my case, the hottest part of the car would be the windshield, until the rest of the car heats up. So a fan sucking in air from the car's interior and blowing it over the heatsink would be cooler. As the ceramic tint is capturing most of the heat until at some point both the tint and car reach an equal temperature.

On a car without tint, the Windshield and Interior would be about the same. So the fan would be sucking in that same hot air and blowing that same hot air over the heatsink. Whether the camera's CPU and Internal Temperature is hotter than the car interior's ambient air would be the question. It might be somewhat cooler, but I'm not sure enough to make a difference. As it'd be very hot air blown over extremely hot CPU.

No, now you don't get the point. If in a black car A139 fails at about 70°C and doesn't fail in a white car at the same temperature, then colour effectively doesn't matter at that point (not to mention the tint which, according to your description, should shield part of the energy and thus make the camera to not heat that much from Sun as in other cars). One only could point out that the black one needs less time to heat up which could lead to a "heat shock", compared to a white car which heats slower and gives a dashcam more time to "sync the temperature".

This really depends. You'd have to place two cars (white and black) side by side and measure the time it took for the interior cabin to reach the same temperature. There will definitely be a time difference, but how gradual or pronounced, I don't know. As ultimately, the sun is still shining through the windshield into the car. It's the car's body that is going to not get as hot as fast.

I suck at math. I don't know how much heat transfer from the body to the interior equates to the interior cabin warming.

With my car, the variable is the ceramic tint. Yes, I have a dark car (doesn't help), but the tint is the driving factor. The tint is sucking up all the heat. And the camera is sitting on that very hot tint. Meaning the camera Base and Camera are getting much hotter before the interior of the car.

You could take a White Volvo S60 with 80% front ceramic tint and run same test. See how much a difference there is. Again, I don't know.

Don't you have a something like hole in ozone layer over Sweden, that more energy leaks to heat your dashcam more? That's why I suggest we should rather bake our dashcams in ovens to finally get to some accurate conclusion, instead of temperature and Sun hunting. :D

The Sun's Rays and UV light are what heat up the car. None the less, I do believe Dash Cam makers run "Over Tests".

If the temperature was higher than 65°C, anything is possible. Unfortunately that was still the time before the meat thermometer.


I'm not sure why you're still pulling this out. Why is it so important? Nobody doubts that your files are getting corrupted.


#57 You had the thermometer sticked into your car's housing, if I recall correctly, and you yourself doubted about the proper placement. That's why I haven't got that measurement into account.


I wouldn't except less.

We're not having to be 100% accurate on the actual temperature. What is important is consistent results. Science is about trying to "Disprove" a result to ultimate determine if the results are valid. I ran this test NUMEROUS times and the results remained the same. Somewhere around 65-67C on the meat thermometer, the car shut off (Camera #1).

On Camera #2, the Camera also shuts off with overheating. Didn't meat thermometer the 2nd one.

Ultimately, what I know from these results is that once the camera overheats, it improperly exits the last file written. Resulting in Corruption.

Ok then, if even Viofo admitted it, you managed to improve their product. Let's hope it won't be "killing" too soon the units which happen to be capable to run at even higher temperatures. I still have PTSDs from my DR900S...

This is my goal. The whole point of DashCamTalk. To provide companies with real world feedback. Viofo has seen my pictures of the heat thermometer. Viofo has seen that at about 65-67C (guestimated) the camera is shutting off and exiting the last files improperly.

Viofo has acknowledged my findings and stated that they hope this next firmware will address the problem.

65C is the cameras "maximum operating temperature", but not a shutdown temperature. So the camera itself is regulating overheating versus say Viofo's Software telling it to shut down before overheating to prevent corruption.

This is going to be a fine balance. I don't own a heat gun. So even if my car is 65C within the interior, the camera itself could be ANY AMOUNT hotter (80C 90C, etc) due to the ceramic tint and sitting on the windshield.

But the point remains, Viofo should devise a means of making the camera shut down properly before overheating. So you don't lose the last files corrupted.

I hope they will also implement a recovery measure when it cools down.


Remember, I have two units as well, both run simultaneously in one car. One is exposed to sun, the another is not. Both were exposed to 70+°C ambient temperature inside the car. Never encountered the same behaviour as yours.

Again, I 100% agree.

1. Recovery Method On overheat = Important. The camera just shuts down and never returns to life.
2. Properly exiting last file written.

See now we're both thinking on how to improve Viofo's A139!
 
That may reduce the temperature of the processor, but it will increase the temperature of the super capacitors, memory card, and sensor...
How exactly pushing air with lower temperature increases temperature of capacitors which are sitting about a milimeter above the processor shield on which is the main heatsink, few milimeters from the heatsink, temperature of a memory card which shares the PCB with the processor (few milimeters from it) and effectively shares the main heatsink as well, and temperature of the sensor which is in separate almost completely closed section with only small openings?
 
How exactly pushing air with lower temperature
But you are not pushing air with lower temperature, as you explained above you are allowing the processor to continue working at higher air temperatures before it shuts down, and at higher air temperatures it is inevitable that the image sensor will be hotter since your new fan does not cool it. The memory card may be close to the processor and share a heatsink, but there is still a significant temperature gradient between them so the memory card will be somewhat hotter, and the memory card is often said to be the critical component.

and thus the dashcam will run just fine even at higher ambient temperatures (because of lower ΔT).
 
Adding cooling to the processor will always help as it's likely the hottest component involved, and part of it's heat goes to the PCB heating all the other components mounted there. Adding cooling to the entire system via forced air (especially across the processor) will also help. The big question is how much will it help?

If it's other components reaching their thermal limits then adding cooling to the processor via improved heat-sinking won't help much, but forced air cooling across everything inside the cam might. It's all a matter of what can be gained and only experimenting will tell. Probably the easiest way to test this is to make a ring of modeling clay or blu-tack encompassing the vents in one side of the cam, then sealing a fan to that from the outside so that all airflow goes through the cam. If it doesn't work then removal and clean-up is easy leaving an unaltered cam to try a different approach with. Experiments and tests needn't be pretty- only permanent solutions need that ;)

The potential drawbacks I see are increased power draw from the car battery and possibly some EMI from having the fan's electromagnetic field so close to the PCB. The latter might be worsened if the fan is sharing the 5VDC power source with the cam. Given the temp extremes and the limits of many electronic components being approached I doubt you'd gain more than 5C degrees of ambient air temps but that could well be enough to cover most folks problems.

Phil
 
But you are not pushing air with lower temperature, as you explained above you are allowing the processor to continue working at higher air temperatures before it shuts down, and at higher air temperatures it is inevitable that the image sensor will be hotter since your new fan does not cool it.
Since A139 is basically a heat chamber, all the components are heated above the ambient temperature by the processor and other components. When you push the air, let's say you input 65°C and it heats up to 67-70°C which could be the exit temperature, instead of suffocate the whole system by 80°C which somehow escapes on by itself, it will decrease the overall temperature inside the camera.

The memory card may be close to the processor and share a heatsink, but there is still a significant temperature gradient between them so the memory card will be somewhat hotter, and the memory card is often said to be the critical component.
Are you sure about that?
Next to my front A139 there is a housing for cameras/sensors of the car so I wasn't able to get a better view, but I'm pretty sure with better angle and more inside the slot I would be able to measure basically the same temperature as on the heatsink (still 56°C vs 55°C...).

Nonetheless, many endurance memory cards are rated for 85°C so they can withstand higher temps, supercaps with higher temp rating exist as well (but I bet the current ones in A139 in 65°C ambient run above the max anyway). I wonder about the sensor, I wasn't lucky enough to find the datasheet for IMX335 but some manufacturers of camera modules rates them in range of 60-85°C so probably even the sensor could be able to handle higher temps.
 
The potential drawbacks I see are increased power draw from the car battery and possibly some EMI from having the fan's electromagnetic field so close to the PCB. The latter might be worsened if the fan is sharing the 5VDC power source with the cam. Given the temp extremes and the limits of many electronic components being approached I doubt you'd gain more than 5C degrees of ambient air temps but that could well be enough to cover most folks problems.
No doubt about the power draw but again, I mentioned fan only as a possible solution for overheating (I'm unsure myself, it's only an idea which seems to me like worth a try). Don't know about you but I would rather sacrifice some time on battery in exchange of working camera while parked. But since I have a battery pack, I probably see it from a different point of view (don't know what are the average times with HWK on car batteries)

I wouldn't be concerned much about EMI, it should be far enough and probably behind a heatsink.
 
Ok, I'm maybe convinced to give it a try and spend some of my time on this in the near future. :unsure: I took an old 5V/0.09A 25x25mm axial fan (pulled out from something long ago due to extensive vibrations), hook it up to a charged Li-Pol battery (means around 4V with such small load), sticked it with double-sided tape on the bottom grill(through which I done the measuring above) over the heatsink and after 10minutes temperatures dropped from 61°C to 45°C at the top of the heatsink, from 60°C to 54°C at the SDcard, 24°C ambient. After 30min, to 44°C heatsink, 46°C SDcard, 24 ambient. Not too shabby.

Of course this was "tested" only in my garage but hey, I was like half asleep, in nothing more than my shorts and with parts which were easily within reach of my hand, so leave me alone. :ROFLMAO: Too bad (or is it? :unsure: ) I haven't encountered any overheating issues yet to be able to tell if it's able to made the difference or not, and summer temperatures are not that high either.
 
If it's other components reaching their thermal limits then adding cooling to the processor via improved heat-sinking won't help much, but forced air cooling across everything inside the cam might. It's all a matter of what can be gained and only experimenting will tell. Probably the easiest way to test this is to make a ring of modeling clay or blu-tack encompassing the vents in one side of the cam, then sealing a fan to that from the outside so that all airflow goes through the cam. If it doesn't work then removal and clean-up is easy leaving an unaltered cam to try a different approach with. Experiments and tests needn't be pretty- only permanent solutions need that ;)

Phil

Are you suggesting mounting a fan to the external case (over the vent holes) of the A139 to suck hot air from the unit itself?
 
Since A139 is basically a heat chamber, all the components are heated above the ambient temperature by the processor and other components. When you push the air, let's say you input 65°C and it heats up to 67-70°C which could be the exit temperature, instead of suffocate the whole system by 80°C which somehow escapes on by itself, it will decrease the overall temperature inside the camera.


Are you sure about that?
Next to my front A139 there is a housing for cameras/sensors of the car so I wasn't able to get a better view, but I'm pretty sure with better angle and more inside the slot I would be able to measure basically the same temperature as on the heatsink (still 56°C vs 55°C...).

Nonetheless, many endurance memory cards are rated for 85°C so they can withstand higher temps, supercaps with higher temp rating exist as well (but I bet the current ones in A139 in 65°C ambient run above the max anyway). I wonder about the sensor, I wasn't lucky enough to find the datasheet for IMX335 but some manufacturers of camera modules rates them in range of 60-85°C so probably even the sensor could be able to handle higher temps.

There's a couple issues at play here, but money is the answer. Unfortunately, heat guns that give a digital readout with a display are around 250 or 300 Euros. The cheap ones just show the temp of what you're pointing the heat gun at. Too touch for my blood, since I already bought this unit and am not a beta testing something I got for free.

I see several issues that are resulting in the overheating (outside of the ceramic tint) on my vehicle.

1. Camera is housed in a black casing. That means the camera is absorbing ALL the heat from the windshield. A light color camera casing, which more noticeable, would not get nearly as hot. The black case gets very hot to the touch.

2. If the car's ambient temperature is 65-70C, the camera mounted on the windshield is most definitely hotter. Not only is the CPU and PCB generating heat, but the black case is absorbing the windshield heat, too.

So the problem here is both the color of case housing and a need for better cooling options. Which still makes me wonder if a mini fan seen on a raspberry pi mounted to the heatisnk would make a difference.

I am sure mounting a camera over the vents externally to suck out the hot air would help. But of course that's extremely unsightly.

Would be nice to get an exact temperature measurement, but I'm not investing 250-300 euros to get exact number on Viofo's Product. It won't change anything as they aren't going to redesign the A139 based upon our suggestions.
 
Last edited:
My Nextbase 522 and 622 both had silver fronts, and they got hot. The 4k 622 moreso than the former. Sometimes I couldn't touch the cam. But it still continued to record., So a light coloured case is not necessarily a solution.
 
My Nextbase 522 and 622 both had silver fronts, and they got hot. The 4k 622 moreso than the former. Sometimes I couldn't touch the cam. But it still continued to record., So a light coloured case is not necessarily a solution.

It's just one element of the solution. I'm not saying the black casing is the cause, but it definitely isn't part of the solution either.
 
Back
Top