Not a Happy Camper with the A139

yea, it's been way hotter where I live, and global warming is a myth. "science"
 
Viofo had asked I try turning off loop record. In my tests, this seems to result in my Sandisk High Endurance Corrupting the Card and/or failed recordings. Problems I don't have when I use Loop Record set to 5 minutes. In Heat test with Loop Record On, Only last files being written corrupt on overheat shutdown. Firmware 1.2
 
I've often wondered if dashcams need to be stuck to the windscreen. If they could be mounted further back then they wouldn't be subject to direct sunlight and the subsequent heat issues.

Difficult one to implement though.
 
@Kremmen There are systems for motorcycles so not that difficult to make.:)

Even simplier since dashcam manufacturers started to use coaxial cables instead of solid USBs.
 
I've often wondered if dashcams need to be stuck to the windscreen. If they could be mounted further back then they wouldn't be subject to direct sunlight and the subsequent heat issues.

Difficult one to implement though.

I think the technology to proper cool 3 channels just isn't available. Meaning 3 channel setups run too hot, and in warmer weather, face the challenges of overheating.

The only working solution at present is to mount cameras on windshield but have these cameras field to a main unit tucked under the glove box or out of direct sunlight. Something no one has designed yet.

Getting the main unit (holding the sd card and handling processing) out of the sun would eliminate the problems I've described.
 
I think the technology to proper cool 3 channels just isn't available. Meaning 3 channel setups run too hot, and in warmer weather, face the challenges of overheating.
You mean heatsinks don't exist? And if one wants a smaller heatsink then a fan is the solution to increase the flow of the medium.

And if even the ambient temperature is too high for the silicon, then the only solution is to use "a better one". But since A139 in 25°C ambient runs at about 60°C, then in 65°C ambient (means the Viofo's advertised maximum) it must be super hot and there is "a room" for improvements on the heatsink side.

The only working solution at present is to mount cameras on windshield but have these cameras field to a main unit tucked under the glove box or out of direct sunlight. Something no one has designed yet.
E.g. Viofo MT1

Getting the main unit (holding the sd card and handling processing) out of the sun would eliminate the problems I've described.
Would reduce the problems. Which I still haven't encountered (despite measured 75°C). I'm still interested in your measurements via something more accurate than meat thermometer (then in your car could be even higher temperature than 70°C).

Actually the fact I still haven't seen the issue is a pity because basically I have the perfect setup for the testing of the proposed idea of not exposing it to Sun. In my case the front unit should fail and the rear should be still working. :unsure:
 
You mean heatsinks don't exist? And if one wants a smaller heatsink then a fan is the solution to increase the flow of the medium.

Heatsinks and thermal paste exist, but are inadequate to disperse the heat sufficiently to properly cool a 3 channel setup. I'm unaware of a camera that uses a fan. Not sure how viable a fanned heatsink option would be in a dash camera. As computers rely on fanned or water cooled setups. But a practical setup in a dash cam seems difficult.
And if even the ambient temperature is too high for the silicon, then the only solution is to use "a better one". But since A139 in 25°C ambient runs at about 60°C, then in 65°C ambient (means the Viofo's advertised maximum) it must be super hot and there is "a room" for improvements on the heatsink side.

From my tests with a meat thermometer, the car is getting 65-70C. Thus, hitting the Viofo's advertised Thermal Max. When this happen (on two different cameras I tested), the camera powers down improperly. Causing the last files (front + rear + interior) to be corrupted.

Now if you turn off loop recording and let the camera record to one huge file, I've found my SD card becomes totally corrupted until formatted. Happened Twice. A problem that I never experienced with loop record set to 5 minutes.
E.g. Viofo MT1

Not a 3 channel setup and it's a two channel motorcycle camera not designed for cars.
Would reduce the problems. Which I still haven't encountered (despite measured 75°C). I'm still interested in your measurements via something more accurate than meat thermometer (then in your car could be even higher temperature than 70°C).

Well considering my measurements have shown two different cameras overheating at the same thermal max, I'm pretty sure my tests are accurate. Seems Viofo's 65-70C is pretty spot on about when the camera shuts itself down. Seeing my car reached those temperatures, and the camera turned itself off, I'm not sure what more you want?

Short of me investing in a thermal gun, there's really no data points to add. My tests verify Viofo's max operating temperature. Or at least what it takes for the inside of a car to reach to shut the camera down unexpectedly.

Actually the fact I still haven't seen the issue is a pity because basically I have the perfect setup for the testing of the proposed idea of not exposing it to Sun. In my case the front unit should fail and the rear should be still working. :unsure:

@jokiin and I discussed this in private messages and have come up with several theories.

1. @SawMaster uses a fan. A van has a greater surface area and interior space to heat up. Meaning it won't get as hot as a Sedan (my car is a Volvo S60).
2. I have a dark colored car (Volvo S60). Yes the windows are tinted, but that only slightly prolongs the cabin heating up.
3. Remember, if you drive a white or light color car, it reflects sunlight. On a dark colored car, it absorbs heat.
4. I have parked my car in DIRECT sunlight for my tests. Again, dark color car heating up.

Not sure what you have attempted or not attempted. I am basing my results on the fact Viofo sent two me cameras that exhibit the same behavior. Leading me to believe the issue with overheating and corrupt files is endemic and not isolated. Seeing how both cameras corrupt the last file being written.

And while I would never turn off loop recording, I've found if you do, the camera corrupts entire SD card due to the same above issue. Thus loop recording isolates the problem to the last file being written.

Maybe a firmware update will change results. Viofo hasn't released or notified me of a released updated firmware to test.
 
On an Aside Note:

1. I've never had the camera overheat while driving. So that's a huge plus
2. It's like mounting your Cell Phone in the Window and using as a GPS. I've had my phone give me the message, Phone Too Hot, Charging Stopped.

I believe the A139 shutting down in an effort to preserve itself from heat damage is similar to my Galaxy S9 ceasing to chare. Both devices have thermal max set in place.

While not convenient or the most ideal, my tests seem to prove the limitations of the A139 and fall in line with Viofo's stated specs. Bit of a bummer the parking mode shuts down on a hot day, but don't think the end user can do more than try to park the car outside direct sunlight. It's a definite limitation acknowledged by Viofo with it's 65C stated operating temperature.
 
Heatsinks and thermal paste exist, but are inadequate to disperse the heat sufficiently to properly cool a 3 channel setup. I'm unaware of a camera that uses a fan. Not sure how viable a fanned heatsink option would be in a dash camera. As computers rely on fanned or water cooled setups. But a practical setup in a dash cam seems difficult.
There are heatsinks which are able to cool down even hundreds/thousands of watts of TDP... And since there is such big difference between ambient and chip temperature in A139, the heatsink could have been bigger...

Fan could be solution in case of overheating(if manufacturer thinks it's necessary). Don't know about you but I would rather have a dashcam which turns on a fan from time to time to cool it down during parking in summer, rather than shuts down and does nothing until driver return.

From my tests with a meat thermometer, the car is getting 65-70C. Thus, hitting the Viofo's advertised Thermal Max.
Oh so you finally acknowledged the Viofo's 65°C, ok then.

Now if you turn off loop recording and let the camera record to one huge file, I've found my SD card becomes totally corrupted until formatted. Happened Twice. A problem that I never experienced with loop record set to 5 minutes.
Well, that's well expected since h264/h265 needs a "proper end" of the file... So in case of segmentation into smaller files the last gets corrupted, in case of one big file the whole file gets corrupted. That's why it's advantage to have dashcam footage divided into smaller files (to eventually corrupt only the last ones).

Not a 3 channel setup and it's a two channel motorcycle camera not designed for cars.
That's why it's an example. But "it's not that hard" to add a third de/serializer combo into A139 to make the front sensor remote as well. The point is, there is such solution...

I wonder what do you think it's the limiting factor to not be able to use MT1 in car.

Well considering my measurements have shown two different cameras overheating at the same thermal max, I'm pretty sure my tests are accurate. Seems Viofo's 65-70C is pretty spot on about when the camera shuts itself down. Seeing my car reached those temperatures, and the camera turned itself off, I'm not sure what more you want?
I thought your A139 were shutting down at around 70+°C. At least I remember questioning myself why you are dissatisfied with a camera not able to withstand 70+°C ambient despite the manufacturer claimed from the beginning it's for 65°C max...

My point is that all thermometers have a characteristics, and meat thermometer is not the kind which has to be accurate... So despite you see 65°C on the meat thermometer, it could be even 60, 70, 75°C or even something else. You know, cooking is not science...

2. I have a dark colored car (Volvo S60). Yes the windows are tinted, but that only slightly prolongs the cabin heating up.
3. Remember, if you drive a white or light color car, it reflects sunlight. On a dark colored car, it absorbs heat.
Well, these I said several post before, so yeah... :unsure: But still, 70°C ambient in a white car is the same temperature as 70°C in a black car.
 
Last edited:
2. I have a dark colored car (Volvo S60). Yes the windows are tinted, but that only slightly prolongs the cabin heating up.
In my testing, dark tinted windows absorbed more heat, and the temperature next to the glass where the dashcam is was hotter, the temperature of the seats was lower as less sunlight reached them, but that doesn't help the dashcam.

1. @SawMaster uses a fan. A van has a greater surface area and interior space to heat up. Meaning it won't get as hot as a Sedan (my car is a Volvo S60).
I think probably it is mainly that the van is taller. The hot air collects in the top few inches of cabin space, in a car that is where the camera sits, in a van the camera is down in the cooler air.

As for the fan, fans work well when they blow cool air over a hot heatsink, they don't work well when they blow warm air over a warm heatsink, which is what would happen in a dashcam, also fan life would not be great working in 65 degree plus air, they are all designed for 20 degree air, so they would add an unreliability problem. And as always, fans consume power, and that generates heat, and when they are not making much difference because the incoming air is already hot, adding heat from the fan makes them even less effective!

4. I have parked my car in DIRECT sunlight for my tests. Again, dark color car heating up.
If you have to park in direct sunlight, park facing north, problem solved.
 
As for the fan, fans work well when they blow cool air over a hot heatsink, they don't work well when they blow warm air over a warm heatsink, which is what would happen in a dashcam, also fan life would not be great working in 65 degree plus air, they are all designed for 20 degree air, so they would add an unreliability problem. And as always, fans consume power, and that generates heat, and when they are not making much difference because the incoming air is already hot, adding heat from the fan makes them even less effective!
But you need to define what "hot air" and "hot heatsink" mean... As long as there is an energy difference, you will always be able to blow hot air over hotter heatsink and get better results than let it to difusion. In case of A139 we are talking about around ΔT=40°C, where if size is the limiting factor, you can use fan as the workaround.

+1 on the power consumption increased by the fan and therefore reduced runtime in parking mode. But still, at this moment it seems like when A139 gets overheated, it won't turn-on until ACC. I wouldn't consider a small fan which runs on its lower RPM as somewhat a large source of heat and therefore to be ineffective.
 
I've often wondered if dashcams need to be stuck to the windscreen. If they could be mounted further back then they wouldn't be subject to direct sunlight and the subsequent heat issues.
Dashboard reflections tend to become worse as you move the lens away from the windscreen. Plus, the windscreen is a nice stable thing to mount a camera on.
 
In case of A139 we are talking about around ΔT=40°C,
I don't think it is that much, if it operates up to 65 then 65 + 40 would be 105, which is too much for reasonable reliability, even supercapacitors die quickly at that temperature.

I wouldn't consider a small fan which runs on its lower RPM as somewhat a large source of heat and therefore to be ineffective.
Remember the camera only uses about 4 watts of power when using 3 channels, so adding a 1 watt fan is a significant power increase.

And how much cooling is a 1 watt fan going to provide with a ΔT=20°C ? Not very much, computer fans normally have a ΔT=60°C and use more than 1 watt if the computer is doing much.
 
Folding dashcams ( the ones that in the case of a front collision flip up and just film the sky after the crash ) hose are bad, i am as much interested in after event footage as i am of the lead up to and actual event.
And for that not to happen is hard if it is mounted further back, also the sensor array on the windscreen of many cars these days, mean you cant really do it or would have to install very offset to a side.

IMO the best solution moving forward are the remote cameras we have started to see on some systems, this should also be better for parking guard in sunny / hot enviorments, though i have not been able to verify this as it would nessitate several "similar" systems, and a more summery summer than the one we Danes got this year.
 
Dashboard reflections tend to become worse as you move the lens away from the windscreen. Plus, the windscreen is a nice stable thing to mount a camera on.
Also, any rain/dirt/scratches show up far more easily once they are not completely out of focus.
 
Also, any rain/dirt/scratches show up far more easily once they are not completely out of focus.
Not to mention windscreen heater wires!
 
The only working solution at present is to mount cameras on windshield but have these cameras field to a main unit tucked under the glove box or out of direct sunlight. Something no one has designed yet.

Getting the main unit (holding the sd card and handling processing) out of the sun would eliminate the problems I've described.
The SG9663DR, like the MT1 and others, show that dual remote cameras are technically viable. Going to triple remote is not such a big step given the hardware in the A139 is capable of processing three video streams, but as @jokiin has already pointed out it may not be commercially viable given the products already in the marketplace.
 
Indeed. You probably stand a better chance making a dual remote camera, and add some of the popular smarts to it.
Even if there is a substantial number of ride share drivers out there in the world ( none in Denmark as only Taxis are allowed here )
 
Going to triple remote is not such a big step given the hardware in the A139 is capable of processing three video streams,
Not sure it can receive three remote video streams, and even if it can, the local (front) stream uses a different communication method which allows higher resolutions than FHD, these days dashcams need higher than FHD resolution on the front. So you need a higher specification processor and remote transceivers including power transmission, and the extra remote camera, all of which makes them uncompetitive as general consumer cameras.

If you only want FHD resolution then no problem, the Viofo MT1 and SG9663DR do that, and presumably the upcoming generation of processors will be able to manage 2K remote cameras, at the cost of using the latest, greatest, most expensive processor for an unspectacular result compared to the 4K+2K resolution others will be using it for.
 
Back
Top