Capacitor Standby Power Mod

@upsss, Have no idea what you mean by this. too late at night here.

When the camera turns off, Vcap falls instantly from 3.8V to 2.9V. The voltage then starts to fall over time until it reaches a low enough level that settings are lost. The mod is designed to maintain Vcap at 2.9V from the converter which is connected to an "always on" 12V source.
 
The graph on the data sheet shows the diode conducts 10mA at 0.6V forward voltage. It doesn't show what happens at lower voltages, but suggests it never actually drops to zero. Without testing it would be hard to predict what voltage the capacitor would reach without the resistor. I think it would be OK but it should be checked.

1N400X.jpg
 
Last edited:
The capacitor will never charge to more that 5V minus the 3 junction diodes. With the camera Off the capacitor sees a constant load (although very small) from the RTC and the processor which will be in sleep mode including its own leakage. Wasn't that the original problem that the OP tried to solve where the capacitor over time lost charge and the camera lost its settings?

the real problem is the camera doesn't have an RTC button cell, if it did none of this would be needed, unfortunately it's a battery based design which has been modded (hacked poorly) to use capacitors
 
The factory cap conversion may be a 'hack job' but it's still a very usable 'hack' for most of us living in hot climates. Few if any other 'budget' cams let you swap back and forth between battery and caps this easily if at all, and if it were not for this I'd be putting in a new battery about 3 times a year :eek: There's still a lot to be said in favor of this older design even with it's quirks, Mobius has become iconic in the history annals of dashcams already and it was never really designed for this use.

The temp restrictions of cam batteries is a real problem as about half the world really needs something better for dashcam use. That's not a direction the battery manufacturers are currently going in since almost everything else does OK with today's battery types. The push with them right now is toward improved energy densities and higher amperage capacities, and unless some new battery technology comes along and gets into large-scale production we're still going to be using caps in our dashcams ten years from now for lack of a suitable and affordable battery :(

So we're still stuck where we have to either settle for less than we need and make compromises, or find alternative solutions like Cobber has done to gain what we want somehow. The perfect dashcam will never exist but we're not really even close yet. That is a shame because so much more is possible right now but it's not happening nearly as fast as it should.

Phil
 
The graph on the data sheet shows the diode conducts 10mA at 0.6V forward voltage. It doesn't show what happens at lower voltages, but suggests it never actually drops to zero. Without testing it would be hard to predict what voltage the capacitor would reach without the resistor. I think it would be OK but it should be checked.
View attachment 28531

@upsss, Have no idea what you mean by this. too late at night here.

When the camera turns off, Vcap falls instantly from 3.8V to 2.9V. The voltage then starts to fall over time until it reaches a low enough level that settings are lost. The mod is designed to maintain Vcap at 2.9V from the converter which is connected to an "always on" 12V source.

Exactly as I predicted (without the 360 ohm resistor) Simulation 101, Camera On capacitor voltage 3.8V, camera Off ~3.4v.

R1 simulates the processor standby load. R2, R3 simulates PS source impedance, .IC Initial capacitor voltage all needed for proper simulation.
 

Attachments

  • Camera ON.PNG
    Camera ON.PNG
    17.8 KB · Views: 13
  • Camera OFF.PNG
    Camera OFF.PNG
    17.6 KB · Views: 13
Last edited:
the real problem is the camera doesn't have an RTC button cell, if it did none of this would be needed, unfortunately it's a battery based design which has been modded (hacked poorly) to use capacitors

As I found, it can be re-hacked :) and RTC cell can be added very easy : I used CR2032 laptop backup battery with wires.

Just disconnect D3 that supplies RTC power and connect 3V battery (+) to this point. The (-) of the battery connect to any GND point.
I suggest to make 1st connection via 10 kOhm series resistor and to measure battery discharge current: the scale is 10mV/uA for voltage measured on 10k . In my measuring the RTC consumption is about 1.5uA, that gives for 225mAh CR2032:

(225000uAh/1.5uA/24h/365d) *0.75= ~12year. (0.75 is 25% self-discharge factor).

The battery should be placed as close to the PCB as possible due to speaker, thus I didn't use any glue.
I tested it for ~24h, seems to be OK, before it lost date in ~5-7 hrs.

About a question: why not to do this in a "classic" way with 2 diodes to prevent battery discharge during power-on? The answer is that this causes more bad than good :) . I checked this point, it works, and this can really save battery power if you drive more time than the car is in standby. But my car is 80% of time is in standby, so I can save only about 20%. But even for good schottky diodes leakage at hot temperature (during parking) is 2-4 times more than RTC consumption, thus we save 20% but loose 300%. You can use low leakage silicon diodes - but their voltage drop is too high and you also loose battery lifetime. If you need - you can place 2 diodes - it works, but don't use original D3 - it leakage is unacceptable :)

G1W_RTC.JPG G1W_final.JPG G1W_current.JPG
 
@EddieL
Very interesting. I still haven't installed my modded camera. Firstly because I had to get an auto electrician to run some cables from the fuse panel to the trunk. Secondly because it's been too hot and I wanted to hardwire my A119 as my front camera first.

I am using the G1W temporarily via the cigarette lighter socket at the moment and the date resets overnight. All my other settings are remembered.

I'm beginning to wonder if that D3 you mention may be open-circuit? Hopefully I will get it installed soon to test my mod. At the moment I'm having trouble finding a suitable way to mount it that doesn't just give me big balls of light from headlights at night. Need it to point down lower, but the suction cup mount doesn't let it.
 
@EddieL Nice and clean- good job :D I am wondering where the red wire from the RTC (+) pad goes, looks like it's headed to the USB power input socket. That would be bad news if the battery is seeing 5 VDC+ via the PS when the cam is powered up. I'd look for myself but I don't have a G1Wc on hand ;)

Phil
 
looks like it's headed to the USB power input
No, this net (D3 cathode) supplies power to the real-time clock circuit only (dedicated pins of the processor). D3 anode is connected to a local 3.3V regulator, thus D3 supplies 3.3V to the RTC from USB/main battery/supercap. When RTC is supplied by an external source, D3 is redundant.
 
Back
Top