Mini 0803 capacitor version

[jokiin]

the caps need to charge prior to recording commencing if you want it to work reliably, delayed startup and shutdown are both supported in the SDK

My 0803 is plugged into a circuit which the car ECU always ensures is on for a minimum of 4 minutes so as long as my supercap charges within 4 minutes it will always work.

not sure what you're trying to say here

 

[Nigel]

the caps need to charge prior to recording commencing if you want it to work reliably, delayed startup and shutdown are both supported in the SDK
not sure what you're trying to say here

Just that because of where I plugged it in, my camera will always have 4 minutes to charge the capacitors before being turned off so as long as the capacitors can be charged in that time it is safe for the camera to open a file and start recording as soon as it is powered up.

Just used the capacitance calculator: http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html#c2

To charge 10F at 0.5A to near capacity requires 5 minutes with a 10 ohm (5W) resistor.

So it's not surprising it takes a while for them to charge, the camera wont be supplying that much current, not sure the standard 12-5V power adaptor can supply that much current. Maybe 10F is over doing the super caps?

 

[Nigel]

Although, I have a feeling that even with 12Fs I'm going to run into problems. The problem seems to be getting a full charge during my short (10-20 minute) drives which I can't quite understand.

What is the smallest capacitance that works if you give it half an hour to charge?

The smaller the capacitance, the faster it will charge - you are going in the wrong direction by having more capacitance than required!

 

[jokiin]

Just that because of where I plugged it in, my camera will always have 4 minutes to charge the capacitors before being turned off so as long as the capacitors can be charged in that time it is safe for the camera to open a file and start recording as soon as it is powered up.

they need to get charge into them and brought up to a stable voltage before recording starts

 

[niko]

I have experienced same problem with not saving last video file using stock firmware on my home-brewed DIY B40 Cap.
The only solution back then I found is manually to stop recording and then manually to turn off unit before turning off an ignition. This way it worked fine, but problem is that in 2/3 cases I was forgetting to do that manual stop / turn off procedure, so it is not practical solution, but still doable ( workable ) if needed.

 

[mr_milo]

OK I'll try to post some of the things I know.

Maximum charge seems to be 4.2V although after it reaches this value it will slowly lower to I believe about 3.9V before it starts to charge up again.
Minimum operating voltage seem to be about 3.0V for a successful shutdown.
With the camera off and just charging, the camera draws 25mA (<- Correction 250mA not 25mA). Yes only 0.025A. I feel this is the BIG problem and most likely the reason caps won't work unless the charge circuit can be modified. (<- Correction this does not seem to be an issue.)
The minimum capacitance that I have tested and got to work (on the bench) with a full charge was 6F. (<- Correction 3F not 6F).
On the bench (ie no GPS) my camera draws less than 500mA while running and charging. It does have a spike or two up to about 700mA but steady state is less than 500mA.

I knew the larger capacitance will take longer to charge but I'm also hoping it will take a lot longer to discharge. What I have noticed is that the lower the voltage gets the faster it loses charge. That makes sense if the power draw is constant based on P=VI. So far with the 12F (<- Correction 7.5F not 12F) bank I have had 4 on/off cycles with it operating as expected.

Later today I will try to run some tests to see how long the 12F (<- Correction 7.5F not 12F) bank lasts from full charge to camera off and how long it takes to charge.

Lastly, I looked at the code to try and find the values to edit. Unfortunately, I'm a bit illiterate when it comes to this kind of coding. I think I found the 1 min & 5 min values but couldn't find the rest. That confused me so I gave up for now. Really hoping someone that's better at it could modify it for me to test.

 

[Nigel]

The minimum capacitance that I have tested and got to work (on the bench) with a full charge was 6F.

To avoid possible confusion, am I right in thinking this is two 3F capacitors in series? (Which does not equal 6F.)

How much voltage is typically left in the capacitors after leaving it overnight? If they are still half full that makes the recharge significantly faster than if they are nearly empty...

 

[mr_milo]

To avoid possible confusion, am I right in thinking this is two 3F capacitors in series? (Which does not equal 6F.)

How much voltage is typically left in the capacitors after leaving it overnight? If they are still half full that makes the recharge significantly faster than if they are nearly empty...

No they were not two 3F caps in series. I used two 6F 2.7V Super Caps, Eaton Bussmann P/N: HV0830-2R7605-R, in series to get the voltage required. With 6F caps there is just above 3V left after a shutdown so I believe it fails overnight, but have never checked the voltage in the morning.

The first two revisions going from 6F (<- Correction 3F not 6F) to 9F (<- Correction 4.5F not 9F) were shown in my post above. In the final revision I added a 3F 5.4V Super Capacitor, Eaton Bussmann P/N: PHV-5R4V305-R, in parallel to the rest. Therefore the progression has been:
Two 6F 2.7V caps in series - Rev 1.
Two 6F 2.7V caps in series, which are in parallel to another two 3F 2.7V caps in series - Rev 2.
Two 6F 2.7V caps in series, which are in parallel to another two 3F 2.7V caps in series, which are in parallel to a 3F 5.4V cap. -Rev 3.

The 3F 5.4V super cap is really nothing more than two 6F 2.7V caps in series.

 

[Nigel]

No they were not two 3F caps in series. I used two 6F 2.7V Super Caps, Eaton Bussmann P/N: HV0830-2R7605-R, in series to get the voltage required. With 6F caps there is just above 3V left after a shutdown so I believe it fails overnight, but have never checked the voltage in the morning.

Thanks,

So actually 3F then. If it drops from 3 to 0 overnight then that something like doubles the recharge time which may make it impractical for a 10 minute drive. Would be interesting to know how much remains in the morning (after 12 hours or whatever).

 

[mr_milo]

Thanks,

So actually 3F then.

Oops! Houston we have a BIG mix-up / mistake! Thanks for catching that....

So lets correct this.

Rev 1 - was only 3F total
Rev 2 - was only 4.5F total
and
Rev 3 is actually 7.5F total

 

[Nigel]

Oops! Houston we have a BIG mix-up / mistake! Thanks for catching that....

So lets correct this...

When we use the calculator it will charge twice as fast now while using a reasonable current

 

[mr_milo]

OK I just got back from lunch and grabbed the camera. It's currently at 3.19V.

I charged it last night until it was at 4.21V, then did the following:
Drove approx. 21 minutes and parked the car for 14 hours.
Drove approx. 10 minutes and parked the car for 5 minutes.
Drove approx. 8 minutes and parked the car for 3 minutes.
Drove approx. 12 minutes and parked the car for 3 1/2 hours.
Drove approx. 8 minutes and parked the car for 35 minutes.
Drove approx. 8 minutes, parked and checked the voltage.

I will do more tests later and post the results.

 

[Tobi@s]

The problem with the modification is this:
The menu settings represent one value each.
Off = 0
10 sec = 1
30 sec = 2
etc etc..
What I'm able to do atm is to change the value set by menu.
Means I could change 10 sec to 0 instead of 1.
But this wont help us here. Because the decission on shutdown delay is made elsewhere.

Something like

switch(mSettingVal) {
case 0: setShutDelay(-1); break;
case 1: setShutDelay(10); break;
case 2: setShutDelay(30); break;
etc..
}
This part of code is compiled and needs to be found in the binary (PRI). I'm not good in ARM Assembly so I have no idea how to find this (except searching for little endian shiorts, integers with value 10, 30, etc)
But that takes much time.
I can't do this now

 

[mr_milo]

Thanks for the info I don't think I could even find the correct menu setting.

 

[Tobi@s]

I already posted the neesed information regarding menu settings a few posts before

 

[mr_milo]

I already posted the neesed information regarding menu settings a few posts before

I know that's what go me going and confused me all at the same time since I couldn't find what I thought I was looking for.

OK had time for some more testing and now feel a firmware mod is the only chance to get this working.
I checked the voltage again and it went from 3.19V to 2.95V in about 3 hours. The surprising thing is that the mini0803 actually powered on at that voltage, then crashed. When it crashed the voltage was 2.73V.
I applied power to the unit, it turned on and started recording / charging. The current draw (without GPS) was 400mA and the capacitors reached 4.20V in 57 seconds.
I then powered off the unit by the power button and removed the power source. When I turn the unit on via the power button, it lasted 22 seconds until it crashed at 2.73v.

Also a correction to my above statement about charge current. I just checked again and the charge current after I turn the unit off is 250mA not the 25mA I stated earlier. Therefore the charge circuit doesn't seem to be an issue. I just have to get my decimal points correct.

 

[mr_milo]

Just an update.....

I have been running my current setup since last Wednesday (about 4-5 days) and the system is working GREAT! I even left the car sitting all weekend long, then checked my files today after my drive to work and everything looks good. The time & date held all weekend and the camera has been correctly powering on and off. Therefore it seems the Mini 0803 can be made to work with Capacitors if you are willing to externally mount at least some of them.

As it sits today my setup is as follows:
Internally I have two 6F 2.7V Super Caps, Eaton Bussmann P/N: HV0830-2R7605-R, in series, which are in parallel to another two 3F 2.7V Super Caps, Eaton Bussmann P/N: HV0820-2R7305-R, in series. These internal capacitors are in parallel to a 3F 5.4V Super Capacitor, Eaton Bussmann P/N: PHV-5R4V305-R, which is mounted externally. I therefore have a total of 7.5Fs and this is working perfectly with the standard iTracker firmware.

I will try and update with pictures of the camera with externally mounted capacitors later today.

 

[Nigel]

Well done, so no change to the charging circuit needed?

Sitting here looking at an 0806, it's 5mm shorter than the 0803 and there is zero space for supercapacitors inside! Not really any space outside either. Thankfully the case seems to run at least 5 degrees C cooler than the 0803 and the inside is considerably better ventilated

 

[mr_milo]

Nope, no change to the charging circuit is needed. I'm not sure why it seems that it would earlier. I must have just mixed up my numbers as I had done with the initial farad calculations.

It's still working great and I'm happy that I finally seem to have a solution to get the Mini0803 working without batteries!

Here are some (bad) pictures of the external battery. I just ran some 24ga wires into the case through the speaker holes and soldered them onto the battery pads with the other wires. The internal capacitors are shown here, currently using Rev. 2 configuration with the external pack as well.

I'm sure something similar could be done with the 0806 using only larger external caps. They would be a little more bulky but I don't think it would be an issue if you could handle the looks.

 

[Nigel]

Not quite how I imagined it, but it shows there is probably room for external only without it looking too bad. You could put a cover over that to make it look part of the camera rather than an add on.

On the 0806, your positioning of the caps would make the top 128GB card slot inaccessible and completely block the top cooling vent, there is room for just one capacitor between the cooling vent and it's CPL filter but no room for a second at the bottom without blocking vents. Maybe they can go on the end, it's not round like the 0803.