Replacing batteries with capacitors in battery based Dashcams

[Kip]

I'm not sure how many folks will actually benefit from this but I've wrote it all out now so here we go!

Spoiler

Note 1: If you simply want to replace a defective capacitor in a capacitor based dashcam, simply buy something at the same voltage rating with equal or higher farad rating than the original. And make sure it will fit the housing!

Note 2: If you just want a quick step by step process and aren't interested in learning the why's and how's, jump to the bottom of post 3.

I've been trying to understand the difference between the way capacitors and batteries store electricity, and how to compare them to each other. I've learned quite a lot myself while writing all this so hopefully it is correct. The more I learned, the bigger this document got. I do have an actual conversion in mind myself, which is what started me off down this path.

Someone who is considering replacing a dead battery with a supercapacitor will want to know how many Farads their new cap needs to replace xxxmAh battery. It's not quite as straightforward as that. A better way of asking things is: what specification capacitor would allow my dashcam to work normally as it would when using a battery?

This thread will attempt to explain just that. In short and in theory, YES, you can replace a battery with a capacitor in a battery based dashcam. However, that is disregarding other things such as capacitor dimensions, available space inside the housing, how far you are prepared to go to mod the case if needed and how often the camera will be used. I'm not an electrical engineer of any sort so only attempt a conversion at your own risk. I'm just an enthusiast trying to solve a problem in front of me and sharing my workings.

Firstly, we need to understand the difference between how batteries and capacitors charge and discharge their energy.

A 12v car battery is considered fully charged at around 12.6 volts and considered flat at around 11.6 volts. In comparison, a 12v capacitor is fully charged at 12 volts and flat at 0 volts. A car battery at 50% charge (let's say 12.0v) might be able to start your car, but, a 12v capacitor at 50% charge would only be at 6.0v. No matter how much current might be available, nothing at 6v is ever going to run a 12v car starter motor well enough to start an engine (although, do check out YouTube for folks starting cars with supercaps). Also, a 12v car battery will easily tolerate being charged at 14v or more, a 12v capacitor could be permanently damaged if subjected to the same 14v.

Just like your car requires a certain voltage from its battery, a dashcam designed with a 3.7v lithium battery will expect to see a certain voltage too. This is between 4.2v (fully charged) and 3.0v (flat) in order to behave normally and most importantly, to save that crucial last file in an emergency. Therefore, we need capacitors that can safely tolerate and keep within these voltages in day to day use.

To tolerate the dashcams charging circuit, we need either a single 5.4v cap, or, two 2.7v caps wired in series to create a 5.4v capacitor bank. Since series wired caps are best when used with a balancing resistor or a balancing board, a single 5.4v cap is easier and might take up less space. Another reason to go parallel is that in series, one bad cap can slowly damage the other and it isn't always obvious. Caps in parallel always see the same voltage and don't have this problem. Remember, if you do buy two 2.7v caps and put them in series, you will halve the farad value of the finished bank. eg. Two 2.7v 10F caps in series will make a 5.4v 5F bank. Although it might first appear that a series wired capacitor bank will store less, when you work out the amount of energy it can store in either series or parallel, it is actually the same figure. An optional explanation of this can be found below.

Spoiler: Stored energy, series vs parallel

Let's look at two fully charged 2.7v 10F capacitors in both series and parallel configurations.

V   = Volts
C   = Coulombs
J   = Joules
F   = Farads
mWh = Milliwatt hours
mAh = Milliamp hours

F * V = C
V * C = J
C / V = F

J * 0.277778 = mWh
J / 3.6      = mWh (alternate method, same result)

C * 0.277778 = mAh
C / 3.6      = mAh (alternate method, same result)

V   * mAh = mWh
mWh / V   = mAh
mWh / mAh = V

Series   | 5.4v | 5F  | 27C | 145.8J | 40.5mWh | 7.5mAh
Parallel | 2.7v | 20F | 54C | 145.8J | 40.5mWh | 15mAh

The figures for joules and mWh are the ones to take note of, both configurations hold the same amount of energy.

Keeping the cap at or under 4.2v is simple, the dashcam will do that bit for us. The cap should be rated above any voltage it will ever see (I'd suggest 5.4v to eliminate any risk of overcharging). Once the cap is at 4.2v, the dashcam will assume the "battery" is full and should not charge the cap beyond 4.2v.

Keeping the cap over 3.0v is the complicated bit. We need the cap to provide enough energy for the cameras normal shutdown procedure to finish when the engine is switched off, plus, hold some leftover charge to keep the clock alive and retain user settings on some cameras. For this, we need to know how much energy our camera is using and for how long after the power is cut.

As an example, I will use an imaginary dashcam that takes 10 whole seconds to shut down and consumes a constant 1 Amp during this time.

1 Amp consumed over 1 hour is called 1 Amp hour
1 hour is 3600 seconds
1Ah (amp hour) is the same as 1000mAh (milliamp hours)

Now we can work out how much power our imaginary dashcam uses over time:

1A over 3600s = 1000mAh
1A over 360s  = 100mAh
1A over 36s   = 10mAh
1A over 24s   = 6.666mAh
1A over 18s   = 5mAh
1A over 12s   = 3.333mAh
1A over 10s   = 2.777mAh
1A over 9s    = 2.5mAh
1A over 8s    = 2.22222mAh
1A over 7s    = 1.94444mAh
1A over 6s    = 1.666mAh
1A over 5s    = 1.388mAh
1A over 4s    = 1.1111mAh
1A over 3s    = 0.833mAh
1A over 2s    = 0.555mAh
1A over 1s    = 0.27777mAh

That's great, but caps are rated in farads. And what the hell are farads anyway?

This is where it starts to get a bit complicated.

One farad is defined as the capacitance across which, when charged with one coulomb, there is a potential difference of 1 volt. A coulomb is 1 amp over a 1 second time period (see table above). This means if we have a 1F capacitor charged to 1v, we have 1C, the battery equivalent of 0.27777mAh and we could power a 1A load for 1 second (assuming 100% efficiency and a load that could actually work all the way down to 0v).

 

[Kip]

There are a few voltages we need to define:
Max circuit voltage: The maximum voltage our dashcam battery operates at, 4.2v
Min circuit voltage: The minimum voltage our dashcam battery operates at, 3.0v*
Useful voltage: The difference between the two figures above, 1.2v. This is the most important one as it tells how much of our capacitors charge we actually have available to use. Unlike a battery, we can't just run our cap flat.
Capacitor max voltage: The maximum voltage of a circuit that a capacitor can be used in. For a battery based dashcam, a 5.4v rated cap (or an equivalent series wired cap bank) will keep us safe.

*Some circuits such as the clock, might continue to run below this voltage but for reliable operation of every circuit, we want no less voltage than a flat lithium battery.

(Hint: Google can convert Coulombs into milliamp hours like the table above)
https://www.google.com/search?q=coulombs+into+mah

The equations we need...
(I used 1 Coulomb = 0.27777mAh for most of my calculations)

mAh / 0.277778 = C | milliamp hours / 0.277778 = Coulombs
mAh * 3.6 = C | milliamp hours * 3.6 = Coulombs (alternate)

C * 0.277778 = mAh | Coulombs * 0.277778 = milliamp hours
C / 3.6 = mAh | Coulombs / 3.6 = milliamp hours (alternate)

F * V = C | Farads * Useful Voltage = Coulombs
C / V = F | Coulombs / Useful Voltage = Farad Requirement

Some examples...

First, we need to know the Coulomb value for our current draw during shutdown:
mAh / 0.27777 = Coulomb | eg: 5mAh / 0.27777 = 18.0005040141124C (let's just call that 18)

Coulombs divided by Useful Voltage gives us our farad requirement:
Coulomb / Volt = Farad | eg. 18C / 1.2v = 15F

This example with the table above, shows how a 15F cap charged to 4.2v could power a constant 1A load for 18 seconds and still have 3.0v left over. Ta da! Our imaginary dashcam appears to be covered nicely.

However, if we only want 1A over 10s, we only need 2.7777mAh.
1 Coulomb = 0.27777mAh
mAh / 0.27777 = Coulomb | eg. 2.7777mAh / 0.27777 = 10C

Coulomb / Volt = Farad Requirement
10C / 1.2v = 8.3333F
10C / 1.0v = 10F
10C / 0.6666v = 15F
10C / 0.5v = 20F

So an 8.5F cap should just about work. This would keep us within the 4.2v - 3.0v range that the dashcam will expect. Obviously, the bigger the cap you can fit, the more reliable your camera will be. Also, 8.5F wouldn't leave much headspace for running the clock and so on. If we put a 15F cap in our imaginary dashcam, on shutdown, the voltage would drop from 4.2v to 3.53333v.

If you know how much current the clock and anything else draws when the camera is off, you can work out how long you can run it for until your "battery" goes "flat". Unfortunately though, the cap will self-discharge much faster than this and that more than anything else will determine how long you can leave it without losing time/date/user settings.

Spoiler: Optional

Taking the example above, 3.53333 - 3.0 = 0.53333v of Useful Voltage.

15F * 0.53333v = 7.99995C
7.99995C * 0.27777 = 2.222mAh OR 7.99995C / 3.6 = 2.222mAh

That's the same consumption as 1A over 8s, and we've already used 1A over 10s.
We've also previously worked out a 15F cap can run 1A over 18s over a voltage drop of 1.2v. It all adds up as expected.

If this is true (sorry, random internet find):

The timekeeping current of the DS1307 RTC (with the square-wave output off) is specified as 500nA maximum. A BR1225 lithium primary cell is rated at 48mAh. Therefore, (0.048 / 500e) - 9 = 96,000 hours, or 4,000 days (almost 11 years).

Then our 15F cap at 3.53333v should take over 4444 hours (or 185 days) to reach 3v running just a real-time clock. As mentioned earlier though, the self discharge rate of any cap will drain it much faster than the clock could.

A solution to this (space and/or ingenuity permitting) might be to install a coin cell non-rechargable lithium battery with a small circuit that keeps the clock alive but isn't used for anything else. Another problem for another thread.

 

[Kip]

Let's say you're not sure how much current your dashcam uses during shutdown, but, you do know how long it takes. You could always guess the current requirement by looking at the original power supply that came with your dashcam and use that number. So if it says something like 'Output DC 5V 1500MA' then assume your camera will drain no more than 1.5A.

10F * 1.2v = 12C

12C is 1A over 12 seconds, so, from that we can work out how long 12C would last for other current requirements.

From a 10 farad cap charged to 4.2v, without dropping below 3.0v, we can get:
0.5A for 24s
1A for 12s
1.5A for 8s
2A for 6s
3A for 4s
--------------------------------------------------
From a 15 farad cap charged to 4.2v, without dropping below 3.0v, we can get:
15F * 1.2v = 18C
0.5A for 36s
1A for 18s
1.5A for 12s
2A for 9s
3A for 6s
--------------------------------------------------
From a 4 farad cap charged to 4.2v, without dropping below 3.0v, we can get:
4F * 1.2v = 4.8C
0.5A for 9-10s
1A for ~5s
1.5A for ~3s
2A for ~2.5s
3A for ~1.5s
--------------------------------------------------
From an 8 farad cap charged to 4.2v, without dropping below 3.0v, we can get:
8F * 1.2v = 9.6C
0.5A for 19-20s
1A for ~9-10s
1.5A for ~6s
2A for ~5s
3A for ~3s
--------------------------------------------------

Another (longer) way of doing things is to see how many mAh you can get out of a cap of a certain F rating, over the voltage range you require.

Useful Voltage * F * 1C = mAh | eg. 1.2v * 20F * 0.27777 = 6.666mAh
Useful Voltage * F / 3.6 = mAh | eg. 1.2v * 20F / 3.6 = 6.666mAh (alternate)

Looking at the table near the top again, we can work out that a 20F cap charged to 4.2v will run a 1A load for approximately 24 seconds (that's 24C) and have 3.0v remaining. Or, 2A for 12 seconds, etc. If your figure isn't in the table, convert mAh to coulombs and you then have your 1A over x seconds. From there, you should be able to work out how long your camera will last based on its current draw.

So there you have it, hopefully there is enough here for most folk to work out what they'd need if converting a battery dashcam to capacitor. Finding something of a suitable size, price and being able to install it is a whole other challenge in itself. Do remember all figures I've used are purely theoretical and real world performance which is affected by wear and tear, efficiency ratings, temperature, etc, will be different. Hopefully this is all correct, if you spot anything out of place, please let me know. Especially let me know if you found it useful or educating in any way, as it took quite some time to figure out myself and write!

Spoiler: QUICK REFERENCE

- Work out or guess dashcam current draw in amps and shutdown time rounded up to the nearest whole second.
- Convert your amp value and time reading into mAh using the table near the top as an easy guide. Some maths may be required.
- Convert mAh into Coulombs:

mAh / 0.277778 = C | milliamp hours / 0.277778 = Coulombs
mAh * 3.6 = C | milliamp hours * 3.6 = Coulombs (alternate)​

- Divide the coulomb value by your Useful Voltage value (most likely 1.2v) to get your absolute minimum farad requirement:

C / V = F | Coulombs / Useful Voltage = Farad Requirement​

 

[Yooshaw]

Wow, great write-up. As it turns out, I'm experimenting with 2x 5.4V 4F capacitors with my Innovv C2. I found some button style ones on Ali (very thin, like a button cell battery), but even as thin as they are, am still working to make them fit.

Also still unsure how long it will hold the date after shutdown. Having fun trying to get it to work. May try and do some math thanks to your explanation.

 

[Kip]

Thanks, glad it's useful to someone. I'm looking at putting a 15F 5.6v cap in a Shadow GT550WS that I received from a DCT member. It won't fit the case looking at the dimensions for it but hopefully shouldn't take too much work.

Also still unsure how long it will hold the date after shutdown.

That will more or less be down to the leakage current of the cap itself than the drain of the clock. If your cap has a datasheet you should be able to find out what the leakage current is, probably in microamps, divide by 1000 to get milliamps. To be honest though, just timing it and experimenting is probably easier as even individual caps within the same product range can have different values. Plus, you probably won't know the voltage the clock stops working at. If you knew those two things though it could be calculated. Good luck with your project.

 

[SawMaster]

As you know, there are different ideal charging parameters for caps and batteries, but as seen with Mobius as well as with your information here, there can be a workable method to go from battery to caps I'm just not sure that a given cam's charging circuitry components can handle the changed load they see as the initial charge rate for caps is higher (less internal resistance when discharged) than with a battery. If the component ratings were marginal for a battery they might toast under the heavier load I think that would only happen with the cheapest cams since a good design always uses higher-rated components than is absolutely necessary.

To the uninitiated I would like to point out that doing the reverse of this- changing from caps to a battery- should not be attempted as there probably will not be a limit on charging designed in, which could cause the battery to overheat and with Lithium batteries that can be dangerous. Caps self-regulate in that regard while batteries do not

I'll be watching the further developments here keenly, this is very interesting indeed. Good job Kip

Phil

 

[Kip]

Good points about the very low resistance of caps. I did wonder if it would have some sort of impact, perhaps a resistor or some kind of special diode somewhere could help with the initial surge in current? I don't know enough about electronics unfortunately, just trying to learn bits and pieces here and there, it's pretty fascinating stuff IMO.

Edit: Come to think of it, if we're talking USB 5v powered cams, wouldn't the 12v to 5v power supply determine the maximum input current to the cap? The camera can't release power to the cap it doesn't yet have.

 

[Dashmellow]

Several years ago I experimented with replacing the battery in a Shadow GT680 dash cam with super-capacitors. I learned from many months of testing two different sets of super-capacitors that it can work to a certain degree but the concept is doomed to failure regardless of the size or value of the capacitors unless the camera has firmware installed that is specifically designed to accommodate super-capacitors.

For weeks and even months the camera would work perfectly, hold the date and time and properly save the last file on shut-down. In fact, the capacitors I used would power the camera for as long as eight to ten seconds after power was withdrawn. Unfortunately, sooner or later I would experience intermittent random corrupted files. In the long run the last files would no longer be properly saved either. The best I can tell is that the super-capacitors eventually become stressed and fall out of spec. A few other intrepid DCT members have tried this too and all have reported eventual failure.

 

[Rajagra]

wouldn't the 12v to 5v power supply determine the maximum input current to the cap?

Yes, but it would look like a short to the PSU. Might not handle it well, could even blow a fuse.

 

[Kip]

Several years ago I experimented with replacing the battery in a Shadow GT680 dash cam with super-capacitors.

Yeah I read that one and possibly some others too. I know someone had a series cap set up that failed due to one or both caps going bad. As you say that might have been the caps being stressed or just bad caps in general. All I know is, I have a battery based cam to hand and want to at least try to convert it to using a cap. If the cap does eventually fail then fair play, a battery will replace it. I do believe it can work, maybe it needs more than just the cap to make it work long term. Only by experimenting will we find out.

Yes, but it would look like a short to the PSU. Might not handle it well, could even blow a fuse.

Didn't think of it like that but yeah that makes sense. I'm thinking something that can add battery-like resistance to the cap when it is flat, but, lower the resistance as the cap charges up. Is there any component(s) that could do such a thing?

 

[Dashmellow]

Yeah I read that one and possibly some others too. I know someone had a series cap set up that failed due to one or both caps going bad. As you say that might have been the caps being stressed or just bad caps in general. All I know is, I have a battery based cam to hand and want to at least try to convert it to using a cap. If the cap does eventually fail then fair play, a battery will replace it. I do believe it can work, maybe it needs more than just the cap to make it work long term. Only by experimenting will we find out.

Yeah, there's no reason not to try it. I had a lot of fun and learned a thing or two. Keep us posted on your results!

P.S. Look into the concept of using a voltage balancing resistor between the two caps. I never quite got around to doing that but it may well help in keeping the caps from becoming stressed and eventually failing.

http://www.kemet.com/Lists/TechnicalArticles/Attachments/125/2013-11%20Cell%20Balancing%20and%20KEMET%20Supercapacitors.pdf

 

[Rajagra]

Inductors act to stabilise current going through them (by limiting rate of change of current), but I don't know if they are a practical solution to this. They are the ugly stepchild of the electronic component family, they are not something I've ever used or seen recommended for the type of projects I've done.

 

[Yooshaw]

Thanks, glad it's useful to someone. I'm looking at putting a 15F 5.6v cap in a Shadow GT550WS that I received from a DCT member. It won't fit the case looking at the dimensions for it but hopefully shouldn't take too much work.


That will more or less be down to the leakage current of the cap itself than the drain of the clock. If your cap has a datasheet you should be able to find out what the leakage current is, probably in microamps, divide by 1000 to get milliamps. To be honest though, just timing it and experimenting is probably easier as even individual caps within the same product range can have different values. Plus, you probably won't know the voltage the clock stops working at. If you knew those two things though it could be calculated. Good luck with your project.

Easier to experiment right now. The fitment is extremely tight, so it's a balancing act of fitment and capacity.

Several years ago I experimented with replacing the battery in a Shadow GT680 dash cam with super-capacitors. I learned from many months of testing two different sets of super-capacitors that it can work to a certain degree but the concept is doomed to failure regardless of the size or value of the capacitors unless the camera has firmware installed that is specifically designed to accommodate super-capacitors.

For weeks and even months the camera would work perfectly, hold the date and time and properly save the last file on shut-down. In fact, the capacitors I used would power the camera for as long as eight to ten seconds after power was withdrawn. Unfortunately, sooner or later I would experience intermittent random corrupted files. In the long run the last files would no longer be properly saved either. The best I can tell is that the super-capacitors eventually become stressed and fall out of spec. A few other intrepid DCT members have tried this too and all have reported eventual failure.

You're probably right - my testing so far with 2x 5.4V 4F capacitors in parallel is yielding inconsistent results There's definitely enough capacity to power down the camera, but it's not saving consistently. Having fun tinkering and learning though.

 

[Dashmellow]

Easier to experiment right now. The fitment is extremely tight, so it's a balancing act of fitment and capacity.



You're probably right - my testing so far with 2x 5.4V 4F capacitors in parallel is yielding inconsistent results There's definitely enough capacity to power down the camera, but it's not saving consistently. Having fun tinkering and learning though.

As for fitting larger capacitors one can try mounting them outside the camera's housing as I did in my DIY project. It allows one to go with as big caps as you want.

As mentioned in that thread two DCT members who are much more knowledgeable than me on these matters recommended the use of balancing resistors but I never did get around to trying it.

In the meantime, more recently, I came across the link I posted above that finally offers a definitve explanation regarding why the super-capacitors will eventually fail.

Here is the quote from the link that explains what happens when super-capacitors with slight variations in current leakage will experience voltage variations between each of the cells.

"For a series of cells that remain on charge for an extended period (that is are being charged by a constant voltage source), cells with higher leakage currents will have a reduced voltage, which in turn will cause the remaining cells to increase in voltage. Over time, this phenomenon will reduce the life of some of the cells and create premature failures".

"One technique to compensate for variations in leakage current is to place a bypass resistor in parallel with each cell, sized to dominate the individual cell leakage current."

If I were to try this project again I would definitely explore using resistors as that seems to offer an explanation for why two sets of different brand super-capacitors failed after a few months in service.

Of course, having the proper firmware in the first place that saves the last file and then triggers immediate shut-down would likely help avoid stressing the caps to that degree.

 

[SawMaster]

There are differences in the charging circuits (and probably the software design too) between the two when a manufacturer designs and offers a cam either way. They would also use close-tolerance caps from one batch to ensure matching. Those can be hard to find at retail (even online) since they're rarely called for in most circuits and since most of that production us taken by manufacturers. Plus they are relatively costly. I suspect that Mobius does it this way to allow their cam to swap between the two. Not the ideal way, but a workable one Or they could be spending time actually testing and matching caps but I doubt it given the prices cap packs for Mobius cost.

Until super-caps came along, caps were not used for primary power storage and they 'leaked down' their stored power comparatively quickly when the input power was cut, but as the device was being turned off that didn't matter. This also allowed caps to better handle being powered-up for long periods. Super-caps lack most of the internal resistance plain caps have which causes that, so they are more quickly worn out by long-term charging. They both tend to self-regulate as their resistance rises as their charge rises. Lithium batteries also do that but they have a lower limit which if exceeded will destroy them; they are not as tolerant of over-charging as caps are. Thus a design meant for batteries will likely under-charge caps, but enough energy might be stored to make the swap workable (and apparently it is).

Caps also wear over time, losing their initial capacity and going out-of-tolerance. The more stressed they are in a circuit the sooner that occurs. And like all electronic components they do not like high heat. When pushed beyond their limits they can explode but as small as the one's we're speaking of are I doubt there is much danger of that. If you look at these electrolytic-type caps, you'll always see an inscribed "X" on the end. That's an intentional 'weak point' which is meant to break open early in a failure, thus keeping the 'explosion' controlled and damage limited to only that one component. Older caps had PCB's in them which made a burst cap potentially dangerous, the newer ones are far less toxic. When you get down to component-level electronics is pretty simple. It's in how they work together that can get quite complex. And I'm still amazed at the miniaturization we have now. If a dashcam circuit of today were built in 1965 it would take a case the size of several shoe-boxes to hold it all. Imagine that sitting behind your windshield while driving

Phil

 

[Dashmellow]

There are differences in the charging circuits (and probably the software design too) between the two when a manufacturer designs and offers a cam either way. They would also use close-tolerance caps from one batch to ensure matching. Those can be hard to find at retail (even online) since they're rarely called for in most circuits and since most of that production us taken by manufacturers. Plus they are relatively costly. I suspect that Mobius does it this way to allow their cam to swap between the two. Not the ideal way, but a workable one Or they could be spending time actually testing and matching caps but I doubt it given the prices cap packs for Mobius cost.

Interestingly, it turned out that the super-capacitors I used in my DIY mod (obtained from a small specialist supplier on eBay) were the same brand (Nano Force) and nearly the exact same ones (slightly different size) than the those used in the Viofo A118C, A119 and the Street Guardian SG9665GC. In my initial quest for caps to do my project the problem was more about finding super-caps of an appropriate size (small) than about finding quality well matched caps. Suppliers like Mouser Electronics sell a vast range of high quality super-caps and can supply data sheets that definitively state the tolerances and all other important specs. One of the issues I ran across was finding suppliers that will sell in small quantities. The cost of these super-capacitors is "relatively" modest but you do pay more for ordering in small quantities.

As for charging it is interesting that Mobius modified the charging circuitry on the V3 PCB to increase the current in order to reduce charging time for the new 820 mAh battery and they also added a voltage cut-off over 5V to prevent the camera from getting fried. If the external voltage exceeds 5V the camera will not start or charge. Charging is limited to about 140 mA for battery longevity. FWIW, super-cap functionality and charging seems the same as before. With the Mobius the primary firmware setting that makes the installation of a super-capacitor work properly is the setting for immediate power-off disconnect. As far as I know there is otherwise no difference in how the camera operates or charges with or without a super-capacitor.

 

[jokiin]

As far as I know there is otherwise no difference in how the camera operates or charges with or without a super-capacitor.

that's true for the Mobius and more a testament to them starting with a well designed charge circuit though rather than it being a blanket rule, a lot of cameras in the market are very poorly designed in this respect and just swapping caps in isn't going to work reliably, firmware is one part of the equation but if the hardware design isn't great it will never work reliably

 

[Kip]

If a supercap has an ESR rating of 10 ohms, at 4.2v being supplied by the camera that's a potential current of 0.42A. That sounds like a reasonable current draw if the original battery was around 420mAh or more (a LiPo charge rate of 1C). To me at least, that says that a camera designed with a larger battery is more likely to tolerate the initial current draw of a cap at 0v, especially if it has a higher ESR rating. I'm probably over simplifying everything... anyone have a better understanding?

I've looked at inductors but yeah, they are ugly, can't get my head around them. I'm leaning more towards an NTC thermistor in series with the cap, I guess I only need to know the charging amps to select the most efficient one. That though, creates another problem of what happens when the camera is off, the thermistor goes cold and that could affect time keeping. I'd need a way of bypassing it when the camera switches off...

PS. Does anyone happen to know the size in mAh of the original battery in a Shadow GT550WS? It is unmarked, unfortunately.

 

[Rajagra]

On reflection, a simple transistor should do the job of limiting current.
In the following, put the supercapacitor where it says load. When power is applied to Vcc, current I(B) will flow through the resistor to ground. A given model of transistor has a set gain value, multiplying the base current by that factor. So I(E)=I(B)*gain. Choose the resistor to determine the maximum charge current.

 

[Kip]

Looks interesting! But, what about when the cap is discharging, won't the transistor then return to off as current now wants to go from the collector side to emitter?

Edit removed 'cause I'm a numpty and don't know how transistors work... yet.