The dark side of supercapacitors
ฝัง
- เผยแพร่เมื่อ 14 มี.ค. 2023
- Measuring the self discharge of some EDLC supercapacitors (500F 2.7V "Samwha" from eBay, 60F 2.8V from Bigcap) and noticing some weird memory effect. The charging curve. Comparing the energy density of a supercapacitor to a Li-Ion battery.
The catculator:
ohmslaw.eu
Please support my channel on Patreon:
/ diodegonewild
Instagram:
/ savage_danyk - วิทยาศาสตร์และเทคโนโลยี
![[LIVE] : ONE ลุมพินี 69 | คู่เอก "กุหลาบดำ vs นาบิล"](http://i.ytimg.com/vi/PNM6U8e_HMo/mqdefault.jpg)
The dark side on the capacitor is the positive side.
What do you mean??
@@0812264capacitors are typically marked light grey on the negative side and the positive side is typically black, hence the positive side is darker… LOL!
Haa, man, good one! :D Love this. :D
You are one of the absolute best when it comes to simplifying complex things and showing your work. Always super enlightening
Not to mention these beautiful, handmade, yet very precise paper graphs (and schematics of course) in the "technology-in-you-pocket" era. :D
Fascinating tests, love the catculator :)
If brand new, Super caps have to be charged and discharged for various cycles until they reach their normal working performance and start
showing a lower self discharge ratio. Also brand is paramount, and Maxwell is yet the best brand I have come across.
Loveland brilliant video. We're supercats really got exciting from me was their use in pulsed photonics. Lasers and other technologies that require crazy pulsed currents at unbelievably fast rates ... One of my favorite examples of which is the fusing energy experiment that isn't your standard containment experiment. But it's like two really fast magnetic guns firing and using a warehouse sized pulse capacitor.
And seeing the pulsed potential of the energy storage at the national ignition facility ....was one of the coolest things I've ever seen.
What dedication !
Thank you for pointing this out.
Some nice experiments done here. Some really dodgy looking caps too! Meowsome catculator :)
Great video, thank you.
Of course it is obvious that the differences you are seeing come down to manufacture, materials, and price point.
So happy to see someone recording data the same as me, pencil n' paper.
;)
I have seen your stuff on the net (schematics, for example) probably couple decades ago, and you have done a lot of great work.
I hope some day to meet you in person and buy you coffee.
Best wishes from Bosnia.
If you ever come this way, let me know, I will be glad to show, and drive you around.
Super interesting! Thanks for the deep dive 🤔👍
Maybe there's another interesting type to test. The mixture of Supercaps and Lithum Cells "Lithium Ion Capacitor" aka LIC.
Physical quantities named after scientists _(volt, ampere, ohm, watt, hertz, farad, henry, coulomb, kelvin, newton, joule, siemens, weber, tesla, pascal, becquerel, gray),_ unlike their abbreviations, are written in *lowercase letters* to distinguish them from the surnames. (Note that _degrees Celsius_ is different.) You got the first four right in the catculator; could you correct the rest? This will fix the awkward _camelCasing_ of _microFarads_ as well.
Thanks for having fixed this!
now its in lowercase ;)
Really interesting video. I was thinking about using supercaps as clock chip batteries instead of very small nicad ones that are used in some Icom radios, that lose capacity over time and become useless in just a couple of years... but if supercaps have such a self discharge (as the big one you tested) then it's better to just install a 2032 non rechargeable battery and replace it in 5 years, otherwise I'd probably lose the clock if I keep the radio disconnected from power for more than maybe 2 days.
You will find supercaps utilized similarly in a lot of designs going back quite a while. I think the first example I've seen of them has been in test equipment and metrology type stuff. But also just really fancy servo drives and things like that from the mid-to-late 2000's I'm guessing
They almost look like a little circuit board mounted coin cell battery but sometimes, they can take other forms also. Even surface mount.
I Remember the first time I found a little 5-volt one that looked like the size of a penny. And had something like five full farades of storage!!!!This was a decade or so ago before I knew much about renewables and battery technologies. I was just absolutely shocked and it's one of the first things that got me interested in that type of stuff
@@hullinstruments smd super caps? thats interesting. I wonder how big the footprint of a surface mount device of farads would be
Curious stuff indeed, I always want to try to use them in replacement for lead acid batteries on computer UPS'. Surely the runtime would be much lower but, maybe we can make a very low maintenance long-term usage UPS, since VRLA batteries doesn't last more than 2-4 years.
They are offered by Siemens for industrial uses, look at capacitor module: 6EP1933-2NC01 and UPS module 6EP1933-2EC41
If your UPS is eating VRLAs in only four years, then you either have horrible power or your UPS is cooking its batteries by over-charging the heck out of them. My APC UPSes' batteries usually last about six years.
@@teardowndan5364 It's funny when you say that "only 4 years" since no one lasts more than that. Usually maximum 3 years for replacement batteries, and a maximum of 4 to 5 for the factory OEM batteries.
@@WagTsX All of my APC UPSes have lasted 6+ years on their original batteries and my old BX1000 lasted another ~6 years on its replacement batteries. Haven't had to replace batteries on my six years old BR1300G yet.
@@teardowndan5364 Ambient temperature is a big factor in the lifetime of these batteries, in the tropics without AC it's common for these things to not last very long in standby use.
8:43 😆Great video as always Daniel! THX!☺
Very interesting!
It seems like most electrolyte capacitors behave in a similar way according to my experience. First time I observed it, was when I tested a big 450 V, 3.3 mF capacitor that I plan to use for later projects - I charged it up with the output from a mains rectifier (connected to a variac, so I could crank up the voltage slowly and avoid a huge current spike). When disconnecting it, I taught it was faulty at first because the self discharging seemed so fast (it's voltage dropped more than 1 volt per second right when it was disconnected). But after having charged it up again a few times, the self discharge got slower and slower.
When I was done, I discharged it thru a resistor (to avoid getting a zap from it later if I forget it's charged or have it accidently short thru something with a bang similar to a gun shot, haha). Later when measuring the voltage on the terminals, it was 10 V - I was like "what - I discharged it all the way down to 0". Then I tested charge and discharge again and after discharge, the voltage rose slowly again for a few minutes before settling at like 12 V. Then I tested some lower voltage capacitors (typical ones used in electronics - like 16 V, 470 uF) and saw they in fact behave in the same way.
The only thing is that all voltage changes is proportional. For a low voltage capacitor it doesn't feel like big deal if it has 0.4 V left in it after use - but one that has 12 V in it when supposed to be empty really catch the attention. Proportionally if the capacitor have 12 V left in after being previously charged to 325 V and discharged, it's the same as if one that was charged to just 12 V, have 0.4 V left in it). Although, the 60 F ones here where more like the typical behaviour I have also observed. The 500 F with it's more continuous behaviour, was more unusual.
Please, take in account the fact multimeters have a resistor (may be around 1mega ohm) to check the voltage and this is load on the capacitor when you test it, so small capacitor will discharge very fast when the leads are connected. By small capacitor i mean in uF range not the voltage, so 3300uF with 1mega ohm load will drop pretty fast from 400volts.
@@yankozlatanov Yes, but in my multimeters (as in most) it's a resistor of 10 mega ohms. So if the capacitor is charged to for example 330 V, that means a measuring current of 330 / 10 000 000 = 0.000033 A (or 33 uA).
Which means for a 3300 uF capacitor, the voltage if there was no leakage current would only drop by 0.000033 / 0.0033 = 0.01 volts per second right when the power supply is disconnected. Which is very low, compared to the about 1 volt per second I got. And I know the capacitor isn't faulty, because at much higher charge/discharge rates, it gets very close to it's rated value.
Same thing for the smaller capacitors. For example at 12 V, the measuring current is only 1.2 uA! So if 470 uF is discharged by that - it means the voltage will fall by only 0.0000012 / 0.00047 = 0.00255.. volts per second. It falls way faster than that. Also, the voltage rise after being discharged - that happened despite the 10 megaohm multimeter still being connected (which means the load imposed by the multimeter is pretty much neglectable).
The self discharge can be checked in another way, just leave the cap for some time without any load and then measure it and do the math knowing the initial voltage, the measured voltage and the time passed. In that way the result will be most accurate 👌.
@@yankozlatanov Did you even read my last comment with calculations?
And the whole idea was to see the momentary self discharge. Not to just do a simple measurement by charge and leave for some time. If the voltage drops over 100 times faster than than it should, with a 10 Mohm load, the load isn't the problem.
Also, higher load means less of the strange behaviour would be observed, not more. At higher discharge load of 100+ mA, they behave just like expected for the given capacitance. The strange behaviour only show up at no or very low loads, so there is no way the load be the multimeter is causing this.
Also when testing with other capacitors, that's not electrolytes, they doesn't behave at all the same way.
For example a 100 uF film capacitor I tested right now when writing this and charged up to 30 V (maximum of a bench power supply that I had at hand) will have a measurement current of 30 / 10 000 000 = 0.000003 A - that means the voltage would drop by 0.000003 / 0.0001 = 0.03 V per second. It dropped from 30 V to 29 in 34 seconds (which is pretty much spot on). And that's regardless of it's previous charge state. It also stays at zero once discharged.
Electrolytic capacitors store a portion of the charge in the dielectric, not just the metal plates. The metal plates discharge almost instantly BUT the electrons that have been pushed into the dielectric slowly leak out of the dielectric onto the plates and that is why some voltage re-appears despite being fully discharged.
This is the same phenomenon, more or less, that makes flash memory possible.
I’m guessing all capacitor curves will always be exponential because charge is directly proportional to the surface area of the plates.
I've always wondered if a simple medium-wave radio receiver might charge a supercap from the detuned frequencies and use that power for a simple audio amplifier for the tuned frequency.
Nice. I've got a couple of 3000F ones I should try that test with.
Nice & the calculator website is super fast & without all the useless clutter , this is how to build a website ( I know it looks like 20 years in the past , but it works far better than all the modern junk)
You should try this with premium caps like Maxwell or Skeleton
Hello Diode: What about your 1C15 digital oscilloscope ? Is recommendable today 2023 ? Is there many bugs in it ?
The china 500F Supercaps are poorly made - thats why the selfdischarge is always much higher than with the 60 F cap. I bought 6 of the 500F caps, connected them in series and use them as starter battery for my little motor bike. It works fine, when the engine was running before and is still a bit hot - but when its cold in winter the capacity is far too low to start the bike. The caps are empty before the motor can start. But its enough for 2-3 trials. I also made curves like you and found there are quite big differences between each caps: in capacity, in selfdischarge and also charging behavior. The quality spreads a lot! I hope they can improve the capacity and stability a lot more in the future. They are good for very fast charge and discharge and can be used as energy sources for local busses and charge at each stop quickly. So its a partial replacement of li-ion batteries in special conditions. Cycle life is about 1 Mio. compared to 1000 for li-ion. But of course energy density is still very low.
It's been some time since I watched a DiodeGoneWild video, and while his accent remained the same, his prosody now sounds more like a Greek Orthodox priest chanting.
WOW that's so weird seeing a capacitor behave this way😮 like you said limited use, replacing a rally car battery maybe for weight saving etc cases where you can charge them fire up the car and keep them charged while its running, if the driver stalls the car at any point the energy would be there for a restart of the engine 🤷♂️
Before you have tested the double layer capacitor, did you make a formation of this capacitor?
You load it to 2.7V (in 0.2V steps, hold it for some hours) for a while and this process is doing some type of formation inside of the capacitor, then the parallel resistance will be reduced and much lower, but this needs some time. It need 50-150 cycles till the capacitor is ready conditioned.
After this you can test the capacitors again. It is called "supercapacitors conditioning".
Also have a look at Tadiran Pulse Plus HLC-1550A. They rate their capacity in Ampere seconds!
I want to see disassembly of charged supercapacitor
Since it's a higher capacity, maybe the 500F one needs to be charged for longer to get the same self-discharge curve as the others?
I remeber specifying a supercap for a real time clock. The clock was a nice IC with 500nA draw. The best supercap self discharge was 1uA :(
If your supercap's ESR is so high that it has a visibly RC-like charging curve despite being charged under constant current, I doubt it'll be any good for high pulse current use. These cheap supercaps are more suitable as an alternative for battery backup in clocks, thermostats and other similar low-power devices. For high pulse current applications, you need supercaps designed for high power which use some form of parallel plates construction for greatly reduced ESR.
if we connect in series to increase voltage capability then capacity decresses. 😮
that is exactly the conclusion I came to as well, my first UPS was on these caps - my next is on 18650 ;-)
Bro suoer battery charger 30amp detail. Plz
Gorgeous cat you have here there.
Nice
Is it the cap value that makes the discharge rate different, or is it the quality of manufacture, ie, maybe poor manufacture has more contaminates that discharge it faster.
It's not the capacitor value. If you scale it up to double size, it doubles the leakage current, but also the capacitance, so it cancels out. It's probably the quality and technology.
Maybe that 500F cap off eBay is just very poor quality?
Hi Daniel
Can we use for LM393 instead of TLV3201?
Thanks.
Probably yes, but it might be a bit too slow for some situations...
Take note a 1F 5.5V supercap can take over 100hours to fully charge.
I have a 1F 5.5V it's been connected to 5.4 volts for 36 hours its leakage current has fallen to 16 micro amps and is still falling datasheet says it should drop to 6 microamps.
Down to 7.6 micro amps
To me it seems 2 things about the 500F one: First it is likely a high peak current device, so the internal metalization will be designed to exhibit very low resistance, so the "sections" are not that visible. Second it is expected to be used in a series connected battery for higher voltage operation. And there you need something to ensure the charging voltage gets distributed evenly, so none of the cells get overcharged. In higher current applications the balancer is then required to operate at higher balancing currents as well, so it would need to be some active electronic. And any active electronic need some supply current for the control part and usually this supply current starts to appear when the voltage exceeds threshold voltage of an active device, which uses to be around 0.7V (both bipolar, as well as CMOS).
So the 500F seems to have its balancer section integrated in the case, so at the 0.8V (hence it seems to slow down to near stop when discharging down to the 0.8V) it starts to consume some bias current, which then appears as a capacitor self discharge. On a component intended to deliver power for just a few minutes per charge this current is negligible, but the no need for external balancing is very practical (as you just connect all cells plain in series).
"It's not a fault, it's a feature!" But no, these are simply trash picked capacitors.
@@SinsBird It is just what "other than a fault" came to my mind. Definite answer would come from the datasheet: If it were a true intentional feature, it would be described in full detail, including parameters. And multiple units even from different batches or age would exhibit those currents as very close to each other.
And I agree, more likely it is parasitic behavior, not constant over production so the extra load for the real balancer (if varying among units).
@@annaplojharova1400 I think the datasheet said 2mA max self discharge current, not sure, this was 2 weeks ago. I've got 6 caps: Out of 6 caps 2 have quite reasonable self discharge rate, in 12 hours from 2.58V to ~2.3V, 2 caps self discharged to around 1V, 1 cap to 1.5V and the last cap to 1.8V. So 2 caps seem to be just barely out of spec, self discharge rate comes out to be 3mA I think while others are much worse.
These 60F Caps. seem to have a higher self-discharge with just above a “zener”-like voltage, resulting in lower insulation resistance as voltage grows (if it was electrolytic, I would say it was an electrochemical phenomenon). Imagine for example discharging a large capacitor using a Green LED (Vf=2V) using a resistor 1K in series, for instance. The voltage decay will be exponential, but adding an asymptotic voltage as Vf.
I believe since you charged it slowly the 500f capacitor had more time to saturate resulting in less initial drop off
Oh I see you tested it
That's really interesting. So supercaps are only appropriate if you super high discharge currents.
Yeah, and apparently very long charging time as well, which I didnt expect. I always though fast charging speed was the main benefit, but apparently that comes with a big drop in capacity. (Note that the Energy content in capacitors is proportional to voltage squared, so a few hundred mV is a big deal)
Supercaps are capacitor-battery hybrids.
when it get to a voltage its not necessarily fully charged to that voltage. It will continue to draw current to charge up.
Next time charge one with a current meter instead of a voltmeter and you will see what I mean.
It looks like it almost gets to a voltage but keeps drawing current, which is different than high self discharge. The difference is shown when discharging.
Measuring the capacity of a supercap is easy. Measure the time it takes to charge a cap from with 1A constant for a difference of 1V. The time in second is then the capacity in Farad. A 500F cap should be loaded in 500 sec. from 1V till 2V with 1A constant.
Cool
What about the other super caps you bought?
Also why not semi automate the tests? you can buffer with an op amp the cap voltage and then use an MCU to sample at a given interval. Also not sure if your energy/ volume math is the best since a Li ion cell is usually used between 4.2V and 3V + maybe you can also compare with NiMn and NiCaMn cells too.
The 10F ones might be tested in another video ;). I might automate the test using my DIY digital oscilloscope. Because all factory ones seem to go no slower than 10s/div. The 3.7V voltage is the average for Li-ion, so it's used for the energy calaculation.
@@DiodeGoneWild i think what the commenter meant, was to use a microcontroller's ADC inputs to automatically save voltages over time ( or currents if you use a resistance ) to a file or something using the serial connection, then plot the data using software like excel or something. ( you can even use an arduino or an esp, or if you want another mcu, you can use an ADC module to equipe it with an external analog input )
What is the application for these Super caps in electronics?
I've seen people make starter batteries for cars, but without too much success.
Realistically there is not much use for them currently. Lithium batteries are better in most aspects and most situations.
@@Basement-Science Regenerative braking?
@@msansjr For that you need high charge current capability for the most part - which this video shows, at least SOME super cap types are not great at. (needing the voltage to be applied for hours to get the full capacity out of it)
A large enough Lithium battery in an electric car can handle the generated Power directly, because it is just a few C worth of charge current.
But yes, that is one of the best applications of supercaps *_in THEORY._* Assuming you use the right type of supercap, which is usually hard to find out what model that would be.
Why aren't they successful?
I made a starter battery out of 6S 3000F capacitors and then doubled it to 6S2P and it works great.
You need to use low contact resistance connections like mechanical lugs, crimps, or welded joints when constructing your batteries so you can actually get those high currents.
People who make their own starter batteries have to also consider how much energy it takes to start and how long they can operate the 12V systems while the alternator is off.
If they leave the car for days the self-discharge + the parasitic drain from the car computer and car alarm will drain the caps down past the usable voltage.
My setup uses a solar cell under the sunroof to offset the self discharge while I dont use the car.
Depending on engine size, you could start the car as low as 7V or require a higher 10-11V if you have bad connections
Discharging figure isnt x*e^(-(1/RC))?
Leaving the 60F caps on for many hours may have helped them reform a bit to their original condition after sitting discharged following manufacture, which would reduce their self-discharge. If the big 500F was used before-hand, it may already have had a chance to reform, meaning its already plateaued.
....or it may simply mean that the 60F capacitors are of a much better manufacturing quality than the 500F....
@@blg53 The 60F are likely of better quality as you say. Its the relative change that I was commenting on. I was commenting on the change in the 60F, which could be due to them having a chance to reform. It could also be due to a memory effect as well, which if so wouldn't make them as desirable since they would need to have a long charge time before each use. In some applications it would be nearly as bad as the 500F with its constant self-discharge.
supercaps gone wild
Probably one of the reasons why the S Pen in Samsung phones disconnects so quickly even if not using the button...
Hmm surprisingly high self discharge. Haven't done the calculations but that must represent a huge waste of power if combining with batteries or solar.
If the 60F cap behaved better after soaking at the charge voltage for twelve hours maybe the 500F would do better after charging for around 100 hours.
doubt it, since there was no difference between 0 hours and 12 hours. The effect should at least be noticeable after that time. It's probably a completely different type of supercap from the small ones tested here.
with my 1F supercap the datasheet says it will take around 100hours to fully charge.
sadly my my diode has popped, this resulted due to a pair of very large capacitors discharging suddenly into a load
i have the same flamaster
Just looks like they are poor batteries. it would be good to see how well they respond as high current sources when used in parallel with car batteries, how well they can smooth out high current short time demands, like high power RF transceivers.
Excellent video,
Could you make a video for telecommunications devices for us?
You can show us a display of frequencies of devices, their distance efficiency.
And how can we set up an intelligent device that could take speaking orders from me, and can open PC , switch on the lights , play a beautiful music, using the Bluetooth?
Technology is so beautiful, I wish we could fully benefit from it
Hello , my English is not good ,. So please don't upset for my mistakes.
Maybe 12 h in constant current is not good enough to keep voltage on 500f capacitor.
I make silly calculations ...
(500/60)*12 => 8,33 *12 => 100 hours of constant current.
Maybe the diagram of voltage change if you leave the 500f capacitor to constant current for 100 hours.
You should check out some Lithium ion Capacitors (LICs)
There are 2 places that SuperCaps RULE. Car Audio & Starting your car. You need 3,000 Amps REAL QUICK? This is how you do it. Even if its -30 outside. Dump the current and its heats up to ~67% output in 10 seconds for the Maxwell style caps (Accounting for Age, Cold, and state of charge) .
They're still too darn expensive for what they are in my opinion.
Také jsem si vyzkoušel 500F/2,7V. Tabulku, vybíjecí char. ale netvořil.
Are you standing inside an active volcano?
Good video .... ( so , now we know why they are used for car audio BASS boosters ! ... and NOT long term energy storage , thank you ! ) .................. DAVE™ ...................
Fuzzy ball cat a+
i think a joulethief would run several days from that 500F Supercap.
Considering the miserable efficiency of them, making a series battery from caps and running led from current regulator may be better idea.
Well, regarding calculations: energy density is one thing, but how about comparing power density? Real supercapacitor, Maxwell BCAP0600 600F 2.8V, can provide 280A of peak current so some little welding is possible. This cheap one with those small termimals would melt down - it is not supercapacitor, just capacitor with high capacity.
Supercaps are not called supercaps because of their current capability. It is usually lower than in low ESR caps for SMPS.
supercapacitors are fundamentally constructed differently. Their real designation is Electric Double-Layer Capacitor and there are hybrids that exhibit more chemical battery characteristics called pseudocapacitors or hybrid supercapcitors because they store charges electrostatically as well as undergo chemical reductions. Electrolytic Capacitors usually have a solid dielectric barrier in addition to the separator between the electrodes, but a supercapacitor will just have the porous separator wetted with an electrolyte like salt water which will have a max voltage of 1.23V. The voltage rating reflects the double layer barrier between the electrolyte and the electrode, where parallel capacitors are made on the atomic level. The electrons and the ions from the electrolyte are able to get extremely close to each other at this double layer boundary, which is why you get massive capacitance ratings.
The downside is that the electrolyte will break down at low voltages. Water becomes HHO through hydrolysis at 1.23V and you decompose your electrolyte. This is why capacitors don't often use electrolytes with water, and opt for ionic solutions that increase their decomposition voltage to a higher rating like 2.3-2.7V.
I seriously suspect that typical, cheap eBay Chinese components are not PRIME components, but are instead manufacturing rejects.
Nicr
Bring out your atmega88 based tv to o-scope
I have bought several 500F caps from ebay and only few of them have this self discharge problems. Probably they are defective or production rejects because they self discharge to 0volts after few hours. The rest of them after charged to 2.5volts self discharged to around 2.3 - 2.4 after few weeks and probably its because i have DIY balancing board to them keepingthe max voltage @ 2.5volts per cap. So not all supper capacitors have this problem from what i have experienced.
Do know that the BCAP0350 i got self discharge like you typed later, they can retain above 2-2.2V for weeks.
Also didn't take care in charging them, power supply at max current so 3-5A charge for each of them.
This video is actually a daily reminder to feed your cat. The rest of the video is just to draw your attention.
Mr Louis Rossmann would probably pay you a million Kronas for your cat as his 3 look rather mainstream......
It's only the case with trash picked supercaps.
Why you speak danm singing?
In response to your kitty... Any OCD's watching this channel know what they are getting themselves into. LOL.
10:33 I don't think this is the limit of 18650 cells since I've seen someone modifying his iPhone to be operated using five Tesla 18650 sized cells 5000 mAh each
Proof: th-cam.com/video/HEqQ9S04ki8/w-d-xo.html
(I'm not an authority at electronics since I'm a 1st grade high schooler)
These seem to be slightly larger 21700, not 18650. But anyway, the 18650 being 3400mAh at most might be an obsolete information.
Co tam říkal?
Those are probably junk capacitors from aliexpress or something. I have 3000F supercaps that lose about 100mV in 24 hours . Your best ones are losing nearly an entire volt. Wow!
The 60F ones later have formed. After keeping them charged for couple weeks, they lose about 20mV in 24h.
@@DiodeGoneWild that's much better! One of my 3000F caps has been sitting at 1.001v for 48 hours now. Hasn't dropped a single mV yet. Seems like it only loses voltage when close to fully charged.
So you sleep about 8Hours - Healthy :-)
I really love your videos but I think you should of mentioned that "normal" capacitors charge up exponentially (not linear). Time to reach 63% charge (Tau) is tau = RC, and 3tau is about 95% charge, 5tau is 99%. You can never reach 100% charge due to the fact it is exponential which is why we count time constants. If we look at most of these graphs we can see that characteristic charge curve profoundly at the start where the time to charge is a multiple of the last datapoint. I am not sure how relevent that is to EDLC's. It certainly doesnt apply to hybrid layer supercaps.
There shouldn't be a difference in charge characteristics between capacitors. The RC time constant is based on constant voltage charging, not constant current. Charging at a constant current, the voltage will rise linearly in any kind of capacitor. This is because the physical theory is based on parallel plates accumulating electron charges. Because you are charging at a constant current, there is a fixed amount of coulombs entering the capacitor per second. 1 Amp is defined as 1 coulomb per second, and 1 coulomb is 6x10^18 electrons. Unlike charging at constant voltage, where current is limited based on the resistance and voltage difference between the cap and source, constant current charging is limited based on the number of coulombs flowing per second.
You would see the same thing in a ceramic capacitor, polymer capacitor, anything based on parallel plates.
did your accent get worse? damn i cannot understand your rolling of the tongue
Oh please slow down your talking. At times it sounds like you're saying one long word.
ifitrytomeasurethecapacitorrrrrrrrrrrrrrandeamineeeeeethediscargecurve.
Please just slow down your speech. And learn to not stretch out your words. Like "capacitorrrrrrrr". It's cap-ac-si-tor.
Good videos but I cannot understand your english.