2.7V is the rated voltage. The maximum voltage is typically stated in the datasheets to be 8% higher, so even at the worst case scenario we're just around the maximum voltage and not exceeding it. This module is meant to be used in automotive applications, where the voltage rarely exceeds 14.5V, and for the protection circuits to activate the capacitors would have to go out of whack big time (which with time they probably will..). If you are trying to cram every fraction of a Joule into this in a lab setting - perhaps 2.5 or 2.6V rated chips would be better, so as to keep the voltage below the RATED value as opposed to MAXIMUM value. It's a bit like expecting to see a MOV rated at 245V in a mains overvoltage protection, which would be a nonsense.
I am sure the max voltage is way above nominal 2.7V of these supercaps. If so then 108% (takin max of nominal and hysteresis) would be 2.91V - still probably below the voltage of the capacitor to blow. On the other hand if this sensor detects exactly 2.7V to start charging it, then the capacitor would be always between 2.7-2.91V - not bad, just 10% of the U nominal
Bear in mind that the the data sheet for a 500F one was looking at today, gives the upper voltage limit of 2.85 volts. Seems they can tolerate quite a bit more that 2.7 Volts
Nice video! But I do think it would work rather good if the 2,5V device was used. Unfortunately that version is a bit hard to get. Farnell only has the 2,7V and 4,5V versions... Digikey has them for 64 (Yankee) cents. Mouser about the same. You could try a sample request 0:-)
Having investigated further online, I am fast coming to the conclusion that the capacitors in your device may be either fake or falsely labeled. Elna manufacture no super capacitors in their DZ, DZH or DZN ranges with both a maximum voltage of 2.7v and a capacity of 120F. In fact, anything they make of that sort of capacity is significantly larger than the ones you have - Dynacap DZ range 2.5v 100F is 25mm × 50mm (Dynacap range 2.7v only goes up to 33F) and they do not make any capacitors in their Dynacap range in 1.2F,12F or 120F.
I also researched this subsequent to your comment, nearly 9 months after your comment and ELNA do not list these capacitors on their website. From my initial observations the one thing that does seem genuine is the capacity....
Hi Julian Thank you for the video. I have one of these CAPACITOR modules and going to use it with a crank generator. Some suggested to keep the voltage down to around 14.5V so I suppose I will add in a regulator to ensure it does not go any higher.
How many w/h (energy/per hour)can this bank have? Can it be modded to replace the supercapacitors with for example 200F Supercaps? Can more pe used in series to act as an e-bike battery? Or not worth as a battery, but like a NOS boost like cars have? Can it be done so that a bank like this is charged when breaking an e-bike?
One other issue I see with this circuit is that when the capacitors are discharged (via the soldered on bulb), some will be reverse biased similarly to fully running down a pack of AA batteries where one battery runs down before the rest.
In my opinion, they are not a "protection" feature, they are a "balance" feature, you should not attempt to charge the entire module above the rated voltage.
In the data sheet's OPERATIONAL EXPLANATION, hysteresis operation is made a bit confusing in (1) and (5). In (1) where it is stated: when Vin ===IS HIGHER=== than Vdf, then Vout=Vin. and in (5) where it is stated: WHEN Vin ===RISES ABOVE=== Vdr, then Vout=Vin.
Thanks for the update. I don't take mine to anything above 14.5 volts total across the capacitor bank - hopefully that's well within the specs of each individual capacitor.
You're probably not going over the 2.7V upper limit, but I noticed when the module is fully discharged (courtesy of the 1W bulb) some caps were showing a small positive voltage and the others a small negative voltage. There's no easy way to 'bottom balance' the caps in this module.
Could you monitor the maximum voltages of the capacitors when you are charging and discharging at higher currents within the rated voltage of the capacitor bank to see how well they stay balanced? One datasheet for 2.7v supercaps caps showed an absolute maximum of 2.85v for 1 second non repeating, it would be good to see how long they lasted in an actual application.
He has 6x 120F capacitors in there with a 2.7V voltage rating. The energy stored in each of those 6 capacitors is 1/2 CV^2 At 2.7V that will give you 437Joules and at 0.7V the remaining energy will be 29 Joules so usable energy is 437 - 29 = 408 Joules (same thing as W/second) or 408 Joules/3600 seconds = 0.113333Wh Since he has 6 of those in that module total module capacity is 0.11333Wh x 6 = 0.68Wh He used a 55W rate bulb but that is probably the rating at 13.8 or 14V and he had a bit more in the beginning but that dropped quickly so it maybe started at around 65W in the firs seconds then dropped down to almost nothing in those 40 or so seconds. Since resistance of that bulb is dependent on temperature it will be a bit hard to calculate the energy delivered in this video. If it will have been a 55W constant power load then that 0.68Wh should have lasted for (3600s * 0.68Wh)/55W =44.5 seconds. Since that was not the case in this video those capacitors have probably a much lower capacity than advertised so instead of 120F they are more like 60 or 70F.
Yes they are not that useful in most applications. There are some particular applications where they can work well and even be cost effective compared to a battery. They where used in some large EV in the past to recover the energy from breaking and release that immediately after for vehicle acceleration so there many many cycles/day. As an analogy you can see the super-capacitors as the fast cache memory in computers vs RAM maybe represented by LiFePO4 (more like SRAM) then Flash represented by other high energy density Lithium cells like LiCoO2 and not sure if I want to compare HDD with Lead Acid :) it is worse than that. Super cap can be recharged in a few seconds, LiFePO4 in a few minutes, LiCoO2 in a few hours ....
you could use supercaps a s batteries, but you would probably need to regulate/boost the voltage to get the same performance out of them - as with regular batteries they are relatively delicate and need to be protected. The fact that they don't degrade as much over charging cycles and can charge very quickly also potentially makes them more usefull.
I did read them :P, its just my thoughts on how you could turns these caps into the equivalent of a battery, using them directly wouldn't work well in a lot of applications as the voltage drops down all the way to zero whereas battery remain relatively constant. Not reading the comments is only true for my troll comments :>
Yes I would rather 2.5v version with lower capacity. Rather than risking popping a capacitor. So will be looking to replace mine. What is the part number for the 2.5v IC? .
I am wondering if the chip is actually intended for low voltage load disconnection of Lithium cells... I have the exact same capacitor module, interesting stuff from the datasheet for the chip. Looking at the data for the Maxell supercapacitors (presumably using the same chemistry) it seems they are rated for 2.85v, so this may not be a disaster at all.
I agree that this is the wrong protection voltage, but I believe you misread the data sheet. It's actual voltage design is 2.8V, not the 2.7V that you said it was.
Julian, My 2.5 volt protect c uses lm314 and uses no current under 2.5v and triggers a darlington overvoltage dump. Maybe we should use zeners over other ciruitsts?
You could bodge the ground of the of the voltage detector with two resistors raising the ground. The device is low current so resistance could be a 10s of K so little loss. Better to use a good reference and 6 comparators and just use the FETs and load resistors. Could the FET be a darlington transistor as even low threshold FETS have perhaps 0.5 to 4V, so can't ensure would always work.
replace voltage detector with a TLV3012 comparator that has 1.24v internal reference, choose the cmos output and add potential divider or potentiometer to set wanted voltage, works down to 1.8v.
I have a question to you Julian. You are using the B3606 and B3603 (the step down modules). Im making my own solar charger to charge a power bank and I wanted to use one of these modules in it, my problem is that I dont know how they really work, If I set the limit to 5V, will they stop outputting when solar power voltge drops ? Or do I need something else at the end of my solar panels ?
i) The cap will be 2.7v over full temperature range. ii) the voltage that the chip triggers may well be different than the value on the DVM because of the current path in the PCB. Probably nothing to worry about, that said be wary of eBay crap...
Normally you would want at-least 10% headroom, 20% is even better. Using 2.4 volt Zener diodes in reverse bias is cheap, but effective over voltage protection.
Julian thx for all great videos and many inspirations !!!!!! But i have question , can I use suprecapacitors as battery for portable solar generator (2,7v 500F X 6 ) = 16,2v 82F . supercaps are same brand as those in video . Does it worth to buy them ???
Love your meticulous investigations and demonstrations. Did you try find the right IC (one turning on its output when voltage is 2.7) ? if so you could try replace them on the circuit board and make a video for us :)
***** that's not how it works, usually when designing electronics, you'd pick a capacitor rated at twice or even 3 times the max. voltage you're designing for. That's why if you crack open an led bulb the capacitor is rated around the 400V mark as most mains supplies are 100-240V
Super-capacitors much like other electrolytics, are typically designed and built to allow up to 20% over-voltage with only minor reduction in service life. 2.862 volts is only 6% out. Remember that these are over-voltage protection circuits and not true balance circuits. You should still be using this module with a proper balance circuit. The balance circuit on my super capacitor boost pack uses Zeners, resistors and LEDs IIRC to keep the capacitors balanced. The intended purpose of these super cap packs is automotive use so as long as a decent balance circuit is used. These protection circuits should never even need to kick in. They could have easily dropped the voltage by .15 volts when they programmed the chips but then they would also have to say the max voltage of the pack should really only be around 15.5 which would still be fine for automotive use but would allow for a bit more safety margin.
well OK... that was interesting..So, now what? into the trash bin or will you try to desolder that chip and replace it with something else? Kinda hanging here wondering after that walk through..
Given it's just a circuit board I'm sure it's limited but what Amperage can you charge those at max? Wonder soldering a wire to bridge between capacitors would help. Nice update I can't wait to see you play with this thing a bit more.
Seriously, what's your problem with this Ryobi thing? In my view the main advantages of expensive brand multimeters are better security in case of high voltage/high current and expected better long-term stability. I prefer the brand stuff myself, but often I'm surprised how far people get here on youtube with a multimeter for less than 10 € if they know what they are doing.
The Ryobi is a good DMM, for what he is doing. High price is a decisively bad attribute, making all meters you listed BAD meters, not good, within the parameters needed for these tests. The same could be said about the components on his power supply. Not high end components, but if they were then it would cost multiple times as much. His camera probably doesn't cost $20K either, and the list goes on of what would be a waste of excess because of an inability to recognize what matters and what doesn't.
My trusty Fluke-11. 20 years old, I found it in my pocket when I Left Nokia... Measured it last year against a Lab. standard. Spot on! Doesn't measure current, which is when I pull out my (T)rusty AVO-8. 20 microanp movement - I can see when an interrrupt routine is working....
the only think that this supercaps can be useful, I think is to start a car engine if your car battery is discharged... but then here comes an other question... how long can this caps maintain their voltage? you need to know that, to be able to know when you have to charge them again.
These super caps are too small for that. They have a high ESR and a low current rating. You could use six of the 3kF Maxwell boost caps though. They are rated to supply 1900 amps. Even with those, you would only be able to start an engine once or twice per charge. They will also self discharge in a few days, so you have to keep them on a trickle charger.
Julian Ilett Hi I have a question for you I watched many of your Videos and I am working on a project. Let say you have AC 6 - 12 volt and you rectify it with a Bridge Rectifier to make DC voltage. I know that it will have a ripple but if you was you use this could you make the ripple less with this bank of capacitors? Or would you say get something better?
if they were genuine capacitors they wouldn't get warm when they are slightly overcharged at a low current. so they must be fakes knockoffs or copys of a different brand. with a lower voltage actual rating than the real thing and most likely lower capacity too.
Chinese businessmen buy patents in Europe and other countries from technological companies and not always they check the designs before rushing to produce it :] THe european companies usually sell the patents and licences for designs that are usually old or faulty in some point.
A normal capacitor rated in microfarads isn't exciting when you overvolt it slightly, a bit of magic smoke with some leaking. A supercapacitor is a bit more scary since it stores far more energy. If one of those 120 farad caps went off, that's 437.4 joules of energy. If it shorted out all at once, that's over 4 kW.
8% is the usual supercap manufacturer specified margin between rated voltage of 2.7 and maximum voltage, which would be 2.92V. Everything, including the tires, has a safety margin. A pretty substantial one usually, so the analogy isn't too good. 8% is actually very little leeway, however sufficient for the circuits used on this board to be good enough.
Hi. can i use this particular capacitor pack for my bike starting without much customization. bike is 12 v. i cannot use battery as i barely use bike. battery ends up dying.
The supercapacitor won't fail through lack of use like your lead-acid battery, but it will always be flat when you come to start the bike. Supercapacitors tend to self-discharge quite quickly.
Hi Julian. That was a quick reply. Thank you. I was in to avionics and understand that Caps get discharged quickly. If I do a kick start (manually using kicker) once, can I run the bike whole day using self start? Do I need any additional customization to get this work on a bike which used 12v lead acid battery. Now I use the bike with kick start :-). If no customization, then i am in for it. :-) I found the same product in your video in AliExpress.com for $17.68. Thank you
Hmm, not sure. The discharged capacitor will look like a dead short across all your bike electrics. Not sure what the high voltage generator will think of that! Only one way to find out - try it. You may have an awesome solution - you may have a busted bike ;)
It seems that supercapacitors rated voltage changes with the temperature of the environment. This spec for instance says they can handle 2.7V at 65 celcius and 2.3V at 85 celcius www.vishay.com/docs/28421/220edlcenycap.pdf So at 25 celcius these will be ok at even higher voltage i guess.
so the accuracy error could be up to 20 millivolts at that scale and the chip could be pretty accurate, that said your point about the chip not being that appropriate for that application is on point. btw i really like the work that you do, keep them videos coming :-D ( i have bought many of the gadgets/ electronics parts that you have reviewed)
Well, technically that's the accuracy when it was calibrated. It's more a statement of the meter's technical capability, than its actual present status. Now obviously pricey calibration of a kickabout toolbox meter for hobby use doesn't make any sense ... but if you knew someone that could take it into work and compare your meter to something traceable, it would definitely give you a lot more confidence when you're chasing 10s of millivolts on a meter that maybe had a couple of winters in the shed.
The detector is there to detect and can not dissipate any power, for that you need a MOSFET ( logic level ) and a resistor. Logic level MOSFETs have a RdsON at 3.3V or less. Then: 5V / 3A = 1.66 Ohm resistor. 5V / 4A = 1.25 Ohm resistor. The voltage can be set by a resistor divider, which can be found in the TL431 datasheet, and can then be calculated.
Not really. You'd want something with a hysteresis, which draws almost no current below the threshold, and starts bleeding the cap sharply when the voltage exceeds the threshold. TL431 could be used as a reference (albeit wasteful with its minimum current of 1mA), but together with a low power comparator & a LL MOSFET. And this chip does just that, with a quiescent current of less than one microamp. Not to mention the precise voltage trim and temperature compensation.
And capacitance 😂 I have a lithium cell here from China.. And the capacity is 78% lower than what it says Low internal resistance.. So that's the only good thing
Do you read all comments? I don't think so. So again: if you drain the caps in series down to (almost) zero, chances are you are actually charging one or more capacitors with a negative voltage! This is because of not having any balancing discharge regulation. Just draw the (simple) schematic. 6 caps in series, and a load. Now think of 1 or 2 caps having slightly more power (and thus voltage). The load (55W bulb) is almost a dead short. What will happen is the caps with a remaining voltage will start pushing a current trough the caps having 0 Volts. Tada! You are charging those NEGATIVELY. I won't stress this again. If you don't read it what's the point. I had to find this out just because it happened to me. If you relay this message, others won't make the same mistake. Or you for that matter ;-)
Hardly 'scary' - these are capacitors, not Li-on batteries! Although I agree that, on the surface, the over voltage devices appear to be somewhat unsuitable surely the purpose of the protective circuitry is for rare overvoltage conditions and not specifically for use when connected to a higher continuous voltage than specified. As some people have mentioned, perhaps the datasheet for the capacitors would shed a clearer light on the actual maximum overload voltage and as suggested by others, I suspect an overvoltage of less than a volt would likely, at worst, lead to reduced lifetime of the capacitors rather than some kind of catastrophic failure. What IS perhaps 'scary' is your use of hyperbole in these circumstances!
Even a 0.1V overvoltage on this type of capacitors will reduce the life significantly and a 0.2V can damage them almost instantly. This are also fake low quality super capacitors that do not have the stated capacity and they will probably fail very fast even at spec 2.7V Not to mention that they are unusable due to large difference in capacity, internal resistance and self discharge characteristics. This is based on my experience with a set of 8x 500F 2.7V capacitors I got not long ago from Ebay and look similarly to this also same low cost.
I trust your experiences with these cheap capacitors and it doesn't surprise me that they're a bit rubbish. But the module still has value - for playing, learning and experimenting.
Yes for playing and learning is good but not usable in a real application. I realy annoying that the ones I have will self discharge in 24 to 48h and not all of them will discharge at the same rate. Maybe the 120F are better than the 500F I have in this regard.
+electrodacus +julianilett It would be interesting to determine the actual capacitance of your capacitors. This could be done with an LCR or capacitance meter but most would likely give unreliable readings with such high capacitances. The other option is to use a signal generator, an oscilloscope and a few other passive components and calculate the capacitance value which is theoretically possible with any size of capacitor.
Julian, may I suggest you have a look at using OBS to capture and record video when looking at web sites etc rather than pointing the camera at the screen. It will make things a lot clearer to read. Above all OBS is free obsproject.com/
no., your board id DEFECTED. the moment I saw it go pass 2.5 volt and it not work. that mean it bad defected. I don 't need to see all the video. it can not reach 6 volt. for the protection of the Cap. thanks for your pst. thumb up before see all video
sounds like they pulled whatever they had available,thinking that would work even though its clearly not even designed to act as intended. pretty bad. The truth is,this happends in more electronics than you think. Televisions now days were engineered to only last so long ,then go bad. Those capacitors being a key reason. After time,they go bad. I remember the old commericals for televisions and how they would tell you their products were tested in so many ways,longevity ment something back then. When Dad went shopping for a TV. He bought one knowing it was built to last a long ,long time. after all,money doesnt grow on trees and the longer something lasted,the better it was! They even pushed that. Now days,thanks to ever changing technology,nobody WANTS a television thats going to last longer than 6 years. The fact of the matter is televisions,and electronics in general, are designed for a short life. They produce them so cheaply, that it doesnt matter any more. Its sad to think there is no more pride in the quality of things,that they are no longer designed and built to last for years. That includes most household appliances now days also.
I bought a 100 farad 2.7v super capacitor for testing. and charged it all the way up to 3.4v and it doesnt get even slightly warm Stayed cold and holds voltage for a long time. if these are getting warm at just barely 2.7v then they are ABSOLUTELY fake no doubt about it.
I did. Farads = seconds / ohms. The car bulb ran for about 40 seconds and is 4.6 ohms (when hot). That gives 8.7 Farads. Seems a bit low to me. Can anyone spot a flaw in my maths?
i wouldnt test it with a lightbulb because the resistance depends on the voltage and also there is an exponential decrease in the voltage if it would be an ohmic resistance. I would charge them with a constant current source until they get 1V. You can figure out the Capacity by measuring the Time. Capacity is charge per Voltage is As/V . You can get the Capacity by Multiplying the constantcurrent with the Time and Dividing it with the Voltage this should be really accurate!
+Julian Ilett Do you think you were getting the full 55W while the bulb was bright for 30 seconds? If so, you get (55*30) 1650Joules of energy. Therefore, based on E=0.5*C*(V^2), the capacitance would be over 12F (roughly 12.5F just during the brightest phase). Take into account there are a lot of assumptions in that calculation.
You should use a constant power load or at least consonant current for better precision but yes you are right capacity is much lower same as I observed with the low cost China supercapacitors that I got. Those that I sent to you are much higher quality you can do a test with those also. Here is the calculation same thing that I wrote to another comment here He has 6x 120F capacitors in there with a 2.7V voltage rating. The energy stored in each of those 6 capacitors is 1/2 CV^2 At 2.7V that will give you 437Joules and at 0.7V the remaining energy will be 29 Joules so usable energy is 437 - 29 = 408 Joules (same thing as W/second) or 408 Joules/3600 seconds = 0.113333Wh Since he has 6 of those in that module total module capacity is 0.11333Wh x 6 = 0.68Wh He used a 55W rate bulb but that is probably the rating at 13.8 or 14V and he had a bit more in the beginning but that dropped quickly so it maybe started at around 65W in the firs seconds then dropped down to almost nothing in those 40 or so seconds. Since resistance of that bulb is dependent on temperature it will be a bit hard to calculate the energy delivered in this video. If it will have been a 55W constant power load then that 0.68Wh should have lasted for (3600s * 0.68Wh)/55W =44.5 seconds. Since that was not the case in this video those capacitors have probably a much lower capacity than advertised so instead of 120F they are more like 60 or 70F.
I disconnected the bulb before it had gone out completely - and the filament resistance will have dropped as the bulb cooled. So modifying my measurements to 60 seconds divided by 4 ohms gives 15 Farads (much nearer the 20F pack capacitance).
Well this behavior is totally fine for protection purposes and hysteresis range seems to be quite reasonable, please stop making clickbait out of nothing.
Y-y-yeah, 2.84 is (gasp!) 5 percent above 2.7 volts, outrageous! In all seriousness, I doubt this is ever going to have any significant effect on the capacitors.
These things are ultra fussy about voltage. The absolute maximum surge voltage, not to repeat, (wording in spec sheet) is 2.84V for a quality make - and considering these are knock off, they will get buggered within a week doing this as I'm sure they can't even cope with the 2.7V. Also there is no problem finding new uses for devices like the voltage sense ics, as long as you get the engineering right - this aint right, and well spotted by the chap, even if it took him a week to put his meter on the blinkin relevant voltage.
Martin D A , and I'm pretty sure you've got evidence to support your claim that these capacitors can't handle 2.7 volts, right? I mean, Julian himself is in possession of these for how long, a year? More? If you are right, these caps must be all dead by now. Why don't we ask him?
operating out of specs is quite common in cheap devices. i really dont think with this low of an overcharge anything is gonna happen. 2.89 volts is quite a lot but not a big deal i's say.
2.7V is the rated voltage. The maximum voltage is typically stated in the datasheets to be 8% higher, so even at the worst case scenario we're just around the maximum voltage and not exceeding it. This module is meant to be used in automotive applications, where the voltage rarely exceeds 14.5V, and for the protection circuits to activate the capacitors would have to go out of whack big time (which with time they probably will..). If you are trying to cram every fraction of a Joule into this in a lab setting - perhaps 2.5 or 2.6V rated chips would be better, so as to keep the voltage below the RATED value as opposed to MAXIMUM value.
It's a bit like expecting to see a MOV rated at 245V in a mains overvoltage protection, which would be a nonsense.
I am sure the max voltage is way above nominal 2.7V of these supercaps. If so then 108% (takin max of nominal and hysteresis) would be 2.91V - still probably below the voltage of the capacitor to blow. On the other hand if this sensor detects exactly 2.7V to start charging it, then the capacitor would be always between 2.7-2.91V - not bad, just 10% of the U nominal
Bear in mind that the the data sheet for a 500F one was looking at today, gives the upper voltage limit of 2.85 volts. Seems they can tolerate quite a bit more that 2.7 Volts
Nice video!
But I do think it would work rather good if the 2,5V device was used. Unfortunately that version is a bit hard to get. Farnell only has the 2,7V and 4,5V versions... Digikey has them for 64 (Yankee) cents. Mouser about the same. You could try a sample request 0:-)
Having investigated further online, I am fast coming to the conclusion that the capacitors in your device may be either fake or falsely labeled. Elna manufacture no super capacitors in their DZ, DZH or DZN ranges with both a maximum voltage of 2.7v and a capacity of 120F. In fact, anything they make of that sort of capacity is significantly larger than the ones you have - Dynacap DZ range 2.5v 100F is 25mm × 50mm (Dynacap range 2.7v only goes up to 33F) and they do not make any capacitors in their Dynacap range in 1.2F,12F or 120F.
That doesn't surprise me ;)
Sir, what is the power supply circuit that you have use for supplying the caps? The one with 4 nos of push button switches?
take a look at his channel as i think he reviews it, iirc it's a buck regulated unit powered by his solar system
i might be wrong though
www.es.co.th/Schemetic/PDF/ELNA-DYNACAP.PDF ..I see as high as 300F
I also researched this subsequent to your comment, nearly 9 months after your comment and ELNA do not list these capacitors on their website. From my initial observations the one thing that does seem genuine is the capacity....
Hi Julian Thank you for the video. I have one of these CAPACITOR modules and going to use it with a crank generator. Some suggested to keep the voltage down to around 14.5V so I suppose I will add in a regulator to ensure it does not go any higher.
Hyristisis is not needed. Schmidt trigger also is same same thing folks.
How many w/h (energy/per hour)can this bank have? Can it be modded to replace the supercapacitors with for example 200F Supercaps? Can more pe used in series to act as an e-bike battery? Or not worth as a battery, but like a NOS boost like cars have? Can it be done so that a bank like this is charged when breaking an e-bike?
One other issue I see with this circuit is that when the capacitors are discharged (via the soldered on bulb), some will be reverse biased similarly to fully running down a pack of AA batteries where one battery runs down before the rest.
That wisp of smoke was some angry pixie magic.
Julian, top video, really well explained.
You never disappoint 👍
Thanks Dean - I always appreciate your support :)
I completely agree. This video was very educational.
Cheers! :-)
In my opinion, they are not a "protection" feature, they are a "balance" feature, you should not attempt to charge the entire module above the rated voltage.
In the data sheet's OPERATIONAL EXPLANATION, hysteresis operation is made a bit confusing in (1) and (5). In (1) where it is stated: when Vin ===IS HIGHER=== than Vdf, then Vout=Vin. and in (5) where it is stated: WHEN Vin ===RISES ABOVE=== Vdr, then Vout=Vin.
Thanks for the update. I don't take mine to anything above 14.5 volts total across the capacitor bank - hopefully that's well within the specs of each individual capacitor.
You're probably not going over the 2.7V upper limit, but I noticed when the module is fully discharged (courtesy of the 1W bulb) some caps were showing a small positive voltage and the others a small negative voltage. There's no easy way to 'bottom balance' the caps in this module.
Nice video, and interesting find regarding the protection IC.
Could you monitor the maximum voltages of the capacitors when you are charging and discharging at higher currents within the rated voltage of the capacitor bank to see how well they stay balanced? One datasheet for 2.7v supercaps caps showed an absolute maximum of 2.85v for 1 second non repeating, it would be good to see how long they lasted in an actual application.
anyone have a good recommendation for Supercapactior protector module ?
about 40 seconds of good light off that 55 watt bulb on the module, nice work 👍
He has 6x 120F capacitors in there with a 2.7V voltage rating.
The energy stored in each of those 6 capacitors is 1/2 CV^2
At 2.7V that will give you 437Joules and at 0.7V the remaining energy will be 29 Joules so usable energy is 437 - 29 = 408 Joules (same thing as W/second) or 408 Joules/3600 seconds = 0.113333Wh
Since he has 6 of those in that module total module capacity is 0.11333Wh x 6 = 0.68Wh
He used a 55W rate bulb but that is probably the rating at 13.8 or 14V and he had a bit more in the beginning but that dropped quickly so it maybe started at around 65W in the firs seconds then dropped down to almost nothing in those 40 or so seconds. Since resistance of that bulb is dependent on temperature it will be a bit hard to calculate the energy delivered in this video.
If it will have been a 55W constant power load then that 0.68Wh should have lasted for (3600s * 0.68Wh)/55W =44.5 seconds. Since that was not the case in this video those capacitors have probably a much lower capacity than advertised so instead of 120F they are more like 60 or 70F.
Yes they are not that useful in most applications. There are some particular applications where they can work well and even be cost effective compared to a battery.
They where used in some large EV in the past to recover the energy from breaking and release that immediately after for vehicle acceleration so there many many cycles/day.
As an analogy you can see the super-capacitors as the fast cache memory in computers vs RAM maybe represented by LiFePO4 (more like SRAM) then Flash represented by other high energy density Lithium cells like LiCoO2 and not sure if I want to compare HDD with Lead Acid :) it is worse than that.
Super cap can be recharged in a few seconds, LiFePO4 in a few minutes, LiCoO2 in a few hours ....
you could use supercaps a s batteries, but you would probably need to regulate/boost the voltage to get the same performance out of them - as with regular batteries they are relatively delicate and need to be protected. The fact that they don't degrade as much over charging cycles and can charge very quickly also potentially makes them more usefull.
I did read them :P, its just my thoughts on how you could turns these caps into the equivalent of a battery, using them directly wouldn't work well in a lot of applications as the voltage drops down all the way to zero whereas battery remain relatively constant. Not reading the comments is only true for my troll comments :>
thanks for the detailed explanation very informative on the price/capacity and capacity point :>.
Very useful instructions, tks you very much, I'm waiting for some new products from your hands on.
Yes I would rather 2.5v version with lower capacity. Rather than risking popping a capacitor. So will be looking to replace mine. What is the part number for the 2.5v IC?
.
I am wondering if the chip is actually intended for low voltage load disconnection of Lithium cells... I have the exact same capacitor module, interesting stuff from the datasheet for the chip. Looking at the data for the Maxell supercapacitors (presumably using the same chemistry) it seems they are rated for 2.85v, so this may not be a disaster at all.
I agree that this is the wrong protection voltage, but I believe you misread the data sheet. It's actual voltage design is 2.8V, not the 2.7V that you said it was.
Julian, My 2.5 volt protect c uses lm314 and uses no current under 2.5v and triggers a darlington overvoltage dump. Maybe we should use zeners over other ciruitsts?
It looks as if the leds some how cause a imbalance in the series circuit, it looks that way because the leds never flash in order.
Whats the model name/make of that super beefy looking dc-dc booster on the left? i think it could come in handy.
You could bodge the ground of the of the voltage detector with two resistors raising the ground. The device is low current so resistance could be a 10s of K so little loss. Better to use a good reference and 6 comparators and just use the FETs and load resistors. Could the FET be a darlington transistor as even low threshold FETS have perhaps 0.5 to 4V, so can't ensure would always work.
replace voltage detector with a TLV3012 comparator that has 1.24v internal reference, choose the cmos output and add potential divider or potentiometer to set wanted voltage, works down to 1.8v.
I'm rather surprised at not seeing any inclusion of the datasheet for the capacitors in this...
I have a question to you Julian. You are using the B3606 and B3603 (the step down modules). Im making my own solar charger to charge a power bank and I wanted to use one of these modules in it, my problem is that I dont know how they really work, If I set the limit to 5V, will they stop outputting when solar power voltge drops ? Or do I need something else at the end of my solar panels ?
i) The cap will be 2.7v over full temperature range. ii) the voltage that the chip triggers may well be different than the value on the DVM because of the current path in the PCB. Probably nothing to worry about, that said be wary of eBay crap...
Normally you would want at-least 10% headroom, 20% is even better. Using 2.4 volt Zener diodes in reverse bias is cheap, but effective over voltage protection.
Julian thx for all great videos and many inspirations !!!!!! But i have question , can I use suprecapacitors as battery for portable solar generator (2,7v 500F X 6 ) = 16,2v 82F . supercaps are same brand as those in video . Does it worth to buy them ???
All my balancing boards for my 2.7v supercaps starts turning them down at 2.85v and at 2.7v it shuts off and lets them charge back up.
Love your meticulous investigations and demonstrations. Did you try find the right IC (one turning on its output when voltage is 2.7) ? if so you could try replace them on the circuit board and make a video for us :)
Did you take a look on the super capacitors datasheets? Perhaps they can do more than 2.7 Volts.
Most supercaps have the rating printed on the component, if you watch the postbag video #68 the caps have a 2.7v rating printed on the shell.
blaser
Could be a median or mean voltage. I doubt that the caps are abused when just .1 volts overvolted...
***** that's not how it works, usually when designing electronics, you'd pick a capacitor rated at twice or even 3 times the max. voltage you're designing for. That's why if you crack open an led bulb the capacitor is rated around the 400V mark as most mains supplies are 100-240V
blaser
That's just because there is no cap which is exactly rated for the peak2peak voltage of 230V mains. ;) Like with resistors.
Super-capacitors much like other electrolytics, are typically designed and built to allow up to 20% over-voltage with only minor reduction in service life. 2.862 volts is only 6% out. Remember that these are over-voltage protection circuits and not true balance circuits. You should still be using this module with a proper balance circuit. The balance circuit on my super capacitor boost pack uses Zeners, resistors and LEDs IIRC to keep the capacitors balanced. The intended purpose of these super cap packs is automotive use so as long as a decent balance circuit is used. These protection circuits should never even need to kick in.
They could have easily dropped the voltage by .15 volts when they programmed the chips but then they would also have to say the max voltage of the pack should really only be around 15.5 which would still be fine for automotive use but would allow for a bit more safety margin.
I don't see what difference it makes which way round it is normally used. the hysteresis just needs to be set correctly.
well OK... that was interesting..So, now what? into the trash bin or will you try to desolder that chip and replace it with something else? Kinda hanging here wondering after that walk through..
Yeah me too.
Given it's just a circuit board I'm sure it's limited but what Amperage can you charge those at max? Wonder soldering a wire to bridge between capacitors would help. Nice update I can't wait to see you play with this thing a bit more.
what's the purpose of that diode in the output of your DC converter just before the supercapacitor bank?
Pal, when you gonna get a good DMM ? Fluke 87 , Brymen or Rigol Bench Multimeter ? haha....
Seriously, what's your problem with this Ryobi thing? In my view the main advantages of expensive brand multimeters are better security in case of high voltage/high current and expected better long-term stability. I prefer the brand stuff myself, but often I'm surprised how far people get here on youtube with a multimeter for less than 10 € if they know what they are doing.
The Ryobi is a good DMM, for what he is doing. High price is a decisively bad attribute, making all meters you listed BAD meters, not good, within the parameters needed for these tests.
The same could be said about the components on his power supply. Not high end components, but if they were then it would cost multiple times as much. His camera probably doesn't cost $20K either, and the list goes on of what would be a waste of excess because of an inability to recognize what matters and what doesn't.
theres me using a dollar store dimm to do mains voltage XD XD XD
My trusty Fluke-11. 20 years old, I found it in my pocket when I Left Nokia...
Measured it last year against a Lab. standard. Spot on!
Doesn't measure current, which is when I pull out my (T)rusty AVO-8. 20 microanp movement - I can see when an interrrupt routine is working....
I meant 50 microamp...
take the voltage at the capacitors and see if there a difference
hello new here just a quick question do you buy all your stuff from ebays usa site if so why is it just price or some other reason thanks
the only think that this supercaps can be useful, I think is to start a car engine if your car battery is discharged... but then here comes an other question... how long can this caps maintain their voltage? you need to know that, to be able to know when you have to charge them again.
These super caps are too small for that. They have a high ESR and a low current rating.
You could use six of the 3kF Maxwell boost caps though. They are rated to supply 1900 amps. Even with those, you would only be able to start an engine once or twice per charge. They will also self discharge in a few days, so you have to keep them on a trickle charger.
Well, this was moderately better then watching paint dry or grass grow. :-)
if you take into the account that all capacitors have a +10% voltage tolerance it's ok :) so capacitor should survive almost 3V
Well I am glad I bought the individual super caps and made my own over voltage circuit using 6 zener diodes and 6 transistor's it works better.
Julian Ilett Hi I have a question for you I watched many of your Videos and I am working on a project. Let say you have AC 6 - 12 volt and you rectify it with a Bridge Rectifier to make DC voltage. I know that it will have a ripple but if you was you use this could you make the ripple less with this bank of capacitors? Or would you say get something better?
just use bridge rectifier and regular electrolitic cap 2200uf for 1-3a load you can add a voltage regulator like lm317 as well
if they were genuine capacitors they wouldn't get warm when they are slightly overcharged at a low current. so they must be fakes knockoffs or copys of a different brand. with a lower voltage actual rating than the real thing and most likely lower capacity too.
Chinese businessmen buy patents in Europe and other countries from technological companies and not always they check the designs before rushing to produce it :] THe european companies usually sell the patents and licences for designs that are usually old or faulty in some point.
Could you use the 3v super cap instead?
Any update on the vocoder project Julian?
Haha - that one's in my winter collection
Aw man, I was looking forward to it :( Thanks for the great content though!
Watching the time on the video says you had just under one minute of light.
there is nothing to be nervous of i think. only they start leaking sooner, maybe.
A normal capacitor rated in microfarads isn't exciting when you overvolt it slightly, a bit of magic smoke with some leaking.
A supercapacitor is a bit more scary since it stores far more energy. If one of those 120 farad caps went off, that's 437.4 joules of energy. If it shorted out all at once, that's over 4 kW.
but those little 10 millivolts will not make it go boom, right ?
iceberg789
It's like an over inflated car tire, that tiny last insult can make it go bang.
8% is the usual supercap manufacturer specified margin between rated voltage of 2.7 and maximum voltage, which would be 2.92V.
Everything, including the tires, has a safety margin. A pretty substantial one usually, so the analogy isn't too good. 8% is actually very little leeway, however sufficient for the circuits used on this board to be good enough.
well taking that off the ' to get' list
Are super caps so sensitive to over voltage that .8 volts will damage them?
No idea, but .8 volts on a range of 2.7 volts is 30%, that sounds like a lot.
probaly meant .08
Hi. can i use this particular capacitor pack for my bike starting without much customization. bike is 12 v. i cannot use battery as i barely use bike. battery ends up dying.
The supercapacitor won't fail through lack of use like your lead-acid battery, but it will always be flat when you come to start the bike. Supercapacitors tend to self-discharge quite quickly.
Hi Julian. That was a quick reply. Thank you. I was in to avionics and understand that Caps get discharged quickly. If I do a kick start (manually using kicker) once, can I run the bike whole day using self start? Do I need any additional customization to get this work on a bike which used 12v lead acid battery. Now I use the bike with kick start :-). If no customization, then i am in for it. :-) I found the same product in your video in AliExpress.com for $17.68. Thank you
Hmm, not sure. The discharged capacitor will look like a dead short across all your bike electrics. Not sure what the high voltage generator will think of that! Only one way to find out - try it. You may have an awesome solution - you may have a busted bike ;)
We need it to work at 48v and 20ah
Hey Julian! What camera are you using these days then?
Still the Google Nexus 4 phone, but with the Open Camera app.
It seems that supercapacitors rated voltage changes with the temperature of the environment. This spec for instance says they can handle 2.7V at 65 celcius and 2.3V at 85 celcius www.vishay.com/docs/28421/220edlcenycap.pdf
So at 25 celcius these will be ok at even higher voltage i guess.
does your multimeter have really good specs? cause some of the error could be from the acumulative error
Accuracy would appear to be 0.8% plus 3 digits - manuals.ttigroupna.com/system/files/6464/original/RP4020_664_trilingual_04.pdf?2011
so the accuracy error could be up to 20 millivolts at that scale and the chip could be pretty accurate, that said your point about the chip not being that appropriate for that application is on point. btw i really like the work that you do, keep them videos coming :-D ( i have bought many of the gadgets/ electronics parts that you have reviewed)
Well, technically that's the accuracy when it was calibrated. It's more a statement of the meter's technical capability, than its actual present status.
Now obviously pricey calibration of a kickabout toolbox meter for hobby use doesn't make any sense ... but if you knew someone that could take it into work and compare your meter to something traceable, it would definitely give you a lot more confidence when you're chasing 10s of millivolts on a meter that maybe had a couple of winters in the shed.
I've got one of those precision 5V modules to show in a future Postbag video - I can test all 3 of my DVMs with that :):)
cant w8, almost bought one of those couple weeks ago, i bid to win an auction but i didn't get it in the end.
TL431 is the way to go, it's basically the same thing as you have there but without hysteresis and a 2.5V Vref.
What if I would like to load the caps with more Amps? Lets say 3 or 4 Amps at 5V. Which volt detector would you recommend?
The detector is there to detect and can not dissipate any power, for that you need a MOSFET ( logic level ) and a resistor.
Logic level MOSFETs have a RdsON at 3.3V or less.
Then:
5V / 3A = 1.66 Ohm resistor.
5V / 4A = 1.25 Ohm resistor.
The voltage can be set by a resistor divider, which can be found in the TL431 datasheet, and can then be calculated.
Not really. You'd want something with a hysteresis, which draws almost no current below the threshold, and starts bleeding the cap sharply when the voltage exceeds the threshold. TL431 could be used as a reference (albeit wasteful with its minimum current of 1mA), but together with a low power comparator & a LL MOSFET.
And this chip does just that, with a quiescent current of less than one microamp.
Not to mention the precise voltage trim and temperature compensation.
How's that solar flower going?
+xanataph It kept falling off the table and eventually broke. But I've just bought a new one. And a garden frog 🐸. Both to be over volted.
Actually, what would be far more interesting is if you over volt one of your lithium ion cells. lol
UK seconds most be really long. The bulb was lit for about a minute.
LOL how much ESR does the module have?
Not much by the look of it.
I wonder if Chinese Voltage is different from the rest of the world. (grins).
And capacitance 😂
I have a lithium cell here from China.. And the capacity is 78% lower than what it says
Low internal resistance.. So that's the only good thing
Proved my theory, eh? (smiles).
Do you read all comments? I don't think so.
So again: if you drain the caps in series down to (almost) zero, chances are you are actually charging one or more capacitors with a negative voltage! This is because of not having any balancing discharge regulation.
Just draw the (simple) schematic. 6 caps in series, and a load. Now think of 1 or 2 caps having slightly more power (and thus voltage). The load (55W bulb) is almost a dead short. What will happen is the caps with a remaining voltage will start pushing a current trough the caps having 0 Volts. Tada! You are charging those NEGATIVELY. I won't stress this again. If you don't read it what's the point. I had to find this out just because it happened to me. If you relay this message, others won't make the same mistake. Or you for that matter ;-)
+deslomeslager When I attached the DVM to that single cap, it was showing a negative voltage. About -200mV I seem to remember.
Sigh...
Julian Ilett
Thanks! You rock.
Hardly 'scary' - these are capacitors, not Li-on batteries! Although I agree that, on the surface, the over voltage devices appear to be somewhat unsuitable surely the purpose of the protective circuitry is for rare overvoltage conditions and not specifically for use when connected to a higher continuous voltage than specified. As some people have mentioned, perhaps the datasheet for the capacitors would shed a clearer light on the actual maximum overload voltage and as suggested by others, I suspect an overvoltage of less than a volt would likely, at worst, lead to reduced lifetime of the capacitors rather than some kind of catastrophic failure. What IS perhaps 'scary' is your use of hyperbole in these circumstances!
+Pete Allum But it was a miniscule, microscopic, infinitesimally tiny amount of hyperbole.
Even a 0.1V overvoltage on this type of capacitors will reduce the life significantly and a 0.2V can damage them almost instantly. This are also fake low quality super capacitors that do not have the stated capacity and they will probably fail very fast even at spec 2.7V
Not to mention that they are unusable due to large difference in capacity, internal resistance and self discharge characteristics. This is based on my experience with a set of 8x 500F 2.7V capacitors I got not long ago from Ebay and look similarly to this also same low cost.
I trust your experiences with these cheap capacitors and it doesn't surprise me that they're a bit rubbish. But the module still has value - for playing, learning and experimenting.
Yes for playing and learning is good but not usable in a real application. I realy annoying that the ones I have will self discharge in 24 to 48h and not all of them will discharge at the same rate. Maybe the 120F are better than the 500F I have in this regard.
+electrodacus +julianilett
It would be interesting to determine the actual capacitance of your capacitors. This could be done with an LCR or capacitance meter but most would likely give unreliable readings with such high capacitances. The other option is to use a signal generator, an oscilloscope and a few other passive components and calculate the capacitance value which is theoretically possible with any size of capacitor.
Julian, may I suggest you have a look at using OBS to capture and record video when looking at web sites etc rather than pointing the camera at the screen. It will make things a lot clearer to read. Above all OBS is free obsproject.com/
Very aesthetic!
no., your board id DEFECTED. the moment I saw it go pass 2.5 volt and it not work. that mean it bad defected. I don 't need to see all the video. it can not reach 6 volt. for the protection of the Cap. thanks for your pst. thumb up before see all video
sounds like they pulled whatever they had available,thinking that would work even though its clearly not even designed to act as intended. pretty bad. The truth is,this happends in more electronics than you think. Televisions now days were engineered to only last so long ,then go bad. Those capacitors being a key reason. After time,they go bad. I remember the old commericals for televisions and how they would tell you their products were tested in so many ways,longevity ment something back then. When Dad went shopping for a TV. He bought one knowing it was built to last a long ,long time. after all,money doesnt grow on trees and the longer something lasted,the better it was! They even pushed that. Now days,thanks to ever changing technology,nobody WANTS a television thats going to last longer than 6 years. The fact of the matter is televisions,and electronics in general, are designed for a short life. They produce them so cheaply, that it doesnt matter any more. Its sad to think there is no more pride in the quality of things,that they are no longer designed and built to last for years. That includes most household appliances now days also.
I bought a 100 farad 2.7v super capacitor for testing. and charged it all the way up to 3.4v and it doesnt get even slightly warm Stayed cold and holds voltage for a long time.
if these are getting warm at just barely 2.7v then they are ABSOLUTELY fake no doubt about it.
mvp tengelgeer
Good!
measure the farads
I did.
Farads = seconds / ohms. The car bulb ran for about 40 seconds and is 4.6 ohms (when hot). That gives 8.7 Farads. Seems a bit low to me. Can anyone spot a flaw in my maths?
i wouldnt test it with a lightbulb because the resistance depends on the voltage and also there is an exponential decrease in the voltage if it would be an ohmic resistance. I would charge them with a constant current source until they get 1V. You can figure out the Capacity by measuring the Time. Capacity is charge per Voltage is As/V . You can get the Capacity by Multiplying the constantcurrent with the Time and Dividing it with the Voltage this should be really accurate!
+Julian Ilett Do you think you were getting the full 55W while the bulb was bright for 30 seconds? If so, you get (55*30) 1650Joules of energy. Therefore, based on E=0.5*C*(V^2), the capacitance would be over 12F (roughly 12.5F just during the brightest phase). Take into account there are a lot of assumptions in that calculation.
You should use a constant power load or at least consonant current for better precision but yes you are right capacity is much lower same as I observed with the low cost China supercapacitors that I got.
Those that I sent to you are much higher quality you can do a test with those also.
Here is the calculation same thing that I wrote to another comment here
He has 6x 120F capacitors in there with a 2.7V voltage rating.
The energy stored in each of those 6 capacitors is 1/2 CV^2
At 2.7V that will give you 437Joules and at 0.7V the remaining energy
will be 29 Joules so usable energy is 437 - 29 = 408 Joules (same thing
as W/second) or 408 Joules/3600 seconds = 0.113333Wh
Since he has 6 of those in that module total module capacity is
0.11333Wh x 6 = 0.68Wh
He used a 55W rate bulb but that is probably the rating at 13.8 or 14V
and he had a bit more in the beginning but that dropped quickly so it
maybe started at around 65W in the firs seconds then dropped down to
almost nothing in those 40 or so seconds. Since resistance of that bulb
is dependent on temperature it will be a bit hard to calculate the
energy delivered in this video.
If it will have been a 55W constant power load then that 0.68Wh should
have lasted for (3600s * 0.68Wh)/55W =44.5 seconds. Since that was not
the case in this video those capacitors have probably a much lower
capacity than advertised so instead of 120F they are more like 60 or
70F.
I disconnected the bulb before it had gone out completely - and the filament resistance will have dropped as the bulb cooled. So modifying my measurements to 60 seconds divided by 4 ohms gives 15 Farads (much nearer the 20F pack capacitance).
2.786v
first :)
congrats ;)
Well this behavior is totally fine for protection purposes and hysteresis range seems to be quite reasonable, please stop making clickbait out of nothing.
B-b-b-but it makes the voltage going way over the designed voltage of the caps...
Y-y-yeah, 2.84 is (gasp!) 5 percent above 2.7 volts, outrageous!
In all seriousness, I doubt this is ever going to have any significant effect on the capacitors.
So, what does the data sheets of the capacitors say then? So we can end this dispute?
These things are ultra fussy about voltage. The absolute maximum surge voltage, not to repeat, (wording in spec sheet) is 2.84V for a quality make - and considering these are knock off, they will get buggered within a week doing this as I'm sure they can't even cope with the 2.7V. Also there is no problem finding new uses for devices like the voltage sense ics, as long as you get the engineering right - this aint right, and well spotted by the chap, even if it took him a week to put his meter on the blinkin relevant voltage.
Martin D A , and I'm pretty sure you've got evidence to support your claim that these capacitors can't handle 2.7 volts, right?
I mean, Julian himself is in possession of these for how long, a year? More? If you are right, these caps must be all dead by now.
Why don't we ask him?
operating out of specs is quite common in cheap devices. i really dont think with this low of an overcharge anything is gonna happen. 2.89 volts is quite a lot but not a big deal i's say.