Hopefully now people understand why the graphs cut off there hopefully. I really appreciate that you go THIS far to make it easy to understand for everyone not only engineers that are familiar with the terms and topics. This alone is a fairly demanding task that not even all teachers want to go through so for that you have my respect Dave. Keep up the good work.
MhDaMaster All i hope (but certainly wont happen) is that those guys to whom i wrote a text wall explaining why ESR wouldnt let you even power a led (which my numbers were pretty near to what dave meassured here) and why they graph below 0.8V was useless (1% left doesnt mean you can get it, not with 1Kohm+ esr @ 0,8v) will understand it now. Fun fact, one even told me he was "another experiences electrical engineer" (i doubt he was even an undergrad) and even funnier because i'm a computer sciences guy. It seems to me these kind of people see this as those others see those free energy devices. No one did this so far because evil iluminati corporations wanted you to throw away your batteries half way used. (or 89% full/~1,53V if you want to get that 800% battery life increment) But what does dave know.....the laws of physics are just numbers /sarcasm.
MhDaMaster So much this. Hopefully this wil put to rest the joule thief and near infinite energy from battery chemistry arguments. I really apreciate what this man does to explain simple scientific concepts and guide those who are gullible enough back into the scientific method.
Laharl Krichevskoy "It seems to me these kind of people see this as those others see those free energy devices." Yes this is exactly the reason those guys nitpick all day at every single minor detail/mispelling or whatever to "debunk" any detractors, at least in their heads.
Laharl Krichevskoy Well, 'power a led' really depends on what sort of application it is. Nanopower sensor equipment designed for energy harvesting can tend to get usable energy from a depleted chemical cell, though usually solar, rectenna, or tritium IC power may be preferred. Or fun stuff like lithium-air batteries. Usually you might design a system which draws power to an ultra-low esr organic polymer electrolytic capacitor, then you can convert, store in a capacitor, and enable the device once the stored charge reaches a peak. So you can run a lot of devices on low power sources, usually by adjusting the activity timescale. A pulsing LED for example. If you wanted to be really cute, you could derive energy from the oxidation of the metal the battery is made out of. It's perhaps just a question of 'application'.
Gravity What kind of consumer product would *completely* run like that though. Most if not all consumer products run on demand not when the charge of the capacitor drawing power from your extremely discharged battery reaches a peak or something like that. I honestly dont see that kind of products making their way into the consumer market let alone on a massive scale. On the other hand yes I can see the application for *certain parts* of a product, like the pulsing led from your example. But would it really be worth spending all the extra cash on the parts, design, manufacture, etc. required to get that "extra" energy compared to the savings on batteries you will make on the pulsing led for example?. Sounds like a super LONG term investment probably in some cases at really low power consumption the return rate would be so low that the life span of your product would run out before you make your money back. And thats not even accounting for accidents or anything of the sort. Maybe in the future when that tech is cheaper which would make the length of the investment shorter but I'm not certain for now. We would really have to crack the actual numbers maybe in certain cases its viable but it sounds niche to me.
Serial string commands are a PITA to decode. Changing field length, commands to mistype, sometimes protocols that are case-sensitive, values returned in weird E-notation or non-system decimal point. And as soon as you handle something like status codes you have to deal with a list of arbitrary numbers anyway.
The binary command format is much better because you can write a generic command processor that passes the data on to specific handler functions. Add in the checksum and you get a much more reliable data link.
At the point of "failure" look at the voltages reported by the 3 devices. You'll see as the source decreases, the disparity between them increases. [Verifies that the error is in the wires] At 0.1V@1A (100mW) on the load, the PSU is reporting 0.139V@1A (139mW), The voltmeter is showing 0.112V (112mW)
I use AA (Duracell coppertop) in my GPS. When the batteries are low and done, they are typically at 1.25V unloaded. So that is not too bad. I also noticed that at -20C the batteries last maybe 50% as long as opposed to+20C, and they do not recover when brought back inside (ie I do not get that other 50% back when the batteries are warmed up). I keep the GPS in my pocket during winter hikes but it still gets pretty cold.
Who said that the rule of attraction doesn't work. I was just trying to figure out how to calculate remaining energy in mWh or mAh and long be hold, a follow up video on battery capacity testing, there can never be enough info about that. Thanks for sharing. Keep up the good work Dave. Cheers :)
If the ambient temperature changes, it will greatly affect the characteristics of the battery under test, much more, than the test equipment. At very low power levels, the differences won't be nearly as much, than if you were testing at much higher power levels. Beautiful test rig!
rs232 and serial strings?? are you nuts? if you want to get as much data over it as possible you have to use binary. and its not that hard if you have the specification for it. also strings are horrible to parse. i just implemented a library for a little display that uses serial communication for its commands. wasnt that hard. even though the specification is in bad english. (if ya have the spec, i might take a look at that thing :P)
EEVblog I've read the datasheet for this load and found a little python implementation example that can help. Anyway I'm glad that you provided all this info about batteries discharge curves, it will come handy when I need to design a battery powered project!
Eletronicafg I agree it's not difficult. The checksum is only a modulus. Although if you came in expecting some easy ASCII commands to implement quickly then it's not as easy. The reason they did the packet as for reliability and quick communications. If you waste time with 7-bit ASCII to communicate spending 5 times or more communication time that's what isn't explained.
Eletronicafg Even though everyone may do ASCII if doesn't mean its the right way to communicate. Actually, it's very wrong for the most part. Most companies don't understand software when the engineers write the specification for the equipment communication. We still see communications without a even a simple checksum in industrial devices.
John Borchers Checksum is not always needed. For example USB has error correction integrated, so you don't need it in your protocol. And ASCII is easy to implement. If you use some binary packages, you have immediatly problems like how to detect package boundaries, and you can't debug or test it with a terminal program, which is really helpful for testing new protocols.
sghulten Apparently you are *not* a software engineer yourself, otherwise you wouldn't spread such non-sense. Text-based data is a human-readable format which means you need to write a parser and take care of different command lengths and so on. Sending and receiving hex-based commands is waaaaaay more easier to implement than having to do it in text mode. Plus, it's more reliable as well. I don't understand what makes you think otherwise but I assume you're not a software engineer.
EEVblog Dave, you're wrong here. It's easier to process a hex-based protocol as opposed to a text-based one. A text-based interface means you have to parse the text output, extract the commands, their parameters and deal with different message lengths because every text command has a different length. When you have a hex-based protocol for which you even have the specification which means there is no reverse-engineering involved, you merely have to write a simple send/receive function which puts the hex numbers on the serial interface and calculates the checksums according to the specification. Then you read the answer, strip off the meta data and output it to a text file which you can easily plot with gnuplot. Couldn't be simpler.
for what is worth, the 8600 model actually supports text commands, see bkpmedia.s3.amazonaws.com/downloads/programming_manuals/en-us/8600_Series_programming_manual.pdf
I think you should design a test and logging rig. Something arduino compatible. Then test a bunch of manufacturers, lots, chemistries, etc. Keep "big battery" honest. I really love the video series where you take a set of specs, start with simple circuits, discuss/expand. Then end up with a fully functional product. Like the power supply series.
from a chemists perspective im not surprised to see the sharp drop off. the chemicals inside are only designed (chemically able?) to give a small range of voltage.
Maybe the large voltage drop below 0.8 volts is due to the internal resistance of the battery not allowing the current to flow below that voltage level. That may be the reason for the oscillation, as the internal chemistry changes due to lack of electric charge. The alkali material becomes more corroded (corrosive) as the battery looses charge. That's why alkaline batteries burst and corrode when they are old, destroying PCBs.
The problem with setting the load to constant power and measuring only the voltage is that the power in the load is not exactly equal to the power in the battery, the difference being the power lost in the wires, contrasted with constant current of course. As you showed in the video, the difference can be very significant at high current and low load power where the wire resistance becomes significant compared to the effective load resistance. If you measured the voltage at the load you'd still get an error, just in a different way. If it was constant current you could take only voltage measurements and know the current in the wires, load and battery are all the same, but the energy lost in the wires would still affect the time axis of the plot. So there's no ideal way to do it, all you can do is make the wire resistance as low as possible and/or measure current too. Also, does your load let you know if it can't meet the set point? That might be a source of error if you're not measuring current but assuming constant power. It would all be good enough to prove a lack of energy below 0.8V. Even this video shows that pretty well with 1 minute resolution, saying it takes "under a minute" to drop from 0.7V to 0V at 100mW is nearly as good as saying "it takes so many seconds which is less than a minute", because those seconds are practically useless. Some electronic loads often have oscillation problems in certain modes like constant power or resistance, especially when combined with a regulated source. I think it's because of too many feedback loops trying to control the same thing with different bandwidths.
KX36 For the powers I'm talking about the voltage drops in the wires are petty small, almost negligable. It will only get smaller at lower power settings.
As you know, the closer the voltage gets to 0V, the higher the current gets for constant power and then it _may_ become significant (in a more academic sense; only if you really want a very accutate measure of what's happening in this tiny fraction of the discharge curve). Look at the voltage difference between what the power supply and the load read at 0.1W, 0.1V, 1A on the load, as you noted it's a 40% error at that point. But yes, it is practically insignificant beause through >99% of the battery life because the voltage is far greater than this and therefore the current far less until the last second or 2 of the experiment.
There are some red dot sights for rifles that use a single button cell and have a life of 10 years. You're actually supposed to turn it on and it should glow for 10 years.
I have a small motorcycle battery here, it's lead-acid 12V 3Ah, and it's got some sort of defect: It charges fine, and when it's charged it has no problems giving out power, I can easily use a 12V 50W light bulb on it, it will draw 4A without problem, but when it sits idle without anything connected, the voltage drops at least 0.1-0.15V every 24 hours (roughly). I've let it sit for a couple of weeks now, and it has drained down to 4.35V a few days ago, and now it's around 4.2V, so now it only drops 0.05V every 24 hours. Does this mean "constant resistance", and now there's so much less current flowing that it will drain slower and slower? I've tried cleaning around the acid filler ports (it's a really old battery that has these little knobs you can open to refill), but I haven't gotten around to finding out where the drain comes from. Because, when I see its discharge behavior with higher loads, I think it is still somewhat good for something (like a backup battery for an alarm system or something that needs long standby times without drawing as much as the starter of an engine would), if just there wasn't that annoying drain!
The BK precision meter is using a PID algorithm for constant power output. PID terms are tricky to optimize for all conditions. Not enough derivative term and you get oscillation as is shown by this meter's response. Of course there's no energy here which is why the companies don't show
John Borchers Really? How do you know? Doesn't really make sense to me using a PID, because of the nonlinearities of the 1/x function here.I guess it is simply sampling the voltage at some interval and then setting its classic linear loop current source to the calculation I=P/V.This sampling interval and some group delay from the voltage to the set current combined with the battery impedance/equivalent circuit gives the oscillations, i guess.Maybe Dave can check the BK by giving slow AC lower than 1Hz (a function generator could maybe give 50mA) to the load in constant power mode and measure the current with a scope. I think you will see some aliasing on this current.Might be very insightful to find out how it actually is behaving in constant P mode, as the manufacturers don't give much information.
Raymund Hofmann Because it appears it's overshooting I=P/V. There are also messages in the forum regarding this meter. I stand by the claim it absolutely has a PID control algorithm.
You should actually get a batterizer when they are available on the market (although it means giving them money) and make a video showing how the converter craps itself just a few seconds after the battery alone (or even earlier due to the loss in the converter). And showing how battery gages dont manage to shut down devices sefely before the battery is empty or how the thing shorts the battery out and start a fire because it piered the protective foil of the battery (which is a much stronger point than the performance to me).
Hi, I work in a primary school. The teachers use digital cameras for part of the children's course work. The teachers routinely require new battery's as the don't last long. I have released that the camera's have a very high cutoff point and now when they request new battery's i ask for the old ones because the still have enough energy for IR remote's which last a long time. I think digital cameras are either poorly designed or we're using the wrong batteries. I only use Duracell ultra power.
hue mondesir Your problem is the use of Alkaline batteries for cameras, that's bad! You need NiMh rechargables as they have much lower ESR for higher peak currents.
hue mondesir No, you're not using the wrong batteries. What's wrong is that you use batteries at all! The only real advantage of alkalines over rechargeables is their low self discharge, so you can put it in a device that has to run for month or years. A camera does not need that.
EEVblog Thanks for the reply Dave, but have you ever worked with teachers, they may be able to teach children but when it comes to technology that's another story. I couldn't imagine how many rechargeable battery's would get discharged to the point of no return or just get thrown away. we have 30+ cameras and it's driving me crazy with the amount of batteries we go through.
Good video Dave. Maybe you can come up with a precise explanation for, or eliminate, the noise/oscillation that the load is reporting after the brick wall response. Of course it's gibberish and not actual power remaining in the battery, but you know the nay-sayers will point to that and go "Look! That's the energy we're tapping into and you discounted it!". Because unfortunately it *looks* like more area under the curve, so it has to be explained to cover all bases.
hey dave, could you take a few samples of the ESR as the batteries discharge? ESR is more important that that "extra energy" left as it will tell whether you can get it in a bucket or drop by drop.
It's not about how much energy is left below .8 volts because most electronic devices ,which is prob 99% + , that are on the market won't operate much below .8V due to p-n junctions. In a very scant few devices that can operate below .8V , the time left is ridiculously low for a battery that started out as 1.5 V under load.
WisdomVendor1 Yes, that's the whole point of the video, to find out how much that "ridiculously low" time left actually is. if you can data elsewhere for it please add a link.
It would be nice to atleast sneek in a wang-ho-low/noname brand of batterys and see a discharge curve on those to see if there is more power left after that 0.8 limit compared to some of the more known and proven brands like duracell.
It would be nice to see whether the rated energy storage (1000 mAh?) matches the actual energy based on the measurements you take. Did the thing really leak on the shelf?
You've got a lot of faith in the display on that electronic load... You should also be measuring the amperage independently to verify that the BK Precision is telling the truth...
icesoft1 well, he's power supply did it. And it said it was spot on. The problem is in the software, the load isnt analog, so if it isnt fast enough that can happen. It had the same problem with constant resistance mode which he showed a while ago.
Doh... I misread the display on his power supply during the test. Guess I just wasn't getting any warm and fuzzies from the hysteresis loop in the battery test and the data acquisition software...
Is there a part 2 of this? Sounded like this would be a longer series, but I can't find the second part. Or are "Alkaline Battery Discharge Testing Part 2 + 3" from the second channel the sequel to this, although they came 5 years later?
You might try some different battery chemistries like carbon-zinc, lithium, and the alkaline that you're already running. Would be interesting to see how much power one could squeeze out of the other chemistries (like if there were a better chemistry based on the type of load).
icesoft1 carbon-zinc are crap anyways. lithium i'm not sure, but li-ion and li-poly can catch fire if over discharged. Not sure about the non rechargable lithim ones. They do have lithium metal inside so unles they do not, there's always the posibility for lithium to catch fire (just exposure to air does the trick) if anything, it would be more interesting to meassure the ESR at different charge levels as it would actually tell you when you really cant do shit with it, even if there's still some small percentage of "energy" left
when Dave was referring to his "big,massive" cables, I was like "take a look at my 95mm sq" ones attached to test rig for a controller-motor testing. for you US boys its 0000 cables.
I like having the TH-cam captions on for people with accents. eg. 16:38 "The constant power mode" = "The constant Pelham high road" "Some sort of oscillation" = "Some soda australasian" LOL XD Asian Aussie hybrid
Dave, you are so right about the BK load; it's a great instrument, but the software SUCKS! I had to write my own LabVIEW drivers. Also, the isolated RS-232 cable seems to have very poor common-mode rejection, or at least mine does. You should check that. It could make a very interesting video.
This reminds me of something that happened to my sister a couple of years ago. She called me up saying her Home air conditioning system was not working, The conversation went somethinglike this. My sister, My A/C stopped working, what do i need to check, and how do i check it. Me, do you have a digital thermostat. my sister, what is that. Me, does it have a display that shows the temperature on it. My sister, yes.. Me, check the batteries. My sister, Does it have batteries. I've lived here 14 years and never knew it had batteries. She goes and checks the batteries. Sure enough they were dead. After changing the batteries; the air-conditioning system worked fine. My sister, Who knew the batteries would last 14 years. PS, Before you ask, no I do not know what kind of batteries they were. and yes this really happened.
A single battery-test should not exceed a month or maybe two. Otherwise you may get a problem if your USB stick is formatted in a FAT32 filesystem (4 GB maximum filesize).
Luncher91 maybe 2 months??? 4GB = 2^32 bytes. 2 months = 86400 minutes... Assuming that a single value (V) is being written into the CSV file including the comma is 7 bytes (as in "#.####,"), it should take about more than 1145 years to reach the 4GB file limit... What number are you using???
sorry i dont know what i have calculated. you are absolutly right. i think i divided by 60 one times too often. :-D But i assumed a measurement per second. Correct would be for 7 bytes/second : 2^32 bytes / (7 bytes/second) / (60 seconds/minute) / (60 minutes/hour) / (24 hours/day) / (365 days/year) = 19.46 years So i no problems by the amount of data. Sorry for that absolutly incompetent post!
Hi Dave: How about putting a cap (+/- a resistor in series) to damp the oscillation at the end. Initial thought is that it should not effect the curve other than to smooth the end. Yes/No? Cheers, Mark
Mark Beeunas I think this capacitor will only change the frequency of the oscillation, because the equivalent circuit of the battery, as the recovering of its voltage suggests, is already some kind of RC circuit after the "big capacitor" ie the chemical engergy source.
I have a 50ah sealed lead acid (not a Car one) charged it is 12.8v what is min voltage Can i go to with out damaging it ie: 10.5v or 11.5v open curcuit volt, when i put my camp fridge on the battery volt drops from 12.8v to 12.1v while the load is on. hope explained it
well, why not just drop its energy with a 1 volt zener and current limit then start to log with high density readings a 100mW or just a 10mW that will save your day. though I bet a battery can't survive a .5Amp when its voltage drop, it cant maintain that rising current.
any chance you could do some vids of LiPo vs LiFe batteries as these are getting so popular and happen to be so dangerous, more often than not when hit with a hammer and thrown in a microwave. No that's been done , just like to know the more indepth tech info on the charge numbers especially for LiFe as there's so much opinion out there and it seems the manufactures don't always give the info.
Sean McMunn Better wires would clean up the measurement at the really low end, sure. But what real life products have anything better than what he is using in this video? I think that, in a way, makes this result even more accurate.
KX36 that's what I was pointing at, having good tools (which most often means expensive, especially with electronics) is what makes stuff like this easy
Lol at 4:56: you show what looks a complete list of commands (with the packet structure) that your BK Precision supports and... you immediately abandon that idea because... it's too hexadecimal? Too well documented? ...C'mon now Dave/Dave2, sorting that out looks like 20 minutes of gentle Python coding while sipping a cup of tea... :-)
Got any results from this yet? Though, you can fly to the moon with 100mW. Imagine some AVR monitoring a sensor every few seconds and sleeping (0.1µA) the rest of the time, maybe blinking an LED now and then... What would that take, 1mW on average?
Are you sure the packet structure isn't just Modbus? The Maynuo versions just use Modbus over RS232. I wrote a bit of python code that reads out values (never got around to writing the rest that actually sets values). github.com/MetalPhreak/pyMaynuo
Yay, not nice to try confront the world with some scientific facts. But one thing got me wondered, under 0.6V are you even able to draw 100mW out of it? I would guess there is a point where you can't get it anymore but there is still some voltage on it... that point would be intresting, also like a curve V vs. Power you can still draw from it.
Dave, why you using 9999999999$ ultra precision equipment(that also dont works well for this test) instead of using Accucel 6 charger that is designed to measure batteries, It also have uart>usb connection capability So we can get graphs on PC!
Arek R. What I have works great. Does the Accucel 6 even do data logging to hundreds of thousands of samples? I don't want a PC running for weeks just logging. Also, does it do constant power down to 1mW resolution?
EEVblog No it dont, but does we need very precision data for just ordinary batteries? And I also was thinking about my accucel 6(whoops) that have 1mV(oversampling) read capability, so, you are right about acc6 with standard FW....
sleepib the led draws ~20mA and UV leds are around 2V, so about 40mW on the led, plus the rest of the logic/asic. It probably will work up down to 1.75V (2x 0.9V) So a remote most probably wont work below 0.8V anyway, and if it does the led just wont turn on. IR maybe gets away with 1.3V (red leds are about 1.5v) so maybe around 20mW, still you cant get many mA at such low voltages, remember ESR goes up, (the brick wall you saw in the graph) and below 0.8V it can easily go above 500ohms or more. You wont get 800µW with 500ohm and 0.8V. In both cases most probably it wont work below 0.8V
Laharl Krichevskoy The LEDs of IR remotes are driven with a much higher current than just 20 mA, more like up to 1 A; but of course pulsed with a very low duty cycle, and using a beefy enough capacitor to deliver the pulse current.
Jo Pe still you cant go above the max rms power, else you burn it. The point is you cant even get 1/10 of those 20mW. last time i did the math it was like 800 something µW, you cant do shit with that, plus with no voltage you wont be able to charge that cap to get that pulse
Laharl Krichevskoy Mainly, I'm thinking a basic TV remote control would be an ideal use case for a DC-DC converter, provided your converter's output is regulated to just above the minimum operating voltage. If we assume that's 1.3v, and that you're using 2 AA batteries, A DC-DC converter should be a big improvement in battery life, not because of the lower cutoff voltage, but because you're not dropping almost half your starting voltage in a linear regulator. I think we also need some practical testing of how bright the led actually needs to be, and how long the pulses need to be in order for the receiver to detect it. If we're running the LED at 10x higher power than we need to, that's just more wasted battery life.
sleepib i doubt they would put a DC-DC inside one of those cheap chinese remotes. Also dont thing it will make it last longuer as you maybe will be able to get 10% more out of the battery but you also lose more than 10% (eff is usually around 70-80%). plus dc-dc used chips, diodes, caps and inductors. It's really not worth the cost. It really doesnt make a lot of sense in trying to be greener with a remote that lasts more than a year with a pair of AA. It would be better to just use rechargeable batteries (the already charged nimh ones) or stick a lipo battery and forget about it
EEVblog I'd just watched his latest video, it all started to get confusing for me when he explained there was in infinite amount of numbers between 1 and 2. Certainly food for thought that I hadn't considered before.. When it comes to a battery, I should imagine I'd be equally confused by his complex views on it. Secretly I'm now wondering if your Duracell has an infinite amount of energy even if if was unmeasurable. lol 😳
Batteriser Batteroo 1 month ago +dismayer666 Dave jones uses a power supply box to debunk the product, which completely ignores ESR and internal resistance of a battery. Don't believe everything you hear! Ive personally seen the batteriser work. Dave jones is a dramatic arts school graduate. The inventors of Batteriser are pHDs of electrical engineering and also hold over 700+ patents being used today by billion dollar companies in the sillicon valley, California.
EEVblog I just dont understand the stupidity of well known companys, you have a winning name that everyone knows as good quality, oow lets change it for no reason at all, silly buggers lol.
"Usable" voltage . My kitchen analogue clock is usually the recipient of any .08v batteries . Using half a bee's dick, It will get a extra week of usable life and keep time from a AA battery that wood otherwise be rubbish at .08v cause not much use for other devices . Some l.e.d torches I have won't turn on under 1.0 v ..
Hopefully now people understand why the graphs cut off there hopefully.
I really appreciate that you go THIS far to make it easy to understand for everyone not only engineers that are familiar with the terms and topics. This alone is a fairly demanding task that not even all teachers want to go through so for that you have my respect Dave. Keep up the good work.
MhDaMaster All i hope (but certainly wont happen) is that those guys to whom i wrote a text wall explaining why ESR wouldnt let you even power a led (which my numbers were pretty near to what dave meassured here) and why they graph below 0.8V was useless (1% left doesnt mean you can get it, not with 1Kohm+ esr @ 0,8v) will understand it now.
Fun fact, one even told me he was "another experiences electrical engineer" (i doubt he was even an undergrad) and even funnier because i'm a computer sciences guy. It seems to me these kind of people see this as those others see those free energy devices. No one did this so far because evil iluminati corporations wanted you to throw away your batteries half way used. (or 89% full/~1,53V if you want to get that 800% battery life increment)
But what does dave know.....the laws of physics are just numbers /sarcasm.
MhDaMaster So much this. Hopefully this wil put to rest the joule thief and near infinite energy from battery chemistry arguments.
I really apreciate what this man does to explain simple scientific concepts and guide those who are gullible enough back into the scientific method.
Laharl Krichevskoy "It seems to me these kind of people see this as those others see those free energy devices."
Yes this is exactly the reason those guys nitpick all day at every single minor detail/mispelling or whatever to "debunk" any detractors, at least in their heads.
Laharl Krichevskoy Well, 'power a led' really depends on what sort of application it is. Nanopower sensor equipment designed for energy harvesting can tend to get usable energy from a depleted chemical cell, though usually solar, rectenna, or tritium IC power may be preferred. Or fun stuff like lithium-air batteries.
Usually you might design a system which draws power to an ultra-low esr organic polymer electrolytic capacitor, then you can convert, store in a capacitor, and enable the device once the stored charge reaches a peak. So you can run a lot of devices on low power sources, usually by adjusting the activity timescale. A pulsing LED for example.
If you wanted to be really cute, you could derive energy from the oxidation of the metal the battery is made out of. It's perhaps just a question of 'application'.
Gravity What kind of consumer product would *completely* run like that though. Most if not all consumer products run on demand not when the charge of the capacitor drawing power from your extremely discharged battery reaches a peak or something like that.
I honestly dont see that kind of products making their way into the consumer market let alone on a massive scale.
On the other hand yes I can see the application for *certain parts* of a product, like the pulsing led from your example. But would it really be worth spending all the extra cash on the parts, design, manufacture, etc. required to get that "extra" energy compared to the savings on batteries you will make on the pulsing led for example?.
Sounds like a super LONG term investment probably in some cases at really low power consumption the return rate would be so low that the life span of your product would run out before you make your money back. And thats not even accounting for accidents or anything of the sort.
Maybe in the future when that tech is cheaper which would make the length of the investment shorter but I'm not certain for now. We would really have to crack the actual numbers maybe in certain cases its viable but it sounds niche to me.
Serial string commands are a PITA to decode. Changing field length, commands to mistype, sometimes protocols that are case-sensitive, values returned in weird E-notation or non-system decimal point. And as soon as you handle something like status codes you have to deal with a list of arbitrary numbers anyway.
The binary command format is much better because you can write a generic command processor that passes the data on to specific handler functions. Add in the checksum and you get a much more reliable data link.
Lab is looking unusually clean.
This small and simple stuff is just what I am searching for, hats up for Dave!
At the point of "failure" look at the voltages reported by the 3 devices. You'll see as the source decreases, the disparity between them increases. [Verifies that the error is in the wires]
At 0.1V@1A (100mW) on the load, the PSU is reporting 0.139V@1A (139mW), The voltmeter is showing 0.112V (112mW)
I use AA (Duracell coppertop) in my GPS. When the batteries are low and done, they are typically at 1.25V unloaded. So that is not too bad. I also noticed that at -20C the batteries last maybe 50% as long as opposed to+20C, and they do not recover when brought back inside (ie I do not get that other 50% back when the batteries are warmed up). I keep the GPS in my pocket during winter hikes but it still gets pretty cold.
Who said that the rule of attraction doesn't work. I was just trying to figure out how to calculate remaining energy in mWh or mAh and long be hold, a follow up video on battery capacity testing, there can never be enough info about that. Thanks for sharing. Keep up the good work Dave. Cheers :)
If the ambient temperature changes, it will greatly affect the characteristics of the battery under test, much more, than the test equipment. At very low power levels, the differences won't be nearly as much, than if you were testing at much higher power levels. Beautiful test rig!
rs232 and serial strings?? are you nuts? if you want to get as much data over it as possible you have to use binary. and its not that hard if you have the specification for it. also strings are horrible to parse.
i just implemented a library for a little display that uses serial communication for its commands. wasnt that hard. even though the specification is in bad english.
(if ya have the spec, i might take a look at that thing :P)
Well, I think that this serial protocol isn't so difficult to implement. Just a couple of hex commands.
Eletronicafg Sure, but it's just a PITA so we ditched it on principle.
EEVblog I've read the datasheet for this load and found a little python implementation example that can help.
Anyway I'm glad that you provided all this info about batteries discharge curves, it will come handy when I need to design a battery powered project!
Eletronicafg I agree it's not difficult. The checksum is only a modulus. Although if you came in expecting some easy ASCII commands to implement quickly then it's not as easy. The reason they did the packet as for reliability and quick communications. If you waste time with 7-bit ASCII to communicate spending 5 times or more communication time that's what isn't explained.
Eletronicafg Even though everyone may do ASCII if doesn't mean its the right way to communicate. Actually, it's very wrong for the most part. Most companies don't understand software when the engineers write the specification for the equipment communication. We still see communications without a even a simple checksum in industrial devices.
John Borchers Checksum is not always needed. For example USB has error correction integrated, so you don't need it in your protocol. And ASCII is easy to implement. If you use some binary packages, you have immediatly problems like how to detect package boundaries, and you can't debug or test it with a terminal program, which is really helpful for testing new protocols.
As a software engineer I appreciate the hex-based communication interface!!! Less bandwidth, more data!
sghulten Me too. In the case of an electronics load, speed isn't really a requirement.
sghulten Apparently you are *not* a software engineer yourself, otherwise you wouldn't spread such non-sense. Text-based data is a human-readable format which means you need to write a parser and take care of different command lengths and so on. Sending and receiving hex-based commands is waaaaaay more easier to implement than having to do it in text mode. Plus, it's more reliable as well. I don't understand what makes you think otherwise but I assume you're not a software engineer.
EEVblog Dave, you're wrong here. It's easier to process a hex-based protocol as opposed to a text-based one. A text-based interface means you have to parse the text output, extract the commands, their parameters and deal with different message lengths because every text command has a different length. When you have a hex-based protocol for which you even have the specification which means there is no reverse-engineering involved, you merely have to write a simple send/receive function which puts the hex numbers on the serial interface and calculates the checksums according to the specification. Then you read the answer, strip off the meta data and output it to a text file which you can easily plot with gnuplot. Couldn't be simpler.
from a programming perspective, it makes no difference, text is easier to write libraries with
for what is worth, the 8600 model actually supports text commands, see bkpmedia.s3.amazonaws.com/downloads/programming_manuals/en-us/8600_Series_programming_manual.pdf
This is a good review about test setup. I like it very much. Thanks Dave.
I think you should design a test and logging rig. Something arduino compatible. Then test a bunch of manufacturers, lots, chemistries, etc. Keep "big battery" honest.
I really love the video series where you take a set of specs, start with simple circuits, discuss/expand. Then end up with a fully functional product. Like the power supply series.
from a chemists perspective im not surprised to see the sharp drop off. the chemicals inside are only designed (chemically able?) to give a small range of voltage.
Maybe the large voltage drop below 0.8 volts is due to the internal resistance of the battery not allowing the current to flow below that voltage level. That may be the reason for the oscillation, as the internal chemistry changes due to lack of electric charge. The alkali material becomes more corroded (corrosive) as the battery looses charge. That's why alkaline batteries burst and corrode when they are old, destroying PCBs.
21:08 Okay, after the bench meter fixing fun, I think either your equipment or the workshop itself, but something is posessed for sure :D
I enjoy these battery measurement videos.
Thanks Dave.
Mohammed Algailani Glad to hear. I actually have quite a lot of battery related videos now, I think 10 on the playlist or something.
It would also be interesting to see what some of these low power draws look like when using a low self discharge NiMh battery.
You should shoot the IR camera at the setup and see where the power is going at 100mW. I'm thinking that most of it is happening at the connections.
Dave, what if you used a Batteriser? I heard it gives you 800% battery life! ;)
Stephen Furr If your not just taking the "piss" then try EEVblog #751 and you'll get Dave's answer..
Macca Wasn't it blatantly obvious he was taking piss?
AnthonyH Yes well I suppose the 800% was a dead give'a'way,but you never know..
Yes, I was just trying to be funny. I knew Dave had already reviewed it. Thanks!
Stephen Furr 800% or 8 times? They aren't the same!
Where is Part 2?
The problem with setting the load to constant power and measuring only the voltage is that the power in the load is not exactly equal to the power in the battery, the difference being the power lost in the wires, contrasted with constant current of course. As you showed in the video, the difference can be very significant at high current and low load power where the wire resistance becomes significant compared to the effective load resistance. If you measured the voltage at the load you'd still get an error, just in a different way. If it was constant current you could take only voltage measurements and know the current in the wires, load and battery are all the same, but the energy lost in the wires would still affect the time axis of the plot. So there's no ideal way to do it, all you can do is make the wire resistance as low as possible and/or measure current too.
Also, does your load let you know if it can't meet the set point? That might be a source of error if you're not measuring current but assuming constant power.
It would all be good enough to prove a lack of energy below 0.8V. Even this video shows that pretty well with 1 minute resolution, saying it takes "under a minute" to drop from 0.7V to 0V at 100mW is nearly as good as saying "it takes so many seconds which is less than a minute", because those seconds are practically useless.
Some electronic loads often have oscillation problems in certain modes like constant power or resistance, especially when combined with a regulated source. I think it's because of too many feedback loops trying to control the same thing with different bandwidths.
KX36 For the powers I'm talking about the voltage drops in the wires are petty small, almost negligable. It will only get smaller at lower power settings.
As you know, the closer the voltage gets to 0V, the higher the current gets for constant power and then it _may_ become significant (in a more academic sense; only if you really want a very accutate measure of what's happening in this tiny fraction of the discharge curve). Look at the voltage difference between what the power supply and the load read at 0.1W, 0.1V, 1A on the load, as you noted it's a 40% error at that point.
But yes, it is practically insignificant beause through >99% of the battery life because the voltage is far greater than this and therefore the current far less until the last second or 2 of the experiment.
There are some red dot sights for rifles that use a single button cell and have a life of 10 years. You're actually supposed to turn it on and it should glow for 10 years.
I have a small motorcycle battery here, it's lead-acid 12V 3Ah, and it's got some sort of defect: It charges fine, and when it's charged it has no problems giving out power, I can easily use a 12V 50W light bulb on it, it will draw 4A without problem, but when it sits idle without anything connected, the voltage drops at least 0.1-0.15V every 24 hours (roughly). I've let it sit for a couple of weeks now, and it has drained down to 4.35V a few days ago, and now it's around 4.2V, so now it only drops 0.05V every 24 hours. Does this mean "constant resistance", and now there's so much less current flowing that it will drain slower and slower? I've tried cleaning around the acid filler ports (it's a really old battery that has these little knobs you can open to refill), but I haven't gotten around to finding out where the drain comes from. Because, when I see its discharge behavior with higher loads, I think it is still somewhat good for something (like a backup battery for an alarm system or something that needs long standby times without drawing as much as the starter of an engine would), if just there wasn't that annoying drain!
Test with power levels like those from wall quartz clocks. In those an AA cell lasts normally for months. Must be at about 1mW or so.
The BK precision meter is using a PID algorithm for constant power output. PID terms are tricky to optimize for all conditions. Not enough derivative term and you get oscillation as is shown by this meter's response. Of course there's no energy here which is why the companies don't show
John Borchers Really? How do you know? Doesn't really make sense to me using a PID, because of the nonlinearities of the 1/x function here.I guess it is simply sampling the voltage at some interval and then setting its classic linear loop current source to the calculation I=P/V.This sampling interval and some group delay from the voltage to the set current combined with the battery impedance/equivalent circuit gives the oscillations, i guess.Maybe Dave can check the BK by giving slow AC lower than 1Hz (a function generator could maybe give 50mA) to the load in constant power mode and measure the current with a scope. I think you will see some aliasing on this current.Might be very insightful to find out how it actually is behaving in constant P mode, as the manufacturers don't give much information.
Raymund Hofmann Interesting thought. I've done it before in my own constant power loads as you explain.
Raymund Hofmann Because it appears it's overshooting I=P/V. There are also messages in the forum regarding this meter. I stand by the claim it absolutely has a PID control algorithm.
You should actually get a batterizer when they are available on the market (although it means giving them money) and make a video showing how the converter craps itself just a few seconds after the battery alone (or even earlier due to the loss in the converter).
And showing how battery gages dont manage to shut down devices sefely before the battery is empty or how the thing shorts the battery out and start a fire because it piered the protective foil of the battery (which is a much stronger point than the performance to me).
Róbert Valdimarsson If they stood behind their product, Dave would already have gotten a couple to test for free.
Hi, I work in a primary school. The teachers use digital cameras for part of the children's
course work. The teachers routinely require new battery's as the don't last long. I have released that the camera's have a very high cutoff point and now when they request new battery's i ask for the old ones because the still have enough energy for IR remote's which last a long time. I think digital cameras are either poorly designed or we're using the wrong batteries. I only use Duracell ultra power.
hue mondesir Your problem is the use of Alkaline batteries for cameras, that's bad! You need NiMh rechargables as they have much lower ESR for higher peak currents.
hue mondesir
No, you're not using the wrong batteries. What's wrong is that you use batteries at all! The only real advantage of alkalines over rechargeables is their low self discharge, so you can put it in a device that has to run for month or years. A camera does not need that.
EEVblog Thanks for the reply Dave, but have you ever worked with teachers, they may be able to teach children but when it comes to technology that's another story. I couldn't imagine how many rechargeable battery's would get discharged to the point of no return or just get thrown away. we have 30+ cameras and it's driving me crazy with the amount of batteries we go through.
Hmm, did you ever figure out if it was the power supply or if it was dielectric recovery of the battery?
Hey good job with the echo cancellation panels.
But I wanted to see what happens with the setup if you go down to 0V on your power supply. Infinite current? The load stops (more likely...sad)?
I have a "1,5V led flasher" 3 transistor circuit that has been operating on under 0,8V battery for 24/7 over a year now.
DjResR is that in bee's dick years?
thekaduu Nope. Human size years. And it's still running.
Good video Dave. Maybe you can come up with a precise explanation for, or eliminate, the noise/oscillation that the load is reporting after the brick wall response. Of course it's gibberish and not actual power remaining in the battery, but you know the nay-sayers will point to that and go "Look! That's the energy we're tapping into and you discounted it!". Because unfortunately it *looks* like more area under the curve, so it has to be explained to cover all bases.
Great video! But I can't seem to find part 2. Where can I find the continuation of this? 🙂
hey dave, could you take a few samples of the ESR as the batteries discharge? ESR is more important that that "extra energy" left as it will tell whether you can get it in a bucket or drop by drop.
It's not about how much energy is left below .8 volts because most electronic devices ,which is prob 99% + , that are on the market won't operate much below .8V due to p-n junctions.
In a very scant few devices that can operate below .8V , the time left is ridiculously low for a battery that started out as 1.5 V under load.
WisdomVendor1 Yes, that's the whole point of the video, to find out how much that "ridiculously low" time left actually is. if you can data elsewhere for it please add a link.
It would be nice to atleast sneek in a wang-ho-low/noname brand of batterys and see a discharge curve on those to see if there is more power left after that 0.8 limit compared to some of the more known and proven brands like duracell.
It would be nice to see whether the rated energy storage (1000 mAh?) matches the actual energy based on the measurements you take. Did the thing really leak on the shelf?
Dave, why constant power mode, couldn't you use constant current, and also just log the voltage?
You've got a lot of faith in the display on that electronic load... You should also be measuring the amperage independently to verify that the BK Precision is telling the truth...
icesoft1 well, he's power supply did it. And it said it was spot on. The problem is in the software, the load isnt analog, so if it isnt fast enough that can happen. It had the same problem with constant resistance mode which he showed a while ago.
Doh... I misread the display on his power supply during the test.
Guess I just wasn't getting any warm and fuzzies from the hysteresis loop in the battery test and the data acquisition software...
Laharl Krichevskoy Oh yeah, the "power supply killer" constant resistance mode! Had forgotten about that.
What HP benchmeter did you mean at 5:19? It's Keysight. If they change their name some more time, I need to make a list to remember that it is HP :-)
Is there a part 2 of this?
Sounded like this would be a longer series, but I can't find the second part.
Or are "Alkaline Battery Discharge Testing Part 2 + 3" from the second channel the sequel to this, although they came 5 years later?
You might try some different battery chemistries like carbon-zinc, lithium, and the alkaline that you're already running. Would be interesting to see how much power one could squeeze out of the other chemistries (like if there were a better chemistry based on the type of load).
icesoft1 carbon-zinc are crap anyways.
lithium i'm not sure, but li-ion and li-poly can catch fire if over discharged. Not sure about the non rechargable lithim ones. They do have lithium metal inside so unles they do not, there's always the posibility for lithium to catch fire (just exposure to air does the trick)
if anything, it would be more interesting to meassure the ESR at different charge levels as it would actually tell you when you really cant do shit with it, even if there's still some small percentage of "energy" left
icesoft1 The Lithium AA/AAA's would be interesting. Yes, I might try those after Alkaline.
This is great , thanks dave
when Dave was referring to his "big,massive" cables, I was like "take a look at my 95mm sq" ones attached to test rig for a controller-motor testing. for you US boys its 0000 cables.
Speaking of batteries what about a tear apart apart of a ctek 12v Battery charger it's also never be done on TH-cam
I would love to see you test those batteriser and destroy their claims some more.
Every Duracell I've used in recent years has leaked. Hope yours don't.
Dave would you consider doing a video on calculating Thevenin equivalent circuits?
I like having the TH-cam captions on for people with accents.
eg. 16:38 "The constant power mode" = "The constant Pelham high road"
"Some sort of oscillation" = "Some soda australasian"
LOL XD Asian Aussie hybrid
SNEAKxxATTACK TH-cam translate is a laugh a minute!
Dave, you are so right about the BK load; it's a great instrument, but the software SUCKS! I had to write my own LabVIEW drivers.
Also, the isolated RS-232 cable seems to have very poor common-mode rejection, or at least mine does. You should check that. It could make a very interesting video.
This reminds me of something that happened to my sister a couple of years ago. She called me up saying her Home air conditioning system was not working, The conversation went somethinglike this.
My sister, My A/C stopped working, what do i need to check, and how do i check it.
Me, do you have a digital thermostat.
my sister, what is that.
Me, does it have a display that shows the temperature on it.
My sister, yes..
Me, check the batteries.
My sister, Does it have batteries. I've lived here 14 years and never knew it had batteries.
She goes and checks the batteries. Sure enough they were dead. After changing the batteries; the air-conditioning system worked fine.
My sister, Who knew the batteries would last 14 years.
PS, Before you ask, no I do not know what kind of batteries they were. and yes this really happened.
What camera is used for shooting of this video by the author? Good sharp image.
does the discharge rate above .8 volts make a difference to the behavior under .8 volts?
A single battery-test should not exceed a month or maybe two. Otherwise you may get a problem if your USB stick is formatted in a FAT32 filesystem (4 GB maximum filesize).
Luncher91 maybe 2 months??? 4GB = 2^32 bytes. 2 months = 86400 minutes... Assuming that a single value (V) is being written into the CSV file including the comma is 7 bytes (as in "#.####,"), it should take about more than 1145 years to reach the 4GB file limit... What number are you using???
sorry i dont know what i have calculated. you are absolutly right. i think i divided by 60 one times too often. :-D But i assumed a measurement per second. Correct would be for 7 bytes/second : 2^32 bytes / (7 bytes/second) / (60 seconds/minute) / (60 minutes/hour) / (24 hours/day) / (365 days/year) = 19.46 years
So i no problems by the amount of data. Sorry for that absolutly incompetent post!
Hopefully Duracell sorted out their exploding battery issues, wouldn't want one of them to go off in your lab... :P
Hi Dave: How about putting a cap (+/- a resistor in series) to damp the oscillation at the end. Initial thought is that it should not effect the curve other than to smooth the end. Yes/No? Cheers, Mark
Mark Beeunas I think this capacitor will only change the frequency of the oscillation, because the equivalent circuit of the battery, as the recovering of its voltage suggests, is already some kind of RC circuit after the "big capacitor" ie the chemical engergy source.
MichaelKingsfordGray Could be something like that. Would be interesting to know the exact chemistry happening...
MichaelKingsfordGray
You say, "I suspect · · · " Prefect reason to test.
Raymund Hofmann
You say, "I think · · · " Prefect reason to test.
MichaelKingsfordGray
Nope, just cannot -speal-, -speel-, -sphell-, anyway.
Cheers,
Mark
**************************
I have a 50ah sealed lead acid (not a Car one) charged it is 12.8v what is min voltage Can i go to with out damaging it ie: 10.5v or 11.5v open curcuit volt, when i put my camp fridge on the battery volt drops from 12.8v to 12.1v while the load is on. hope explained it
well, why not just drop its energy with a 1 volt zener and current limit then start to log with high density readings a 100mW or just a 10mW that will save your day.
though I bet a battery can't survive a .5Amp when its voltage drop, it cant maintain that rising current.
I assume Batteriser will send you a sample so you can do this test and prove them wrong ;)
any chance you could do some vids of LiPo vs LiFe batteries as these are getting so popular and happen to be so dangerous, more often than not when hit with a hammer and thrown in a microwave. No that's been done , just like to know the more indepth tech info on the charge numbers especially for LiFe as there's so much opinion out there and it seems the manufactures don't always give the info.
BK Precision? So basically Burger King precision?
HTFCirno2000 British Knights precision
What are those triangular wave tiles in the background?
Some acoustic stuff?
hey Dave @EEVblog, where did you get the spade to binding post connectors?
Very interesting, would you get better voltage drop precision if you had super cooled wires?
Sean McMunn
Better wires would clean up the measurement at the really low end, sure. But what real life products have anything better than what he is using in this video? I think that, in a way, makes this result even more accurate.
Sean McMunn
At low currents the wires are irrelevant. At 10mW constant power you only get higher currents (say 1A+) at the very end (below 10mV).
That depends Sean, are the crystals aligned?
ArtificialDuality Ya that's true, real life examples > theoretical testing. Cant wait to see the results of the experiment
Pook365 Yeah, I heard it depends on if you're using snake oil!
"nice and easy test" (10k yankee dollars worth of equipment on screen)
666Tomato666 easy != cheap
KX36 that's what I was pointing at, having good tools (which most often means expensive, especially with electronics) is what makes stuff like this easy
oh yeah, i see that now. Funny how reading things with different emphasis in your head can change the meaning.
666Tomato666 There is no need at all for using high dollar equipment !! It is just showoff + the man dont have cheap equipment :-)
666Tomato666 So use any data logger you like. Build your own constant power load. You don't have the use the same gear I used.
What about the temperature? How much of an effect would that have at very low loads?
Ben Barbour The lab is about +/-2degC, so not a big deal.
Do you happen to have videos on the 8614?
Lol at 4:56: you show what looks a complete list of commands (with the packet structure) that your BK Precision supports and... you immediately abandon that idea because... it's too hexadecimal? Too well documented? ...C'mon now Dave/Dave2, sorting that out looks like 20 minutes of gentle Python coding while sipping a cup of tea... :-)
Or... 2 mins of hooking up a bench meter that logs.... I'd have done the same as Dave.
Hmm... I wonder why you are building a battery test rig. Bateriser
Let's hope there will be no powerloss on mains. Some UPS should also be a part of this testing, otherwise you testing Murphy's imagination.
Got any results from this yet? Though, you can fly to the moon with 100mW. Imagine some AVR monitoring a sensor every few seconds and sleeping (0.1µA) the rest of the time, maybe blinking an LED now and then... What would that take, 1mW on average?
Are you sure the packet structure isn't just Modbus? The Maynuo versions just use Modbus over RS232. I wrote a bit of python code that reads out values (never got around to writing the rest that actually sets values). github.com/MetalPhreak/pyMaynuo
Yay, not nice to try confront the world with some scientific facts. But one thing got me wondered, under 0.6V are you even able to draw 100mW out of it? I would guess there is a point where you can't get it anymore but there is still some voltage on it... that point would be intresting, also like a curve V vs. Power you can still draw from it.
This was a science project back in high school. LOL
Dave, why you using 9999999999$ ultra precision equipment(that also dont works well for this test) instead of using Accucel 6 charger that is designed to measure batteries, It also have uart>usb connection capability So we can get graphs on PC!
Arek R. What I have works great. Does the Accucel 6 even do data logging to hundreds of thousands of samples? I don't want a PC running for weeks just logging. Also, does it do constant power down to 1mW resolution?
EEVblog No it dont, but does we need very precision data for just ordinary batteries?
And I also was thinking about my accucel 6(whoops) that have 1mV(oversampling) read capability, so, you are right about acc6 with standard FW....
I toss AAs & AAAs if they read less than 1v no load. At least I haven't been throwing out useful power all this time.
James Lamb Nope. At 1V open circuit it's pretty much good for nothing.
How much power does say, an IR remote control, use when active?
sleepib the led draws ~20mA and UV leds are around 2V, so about 40mW on the led, plus the rest of the logic/asic. It probably will work up down to 1.75V (2x 0.9V) So a remote most probably wont work below 0.8V anyway, and if it does the led just wont turn on.
IR maybe gets away with 1.3V (red leds are about 1.5v) so maybe around 20mW, still you cant get many mA at such low voltages, remember ESR goes up, (the brick wall you saw in the graph) and below 0.8V it can easily go above 500ohms or more. You wont get 800µW with 500ohm and 0.8V.
In both cases most probably it wont work below 0.8V
Laharl Krichevskoy The LEDs of IR remotes are driven with a much higher current than just 20 mA, more like up to 1 A; but of course pulsed with a very low duty cycle, and using a beefy enough capacitor to deliver the pulse current.
Jo Pe still you cant go above the max rms power, else you burn it. The point is you cant even get 1/10 of those 20mW. last time i did the math it was like 800 something µW, you cant do shit with that, plus with no voltage you wont be able to charge that cap to get that pulse
Laharl Krichevskoy Mainly, I'm thinking a basic TV remote control would be an ideal use case for a DC-DC converter, provided your converter's output is regulated to just above the minimum operating voltage. If we assume that's 1.3v, and that you're using 2 AA batteries, A DC-DC converter should be a big improvement in battery life, not because of the lower cutoff voltage, but because you're not dropping almost half your starting voltage in a linear regulator. I think we also need some practical testing of how bright the led actually needs to be, and how long the pulses need to be in order for the receiver to detect it. If we're running the LED at 10x higher power than we need to, that's just more wasted battery life.
sleepib i doubt they would put a DC-DC inside one of those cheap chinese remotes. Also dont thing it will make it last longuer as you maybe will be able to get 10% more out of the battery but you also lose more than 10% (eff is usually around 70-80%).
plus dc-dc used chips, diodes, caps and inductors. It's really not worth the cost.
It really doesnt make a lot of sense in trying to be greener with a remote that lasts more than a year with a pair of AA. It would be better to just use rechargeable batteries (the already charged nimh ones) or stick a lipo battery and forget about it
Please, nobody ask Mathew on VSauce how much energy a battery has.. lol
Green Silver Why is that?
EEVblog
I'd just watched his latest video, it all started to get confusing for me when he explained there was in infinite amount of numbers between 1 and 2. Certainly food for thought that I hadn't considered before..
When it comes to a battery, I should imagine I'd be equally confused by his complex views on it. Secretly I'm now wondering if your Duracell has an infinite amount of energy even if if was unmeasurable. lol 😳
we all know that you can get a few weeks MORE out of dead batterys in the remote control things like that use virtualy no power
Batteriser Batteroo 1 month ago
+dismayer666 Dave jones uses a power supply box to debunk the product, which completely ignores ESR and internal resistance of a battery. Don't believe everything you hear! Ive personally seen the batteriser work. Dave jones is a dramatic arts school graduate. The inventors of Batteriser are pHDs of electrical engineering and also hold over 700+ patents being used today by billion dollar companies in the sillicon valley, California.
***** They are wanting a war by the looks. I think they are scared about you trying to unfoil their scam!! Keep it up Dave and guys, spread the word.
***** They seem to have their Daves confused ;-)
"It's hard to get dater on this stuff!" Mwahahahaha.
Is there a life under 0.8V?
Amun-Ra Yes, but it's inter-dimensional aliens.
EEVblog It's quite clear the batteriser is over-unity.
What are you doing to make money in your lab?
infowarrior420 he is making these videos
infowarrior420 You just did it for me by watching this video, thanks!
Thumbnail broke for me..
TheSpeedkarter | Road To 5k Subs Yep, work experience kiddies let lose at TH-cam again it seems.
Dave, Please put a sticker over that stupid keysite label :o)
zx8401ztv I recon there has to be a big market for an old school HP logo sticker...
EEVblog I just dont understand the stupidity of well known companys, you have a winning name that everyone knows as good quality, oow lets change it for no reason at all, silly buggers lol.
MichaelKingsfordGray You hit the nail on the head with your last sentence,
"their marketing execs had to justify their existence".
yep :-D
"Usable" voltage . My kitchen analogue clock is usually the recipient of any .08v batteries . Using half a bee's dick, It will get a extra week of usable life and keep time from a AA battery that wood otherwise be rubbish at .08v cause not much use for other devices . Some l.e.d torches I have won't turn on under 1.0 v ..
dfsilversurfer A week extra life isn't much. Probably more annoying to keep changing them than sucking out every last drop. But each to their own.
It's not a legit test unless you are using Durcacell bunnies.
302!
Dave plase make a video commenting the latest Batteriser video it is sooooooooo funny trust me ;)
:)!Coppahtop
301
109th!
Jake Dancel 172nd :)
This isnt the channel for that bs
TDJDriftersBoss lol
hompalai come on have a little fun, I could have said 4th ;)
TDJDriftersBoss ...... :)
.666 Amps