EcoFlow Delta Pro Battery Capacity TESTED
Вставка
- Опубліковано 31 лип 2024
- Is the EcoFlow Delta Pro battery really 3.6kWh capacity? Well, it all has to do with output referred capacity, let's test it and the linearity of the percentage capacity display.
00:00 - EcoFlow Delta Pro solar storage battery
02:09 - It can barely power a single fridge for 24hrs, what's going on?
05:56 - DaveCAD efficiency loss curve
08:34 - The 1kW load test
10:04 - We've hit 1%, what will it do?
12:51 - Well, it's certainly non-linear! It needs a proper Coulomb counter
19:22 - Oh come on, seriously!
20:09 - Last gasp juice
If you find my videos useful you may consider supporting the EEVblog on Patreon: / eevblog
Web Site: www.eevblog.com
Main Channel: / eevblog
EEVdiscover: / eevdiscover
AliExpress Affiliate: s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
T-Shirts: teespring.com/stores/eevblog
#ElectronicsCreators #EcoFlow #Battery - Наука та технологія
UPDATE: it turns out that you can recalibrate the BMS (which may contain a coulomb counter?) by doing three full 0-100 charge-discharge cycles. This is not mentioned in the manual and is buried away in the FAQ on the website and is barely explained at all. This is poor form, it should be very clearly explained in the manual.
UPDATE2: After two complete 0-100% charge discharge cycles (and turning it off after each cycle) the capacity display appears to be correct now. It only stays on the 1% number for a minute or two. So no more non-linearity at the lower end. Whether or not it stays this way remains to be seen.
the 1 in the display showing the SOC is quite line(ar) or simply a straight line.
I want one but only with the behaviour during the last 1%
Do two more 100-0, then try the test again to check the BMS behavior since re-calibration. Ecoflow owners will love to know this.
Make sure to leave it Off for about a half hour after each full charge. That is important.
I wanted to comment about the fact that it sure has a coulomb counter just that it was not programed with the correct battery capacity but it seems it requires "learning". I designed a BMS and mine requires the user to introduce the real battery capacity for accurate SOC as I do not find "learning" is a good strategy and can result in large errors.
Can you do another test to see if it now "learned" the battery capacity?
There should be no damage if the BMS disconnects at 2.8V
3.3kWh seems about right as modern inverters are around 90 to 92% efficient. So battery is sure 3.6kWh as advertised. The 12V output is through a DC-DC converter maybe 93 to 96% efficient so it will not be that much more efficient than the inverter with 1kW load.
That info really should be readily known and they absolutely should be calibrated from factory as it's a consumer product.
I've been playing with BMS systems lately using a Texas Instruments EV2300 which allows me to read TI BMS chips using the SMBUS. A lot of laptops use the same BMS chip and I wanted to read / edit info on them.
The recommended way they recalibrate the counter is charge to full, leave for 2 hours, then discharge to 0%, leave for 5 hours. Then recharge. Not sure if that changes depending on chemistry or not but I assume that's to allow the charge level to stabilise.
One of the advantages of the freezer is that the door is at the top. When opened, not very much of the cold air spills out. In a regular fridge with a front door, every time you open the door, some cold air spills out and is replaced by warmer air from the top.
Yes, this is why the have supermarket freezers with no lids on them, and they are still fairly efficient.
@@EEVblog2 One thing I do to help reduce air changes is to keep the fridge full, even if it just means putting empty boxes or air-filled bags in to take up the space.
As for supermarkets, several round here have open-front chilled shelf units for the cans and bottles, Some put flaps over when the shop is closed but they must be pretty wasteful regardless. This is despite the now extortionate electricity prices, typically 35 pence per KWh.
@@BG101UK I am pretty sure density or "weight" helps a standard refrigerator run less often when a door is opened.
The cooled items act like batteries. For example if you had a power outage if your fridge or freezer was full of water jugs it would stay cold for days, where if it was empty boxes it would be warm in less than a day, especially if the door was opened.
Cooled density acts as a battery, but neither saves energy long term, thermal loss is always the same.
@@volvo09 I do agree with that, however when it does get re-stocked there is more loss if you're having to remove e.g. water containers which have a higher thermal capacity (I forget the proper term for that).
The thing with the boxes is at least it stops ALL the cold air being replaced when the door is open.
Unfortunately now there are 2 fridges and a freezer in use (plus another old fridge I kept for backup), there isn't roon for a chest freezer etc.. Anyone got a cheap mansion for sale? hehe
@@BG101UK I get your point now, I wasn't considering a walk in freezer or cooler where all the air falls out each time the door is open.
Thanks for putting this up Dave, I was contemplating one of them myself.
There are cheaper solutions. Depends on your needs. If you don't need a portable solution then there are better value and more flexible options.
2:32 - if it's a frost-free type, it might be the heating cycle overnight that's chewing the power.
Have mine, love it
The sound of your Miele washing machine beeping to say it had finished nearly got me out of bed when watching the video to go downstairs and empty my washing machine! Even though I hadn't put it on.
I once came upon a fridge which had a small door and compartment in the main door, in this compartment was room for milk butter and soft drinks.. which meant that the cold air in the fridge didn't fall out every time some one wanted a drink. . . Much more efficient!
@Dave, don't forget the fridge likely has a defrost cycle that runs an electric resistance heater probably about once a week. This can be a huge energy consumer.
Most will defrost a lot more often than that.
Mine defrosts at 400+ watts every 8 ish hours depending on usage.
You test is a calibrating of the device --- resetting the BMS is aligning , at least that is how I thought I understood measuring & resetting / calibrate & aligning ?
Thanks again
I held the on/off button for 20 seconds, I gained the missing 2000 KWh I’ve been missing. I have dp plus one battery. I was getting roughly 1/2 of the rates KWh now instantly getting 6.9Kwh. Thanks. Big difference!
I have a US 120V model, I've noticed some interesting things with mine since I use it in "UPS" mode most of the time. Every few weeks, it will drain the battery down to 80% regardless of the charge setting, and then hold there for about 12 hours before recharging to the set charge level ( I usually keep mine at 20-95%). When recharging to 100%, I have also noticed it stays at 99.49% for quite a while, I assumed this was top end cell balancing since the input wattage is not at maximum.
Impressive!!
The problem with coulomb counting is that you are integrating charge over time. Any measurement errors or variation in battery charge/discharge curve will stack up over time. If you never charge to 100% or discharge to 0% the counter never has an opportunity to reset those errors. An occasional full charge-discharge cycle is needed to keep the coulomb counter in sync.
The very long time to reach full charge hints that the cells had drifted out of balance, which would also throw off a coulomb counting algorithm. Again an occasional 100% charge helps keep your battery in balance.
Totally true.
Measurement errors and also charge efficiency. The charge is not 100% efficient. I wonder if they add the charge efficiency into the counting.
It is most probably just like power banks, the "indicated" capacity is the total absolute maximum capacity of the battery pack used in there. After the conversion losses, the actual output capacity will be around 2/3 of the rating advertised.
please count the coulombs going in to fully charged state too .. Also it would be interesting to repeat the discharge with a maximum load.
I have a Tesla with an LFP battery and I did some reading on LFP. Due to the voltage stability when it isn't full or empty BMS's struggle to estimate the charge in the middle ranges The benefit of LFP is that because of the lower voltage the cells are less stressed and have a much higher cycle count. Tesla recommend charging to 100% each week to help ensure the guessometer has a chance to reset.
This is the Ria N2RJ YT / Dave YT channels cross over content I'm here for!
You make a very good point on what 3.6 kWh actually means. It depends on the load, how you use it (AC vs DC), as well as the temperature of the battery. There should be some sort of government requirement to show the actual expected output, kilowatt hours as well as showing how temperature, load, etc. will affect the entire usable capacity of the battery.
All battery manufacture should be required to list the expected kilowatt hour capacity, expected input, and expected output, values, rather than the internal storage value of the batteries, because the user doesn't care about anything other than how much usage can be obtained out of the battery and how much it cost to charge it.
Just saw you mentioned it in the video. Never mind.
I think it's the battery that's doing to drop in efficiency as load increases more so than the inverter. More current out of the battery means more IR loss in the cells.
I have one in NZ , love it , arc welding smoothly at 110 Amps , hasn't tripped I was impressed , can't remember how many watts it went up too
Just a thought Dave, What's the power factor of your fridge?
I have an Emporia energy monitor in my electrical board that's always showing weird and wacky things when I have inductive loads on those circuits.
Would be nice to see update on it if you calibrate it
You can use like full ups for the fridge?
As it's LiFePo4, then the balancing is best at 95% charge or higher, as the charge/discharge is too linear to get an accurate SOC. Does the manual state that it degrades slower at 90% max charge? LiFePo4 is happy to charge to 100% down to 10%
Most laptop computers do this too. Occasionally running a laptop flat then allowing it to completely recharge helps to recalibrate the charge circuitry.
Dave, Have you considered a lead acid tubular battery based inverter system for your requirement. Not sure about the price/certifications in Australia.
Cool, what was that hissing sound, I was lookin around my kitchen thinking I had a leak till I paused the video.
What are those bugs called making all that noise in the background?
it's also known for LFP batteries in the new Teslas, they recommend 100% charging at least once a week, so if you always kept it at 90% it's highly possible the BMS is messed up and doing a full charge/discharge will help figuring it out it's real battery %
Man, I want one of these! On the fence to buy a max and additional battery or just this pro model
Why do you only load it till noon? Staggered loading of other devices?
I wonder how many times the compressor has to start up in a 24hr period. Could be a large total draw over that time period.
Wow over 20% remaining after it said 1% !
At least it wasn't the other way around.
My stand up freezer draws just over 150VA and cycles several times a day.
Ah, wish i knew this last year during the hurricane when we were rationing our Delta Pro's battery (didn't have a way to charge it in a power outage situation). Make it last 3-ish days.
Arnt most good inverters in 90+++% efficient , at least most of the spec.s I've seen in good brands ?
I assume there will be a low voltage cutout on the battery. If it's within the specs on the battery it should be okay. At low SOC the voltage falls off a cliff.
LiFePO4 are weird to charge in that there is a fairly wide voltage range that will get them to near 100% but they stay in cv longer at the end of charge when using lower voltage. No idea what sort of wizardry you'd need to keep cells balanced without going to full charge
Is it 50 cycle 240 volts
My company uses LFP batteries in their devices, and I can confirm that when it comes to calculate measure energy left, they are pain in the ***. In the 30-70% energy area the voltage change of 1mV is about 5% change in energy left.
It would be interesting to determine the efficiency of the inverter, using a fixed load.
That would be useless cause inverter efficiency is not constant and vary with the output, something like 60% at very low power (
@@davidniquot6423 Specifically, I would be interested in the difference between 12vdc and 120vac using a 72w load. Of course, the resistance would be different for the two different voltages to keep the wattage the same, Ohm's Law.
I'm pretty sure it has some sort of battery fuel gauge chip. A simple voltage measurement would never drift, the relation between voltage and state of charge, while non linear, doesn't really drift much and certainly wouldn't need calibration. You'd also see it change with varying loads due to voltage sag.
Coulomb counting would however drift since you're not gonna be able to measure it 100% accurately and there's spikes that you'd miss. Getting to absolute low and high would fix that kind of thing.
I'd love to see how this behaves with a 3kW load.
Tesla recomends charging to 100% at least once a week with LFP batteries. So I think this also applies to this unit.
Don't worry about your chest freezer loosing power for 5 minutes. These things can hold their temperature quite long. I think the most are rated for at least 18 or 24 hours bevor they thaw. You don't want to go near thawing thou. Set it for a lower temperature, if you're are anxious for testing
I've tested mine recently to see if I can get away with running the thing full blast during the day with solar and shutting it off completely during the night. Turns out I can. It got from -30°C to -17°C in 16 hours. No Battery required
And the worst thing that can happen to your fridge is a warm beer in the morning
I had a little chuckle at the display showing only 6 days when the input and output were identical. The remaining days should have been forever........
APC devices actually require a calibration run to be accurate on the expected time left.
Do it again at 2kW and you'll get another datapoint on the efficiency.
What if they did it to preserve the battery. If you discharge 80% you get a lot more energy in total (cycles * cycle-energy). Or maybe you are right if you do full cycle and it's using coulomb counter it might be calibrated by a full cycle. It'll be interesting to get a second discharge cycle if the capacity indicator changes.
P.S. about efficiency - remember the calculator you commented on. Internal battery resistance increases the losses. I think you might get more Wh out of 100-200W load than 1000W. Anyway I like energy/efficiency measurement -thumbs up!
Oh, just read the update...
At least Twitter support seem to be very helpful. Seems like they messed up the settings for there SOC IC. The Twitter support even told you why you needed a 100 to 0 cycle.
I ran 120w on a 2000w invert and got much longer run time with a 150w inverter
It would be interesting to check the inverter behavior depending on the power factor. With bad ones I bet the losses will be higher and the efficiency will drop.
Batteries only provide VAs, so a load with a 0.5 power factor doubles your battery draw compared to real watts a kill-a-watt meter shows you.
@@taskelam8937 Not necessarily. When the voltage and current are of opposite sign you are putting back energy into the batteries.
@@taskelam8937 Check ua-cam.com/video/kp4lGfLvoTc/v-deo.html
90% top charge mean you also don't get you're pack balanced.. right?
Could explain why you got stuck at 99.49% -> BMS actually doing it's BMS job.
Funny how I was thinking my 3$ aliexpress coulombmeter I use on my solar powered house was crap.
Finally, I like my 3$ coulombmeter :) It get about 1-2% off after a day or two of not reaching my "100%" charge
The BMS of that thing probably cut out when one cell reach the minimum safe voltage. Not great for the battery but it won't damage it by say. Battery capacity will drop to 80% after 4000 cycles instead of 6000 cycles or something like that (if you discharge it to "zero" very often).
Same if you charge it to 100%, it just lower the amount of cycle needed to degrade the pack but still should do 80% after 4000 cycle.
Note that you can have 99% that is the same as 90% in term of longevity. What kill lifepo4 with time, is the voltage. If you cutoff the charging voltage to ~3.45v per cell after 90% charge instead of letting it climp up to 3.55-3.6v, then you will prolong the life of the battery pack while still charging it to 99.92354876%, But it will take a lot longer to charge the last ~5%.
Another edit : If one set the max charging voltage to a lower voltage instead of the max 3.6v, he have to look at the spec of the BMS to know at what voltage the BMS start to do his job (balancing) as if you never reach that voltage (like if you only charge to 90%), then the BMS will never balance the cells and they will get widely unbalaced with time. Mine is ~3.5v
I power metered my samsung fridge and it came in at around 1.6kWh/day in Canadian spring time.
I think that's what mine was last time I measured it. Have started a new long term measurement this morning.
Hmmm, for $6500AUD and 5yr ltd warranty, not sure if its worth it.
Honestly surprised Dave doesn't just make his own battery system (shitload cheaper) with maybe a Pii to control/monitor it. Given his knowledge of electronics.
yes please repeat the test
Wouldn't the USB-C Power Delivery with higher 20V output be more efficient than the 12V car power output? I know the car output is rated higher wattage, but the stepdown from battery voltage to 20V is less in my mind. Only thing making me think the 12V might be the same or more efficient would be less complex circuitry.
Maybe, but I'd still expect the 12V output to be the most efficient I think.
I bought a small glass fronted fridge to keep our most used items in: milk, butter, drinks etc. Stops us opening the main fridge too frequently.
Wouldn't work physically in our kitchen
@@EEVblog2 Same. We put it in a room next to the kitchen.
@@SimonCoates We have another fridge outside in the garage but not nearly as convenient.
It may be balancing when at 99.5% hence why it stops charging quickly
Do you think it might be too close to the warm exhaust from the fridge?
To a family you test put Bootle with water in all the fridge and down the temp to 6/7 Celsius fridge I made that to down the consume of the fridge
Some of my USB power banks loiter at 1% for ages!
Damn. Had my money at 3.3k. Guess this is why I don’t place bets…. 😅💸
This is typical for smartphones too. Those who never drain battery to 0% and auto turn off, or charge it to 100 %. Just charge phone ten times a day by a few percent, this upset charge circuitry after some time. As far as I know most phones don't calculate battery charge by voltage. Just reset counter on 100% charge and guessing remaining percentages by measuring consumption in mA.
Yeah and LFP is even worse than LiPo for SOC. They usually have some algorithm that used voltager together with discharge current etc. to estimate SOC.
Now, how much energy does it take to charege it to 100%?
3600Wh + (unknown) charger efficiency.
You should log that as well to get total system efficiency.
Gotta say it would be nice if all manufacturers would publish all of the technical data that they have on all of their products. But sadly, when a product is deemed to be a consumer product, most technical data is withheld from the public. In all of the companies that I've worked for, the claim is that they do this for 2 reasons. 1) To alleviate the possibility of legal actions, if the product doesn't quite match up with the released technical data. 2) To prevent the competition from getting insights on the design and construction of the product, thus making any and all copyright legal actions more likely to succeed.
I'm sure you could get a similar product from an industrial manufacturer and then you would get all of the technical data, the problem is that the product will likely be about 10 times the price.
Climatic Change ... Running a heater at broad sunlight
Battery was charged from solar, so not really a big deal…
@@danabc322 but the heat is still coming out to environment 😉🤡
@@oreste6076 I’m not sure 1kWh of heat is going to do much in an Australian summer, especially since it’s being vented outside so no extra AC is needed to cool.
so it tries to keep that low discharge a regular thing by indicating it is 1%/ empty.. and you got spare juice in the tank. I rather have this in the wild than the opposite.
It will learn all it's capacity after you flatten it then recharge it
Are your results similar to other reviewers on UA-cam, like Will Prowse? Probably 20% loss to the inverter is not unusual.
Google: Is EcoFlow a Chinese company?
"EcoFlow is one of China's most recent unicorn companies."
% indicators on most battery powered devices are useless it would be good to change the display to wh or kwh maybe 🤔
Have you calibrated the BMS? Charge to 100, run it down until it shuts off, then fully charge it again. Once charged, leave it turned off for about an hour and then try it again. If the BMS isn't calibrated correctly, it won't fully charge the cells but will think it did.
For a product at this high end price range you shouldn't have to calibrate the BMS, it should have a proper coulomb counter. Also, this is not mentioned in the manual at all.
@@EEVblog2 Even with a coulomb counter it would need to be calibrated once in while by a full charge and discharge anyway, a hall sensor current meter isn't perfect, there are some losses in and out of the pack (heat, cell balance etc). Sure, on a 3,6kwh battery it wont account for much maybe, but with alot of use, it needs a full discharge/charge to get a correct soc again. And since your soc went linearly down until last few % of soc it suggest it has a coulomb counter, would be hard to just measure voltage on a lifepo4 pack and get that good soc. Problem is probably they dont calibrate from factory, only program a safe cut off and let user discharge fully atleast once to get correct soc. Just guessing though. 😂
@@kruppin You could auto-calibrate based on a safe zero set voltage. Just one discharge required.
@@EEVblog2 Yeah, thats what I said? When you discharge it fully to set cut off voltage it then count how much charge goes in. They probably just dont calibrate it from factory only program a safe low capacity (or even wrong fw and therefore capacity from lower kwh model?) and count on user doing atleast a full discharge to test it in beginning anyway. 🤷 And it must have a coulomb counter, its almost impossibly hard to predict soc or have charge and/or discharge limit only based on voltage on a lifepo4 pack. A full charge now after full discharge it should know capacity well, but like I said will ofc drift few Wh in time if it wont get another full discharge, but it should be minimal with good matched cells not needing much or any balancing.
Fridges draw around 50W when the compressor is on. This should be the very ineffecient part of the curve. Test again! See you in a month.
Yes, I don't expect that great an efficiency down at sub 100W
Something is acting up in your noise cancelling in this video. Hearing a rather high pitch screeching/sizzling sound.
Or maybe it's the caps in the battery that are slowly simmering off. ;)
6:24 : No your graph is totally wrong, there is near no loss on a LFP, whatever power you ask from it (only if you are at the limit... which for a pack like this one is over 1C = 3600W), on the other hand the inverter DO NOT have a flat efficiency curve and will be the main reason of a capacity difference depending on the load.
Fascinating, not great for the price, though!
94% seems quite okey
Did this person actually breath during the rambling 22+ minute soliloquy?
You fear for fully discharging comes from bare batteries, they should not be discharge below a point. Fully discharge is the best you can do for batteries in such device, because he manufacturer knows best when to stop.
My best guess what that advice was to recalibrate: drain it fully, wait a few hours to cool, then charge in one go without anything connected but the charger.
nah, there's a batterizer in it
I swear none of these work as well as you’d think. It’s like you actually need triple.
No no it's not because those cells are LeFePo4 and that there is a problem with the BMS (LiPo, Lithium got the same kind of problem ... it's basic physics, LFP got a very low voltage variation between 15% and 90% state of charge, the only efficient way to know how much energy is left in the cells is to use a Coulombmeter (Mr Coulomb was living 40 km from my house), to know how much energy is going out or in and do the math to display the correct state of charge. BUT after a while the deviation will start to be too important, reason there is a need to do a full charge/discharge to set the limit again .. and reset the coulomb count.
For example, in my system (i'm living off grid), my shunt (which is a coulombmeter) reset the SOC (state of charge) to 100% when the battery pack is over 53.99V and the charge rate is inferior to 1.5A for more then 10 mins. And it will reset to 0% SOC when undner 48V.
Battery capacity displays based on output voltage are a bit of a joke. when putting batteries into a medical product we had to use a system which as you say records the energy going in vs the energy going out to give us the power remaining. it also kept records of voltage, current, total capacity so we can decide when the quality of the battery starts to drop off to the point where it might not power the display for about 20% longer than it specified battery run time.
Another system than used non-rechargeable batteries where you only have the fresh capacity as a starting point. we know the power drawn from known equipment operation states, so if you keep track of the number of each operation states you can calculate how much power you have taken from the battery and report battery low when you get within in a set point of the battery capacity. Of course the battery s are not cheap as your paying not only for the capacity but that they meet or beat their specification.
I think if you want a completely resistive load, breakout the heater fan to another circuit and test the heating element power only.
The energy meter says a PF of 1.00 so the fan makes no practical difference.
Yeah.. repeat the test.... AFTER you calibrate the unit.
Get yourself a proper inverter / battery system such as victron with a protected output for your important loads. With all the sun you get you should be off grid most of the time.
Not that easy with my two different solar systems and that half my roof is coming down soon for an extension. I may be without any solar for a like 9 months, unless I move one of the systems.
@@EEVblog2 Victron will work with existing ac pv to charge your battery, just needs a meter on the output so it knows what it’s producing. I personally have dc connected solar, but you can have both ac and dc at same time. If you get cheap overnight electricity you can charge overnight and run the house off battery the rest of the day.
18:58 That can only happen if the BMS is trash quality.....
Yeah, but you never know, so this is why I wouldn't recommend 0-100% charge discharge regularly.
You get what you paid for.
It costs at least $6500AUD. I wouldn't call that cheap.
@@theriddick He got it for free.
With your 1kW resistive load, which is 1/3 of the total power capacity of 3.6kW you are certainly in the 90% efficiency range (check inverter efficiency curves) => 3.6*0.9 = 3.24kWh
For a guy that is uspposed to have done a long career in electronics ... it's kind deceiving.
At least 20% below rated amount.
Why are you only charging to 90%? These are LiFepo4, so charge 100% otherwise it can’t really tell the capacity.
(LiFepo4 has a flat curve that falls off fast at the bottom and goes vertical at the top so it doesn’t know what it’s actually at unless you’re over 95% charge or below 5%. )
Also the inverter will be about 95% efficient at 25% load because it’s a transformer-less high frequency unit. At all other loads it will be significantly less and over 50% it’s in the mid-70%
I’d bet the usb c with pd at 20v 5 amps should be the most efficient because the battery is 48v which will make the dc/dc converter for the port at ids most efficient range.
If they simply used a coulomb counter you'd know exactly the state of charge. Charging to less than 100% ensures that there is no way it can accidently over charge any cells and limit their life. This is why it has the charge limit feature.
@@EEVblog2 agreed on the coulomb counter but even those are horrible in most bmses.
And the bms won’t let it go above 3.65v per cell and LiFepo4 batteries can’t be balanced below 3.42v and because of the major amps in those cells balancing with 150 ma will take forever and it won’t report it on the simplistic display so at 90% you’ll never balance the cells. (This is why Teslas with LiFepo4 instruct you to charge to 100% at all times). And even at 100% charge/discharge you’re still taking 3000 cycles+ to 80%. New cells are 6000 to 80%.
Nmc has a much more steep charge curve and can be balanced for most of the cycle these can’t be.
So 0-100% is actually better for the pack and keep it at 100% for a couple of days to get it balanced and do that once a year or so. At the very least take it to at least 95%. (Bottom is 2.5v)
(I run my entire house with Tesla charging off of LiFepo4 so I have lots of first hand experience)
@@jameshancock I found a buried reference in the website FAQ to doing three full 0-100% charge-discharge cycles ro recalibrate the BMS. It's not in the manual.
@@EEVblog2 I’d do that and hold it at 100% for quite a while to get it to balance fully.
And then charge to at least 95% going forward so it can maintain the balance.
@@EEVblog2 a colomb counter will need to figure out what 100%->0%->100% SoC would be, as mentioned elsewhere LiFePO4 aren't the granddaddy nicads, nor Li-ion batteries of phones, it has a very flat discharge curve along with a sharp knee either end of the discharge curve. Looking forward to seeing a retest post calibration! (and yes, this ought to be writ large in the manual)
I start to hate people that waste energy ..... is it that difficult to educate family NOT to open the fridge a zillion times per day ?
3.6kWh PER DAY for a fridge is complete madness... ! Your fridge is far too big for a family of 4 .. !
I'm living with 600kWh per YEAR, all included (not the car).
It's 65-75% efficient, which is DISAPPOINTING! I discovered this after noticing a spike in consumption and been figuring out where it went for weeks!
You can see the stats (at least with the smart home panel) where it tells you total charged/discharged
5:01 : There is an input of 90W and an output of 90W ... so it will last more then 6 days .. ! Of course.. ! If you want to test the remaining time, disconnect input power. You kinf look like a fool right there ..
Boy can you talk on repeat for 15 minutes.
You must be new here, welcome!
Ridiculous product oversight for the price. Totally unacceptable.
You need a voice coach. To train that shrill out of your voice. Its really bad.
OMG please rally for cheap power, not workarounds for bad policies. This is not the solution. No, this complexity is insane. The fact that you need to make videos about your dramas and effort you need to put into these complexities should be a red flag about where we are headed.