I only charge my cells to 4.00V . The actual added capacity gained by charging to the 4.20V really is not that much. If you watch the discharge voltage the cell don't even remain above 4.00V a full minute when even lightly loaded. Preserving Cell Cycle life is way better for me then the little added capacity you realize. .
Well from 4.25V til 4.00V I have already used mine for 5 hours. If I had them starting at 4.00, the voltage would have been 3.75, and thats only 0.05 V from the inverter to say "undervoltage". And It had stopped working. So sometimes you got to evaluate whats best, longer battery time or battery will last 3 months longer (in the end). When i'm working, I can't stop after just 3 hours, so for me, a longer battery time is best. And I know that the laptop I took them from, they was driven to 4.32V. So i'm not the one who's shorten their lifespan.... Its already done.
@@timm3802 LOL, if you have them SOC 4.25 by most standards your overcharged to start with. Bad Ju Ju. Runing LiIon's like that your most likely never going to get the rated Cycle life.
From my experience - 30-60mv difference between packs over the daily charge/discharge curves is excellent. The difference diverges the most as head down the discharge knee - usually 3.4v and below. In my own system, I run 30-40mv difference (between 42 packs) during most of the charge/discharge cycle and it diverges to 70mv difference at inverter off at 3.43v - just at the start of my discharge knee. Agree that this is not 100% DOD, but this 4.0 -> 3.4v is aprox 55% DOD and is expected to extend battery life by 1000(s) of cycles.
Thanks for your information. Sounds like you have a great deal of experience. Mine generally cycle much less than in this test. I barely take mine lower than 3.7 volts truth be told.
It's not disappointing at all! It's a great experiment and your hobby, I guess. In your place I would be concerned about cellblock 4 and replace some of the cells there. Then load the cells up to 4.2 Volts each just for the experiment and discharge it again. After discarging with 100 Watt you should try 400 Watt also and note the difference the load has onto the capacity. Im riding an electric bicycle with a 900 Wh battery and I'm discharging it with 15 Amps (250 Watts) on my long-range holiday trips. After 850 Wh and 3,1 Volt/cell the controller goes into safety mode and the remaining power ist for your headlights. I've a Wattmeter installed and am able to calculate my cruising range in advance. And this is my hobby -- inspired by your experiments.
Yeah, it seems like a lot of these power meters don't like the unclean output of modified sine wave aka or square wave inverters... Had two different ones go wonky after being used with non-pure sine wave inverters.
Thanks Glide. I report all the diyBMS stats into an influxdb database which is then read by Grafana. It means I can design graphs and gauges of my own design to show the information I like to look at most. Colin Hickey forked Stuart’s v3 code to add the report to influxdb but Stuart’s v4 code includes it by default I think.
Presumably you will want all the cell groups to have as near as feasible the same capacity so I presume you will want to check the 1/2 cell groups that discharged first. You can always add an extra 18650 or two to bring up the capacity.
Are you certain that those fuses broke due to mishandling and not due to small long term overloads? Because in these cases the fuse wires will actually look as if they have merely broken rather than "blown" as such. You may well have a few cells that aren't performing, and it could well be that the better cells are the ones passing more current. On the other hand, 3.6MJ is not too difficult to extract from thee or four secondhand car batteries. ;)
Fairly sure yes. I checked the voltage of the cells which were disconnected from the pack and none of them were significantly different from the pack as a whole. So I don’t believe any of them are either significantly self discharging or dead. All the breaks were at the soldering connection. For the fuse to blow via an over current situation I’d have expected that they got hot - and each soldered end is a heat sink so I’d expect the fuse to blow in the middle. That’s my theory anyway! No I’ve been trough them all and done a reasonable discharge and recharge I will pull it apart again and double check all the fuses. If I’m careful I can be fairly sure that any broken fuses would be due to an over current situation now. Cheers Xan
Your setup is closer than mine and your inverter is far more efficient, if you look at my solar monitor there are now some nice interactive graphs on the last month history page. I have an expected 6kwh of storage but the best i can get on a discharge test is 5kwh so far.
Time to check the batteries in cell 4 again since it performed much worse than the others. If cell 4 was in line with the others, you might have hit 1kwhr!
What's the full trip efficiency? For example, now that you're in a discharged state, if you put 500 watt hours in, how many can you subsequently get back out?
Hi guys that looks really good I’m brand new to this at moment I’ve got 20 110 amph 12 battery bank with 12 volt 3000wat inverter being charged at min with 400 watt wind turbine and 400 wat solor how do you work out the size ov these battery wall I would need like say new to these 18650 battery’s any help much appreciated thanks
whats the point in having a bms if it doesn't protect your cells. 2.8v is a bit low, it should of cut off way before that or before you switching the load off.
You’re absolutely right that that bit is missing on this version of the diyBMS. However the designer has just released a new version which does include the ability to control relays based on various situations. I need to upgrade :-)
Just getting into battery packs. I have played around with the TP4054 modules and they work well. What I am having trouble understanding regarding the diyBMS modules is how the modules/batteries can be placed in series. Looking at the schematic it looks like the ground of the module is connected to the negative terminal of the next higher potential battery pack and the Batt+ is just the positive terminal of the battery. Can someone explain how this works as the voltage measured is not referenced the cells negative side.
Well I tested them all and got an average of 2.5 amp hours. But I tested them at 500mA discharge current on my LiitoKala charger. I know most people test at 1 amp - so mine may give a slightly higher result to others.
Interesting to follow your lithium battery adventures. I'm still waiting for the lithium price vs capacity to reach a level I'm comfortable with. I just replaced my almost 9 year old AGM battery bank with some AGMs that are only 15 months old. Two 100AH, two 110AH, total cost of $305US. 420AH * 12 volts = 5040WH, which I never use more than half of (50%DOD) but 50% is 2520WH and there's an inline V/A/W/AH/WH meter to track instantaneous and total power usage - when it hits 210AH, it's time to powerdown. My inverter isn't as efficient as yours - just over 90% vs 96% ;-)
Lithium-Ion cell price to manufacturers is expected to drop to below $100 U/S per kWh within 2 years, and is forecast to bottom out at around $75 after that. Adam only checked his inverter efficiency at one point. It will have varied as input voltage dropped, and as it’s electronics warmed up. Perhaps he should have checked efficiency every hour or so. My guess is that it could easily have fallen to below 90% towards the end of the process. He did mention that the box felt quite warm after several hours of running.
I just ordered the diyBMS controller boards and the bms boards, now waiting 💯🤪 *what software is that you’re using on your tablet that you just showed??*
I've put the EpEver inverter GUI on my workshop PC using another usb to RS485 adapter. I already have the 4210A GUI on my indoors PC. Can I parallel the two RS 485 's to be able to look at both units on my indoors PC?
Interesting test, I like your dashboard, it looks similar to the results I was getting with node-red. Personally I'd have password protected it - it's open to the web. I'll have to see if you have made vids explaining the software setup :)
Yes I know it’s open. It’s read only without a login. The grafana dashboard is behind a reverse proxy, and protected by cloud flare. Never say never, but I think it’s reasonably well protected and if not, it’d take me about an hour to rebuild.
No I didn’t think you were. Actually I’ve been waiting for that very comment from someone because it’s exactly the sort of thing I’d say to someone else! :-)
@@AdamWelchUK Glad I'm not the only weirdo then :) If it wasn't for you and other people pioneering new ideas for putting a system together, researching and testing to the geekiest level and sharing here, I wouldn't have started down the wormhole into lifepo4 in my camper. Then there is the 3 months of testing the diesel heater that'll be the primary use for the 120AH ..... at 24W per hour. Making sure that I have enough solar to ensure I can run that heater in winter. 350W should do it....... Don't want to get cold. OTT I know
Shame the second attempt didn’t produce the 1kw, but at least it’s showing up a struggling pack. More to investigate! Btw, can you show us what your actual load this set up is powering? Is it just a few garden lights?
@@3dmixer552 It is crazy good if you ask me. I have 2 UPS's that I use for my desktop computer. Got curious and checked their efficiency. One is doing 88% and the other one is 90%.
isnt this a bit outdated technology now they have mpp inverters with built in batteries and bms..?you simple plug in the solar and away it goes..then the grid makes up the remaining power if the solar falls short or during dark hours..
Well you won't have that thing (you are talking about) for less then 60£. Thats the cost of the Epever MPPT charge controller here. + the batteries are free (salvaged from old laptops) and then a DIY BMS for 5£ each. Whats the price of your device ?
@@timm3802 your setup is cool but im not savvy enough or patient enough to build such a rig..and i love a bargain..for me paying 2k is worth the hassle..if i had that talent i would be making a business out of it..
Mine out a mixture of two types some LG some Sanyo. Both datasheets say they have a maximum charge voltage of 4.3volts. As I said though, I want to treat these nicely and don’t typically need to extract even half a kilowatt hour.
It's best to charge to as low a voltage as you can get away with based on your power needs. Charging to 80% capacity can significantly extend the number of cycles they can take.
@@keitho3133 Actually every vendor advice you to charge them full before using them for anything. If your statement was true (and in some cases it is), you would only get 80% out of new batteries.
I'm liking them BMS's were they a kit? Is there a link?
Very nice test and good results Adam - nice to see a real-world exercise and data.
Cheers.
I only charge my cells to 4.00V . The actual added capacity gained by charging to the 4.20V really is not that much. If you watch the discharge voltage the cell don't even remain above 4.00V a full minute when even lightly loaded. Preserving Cell Cycle life is way better for me then the little added capacity you realize. .
Interesting thoughts. I have some discharge graphs of single cells I might scrutinise. Cheers
Well from 4.25V til 4.00V I have already used mine for 5 hours. If I had them starting at 4.00, the voltage would have been 3.75, and thats only 0.05 V from the inverter to say "undervoltage". And It had stopped working.
So sometimes you got to evaluate whats best, longer battery time or battery will last 3 months longer (in the end).
When i'm working, I can't stop after just 3 hours, so for me, a longer battery time is best. And I know that the laptop I took them from, they was driven to 4.32V. So i'm not the one who's shorten their lifespan.... Its already done.
@@timm3802 LOL, if you have them SOC 4.25 by most standards your overcharged to start with. Bad Ju Ju. Runing LiIon's like that your most likely never going to get the rated Cycle life.
More like three years.
@@AdamWelchUK www.powerstream.com/lithium-ion-charge-voltage.htm
👍👍👍
95% is pretty good but I can't help wondering if all the loads were kept in DC and using a converter might be better.
Living in the UK I'd never bet on having sun tomorrow, it's usually less than 50/50
From my experience - 30-60mv difference between packs over the daily charge/discharge curves is excellent. The difference diverges the most as head down the discharge knee - usually 3.4v and below. In my own system, I run 30-40mv difference (between 42 packs) during most of the charge/discharge cycle and it diverges to 70mv difference at inverter off at 3.43v - just at the start of my discharge knee. Agree that this is not 100% DOD, but this 4.0 -> 3.4v is aprox 55% DOD and is expected to extend battery life by 1000(s) of cycles.
Thanks for your information. Sounds like you have a great deal of experience. Mine generally cycle much less than in this test. I barely take mine lower than 3.7 volts truth be told.
It's not disappointing at all! It's a great experiment and your hobby, I guess. In your place I would be concerned about cellblock 4 and replace some of the cells there. Then load the cells up to 4.2 Volts each just for the experiment and discharge it again. After discarging with 100 Watt you should try 400 Watt also and note the difference the load has onto the capacity.
Im riding an electric bicycle with a 900 Wh battery and I'm discharging it with 15 Amps (250 Watts) on my long-range holiday trips. After 850 Wh and 3,1 Volt/cell the controller goes into safety mode and the remaining power ist for your headlights. I've a Wattmeter installed and am able to calculate my cruising range in advance. And this is my hobby -- inspired by your experiments.
i remember using this power meter with a modified wave inverter. the power meter stopped working after that. strange!
Yeah, it seems like a lot of these power meters don't like the unclean output of modified sine wave aka or square wave inverters... Had two different ones go wonky after being used with non-pure sine wave inverters.
Your link for the clamp meter shows UT210B, isn't it a UT211B?
Adam can you tell where you got that wrist watch from ?
Video timeline 3:20
It’s an original Apple Watch. A few years old now.
I really like your videos and I have one question, what interface you use for the DIY BMS ? Because the one that Stuart did is quite different.
Thanks Glide. I report all the diyBMS stats into an influxdb database which is then read by Grafana. It means I can design graphs and gauges of my own design to show the information I like to look at most. Colin Hickey forked Stuart’s v3 code to add the report to influxdb but Stuart’s v4 code includes it by default I think.
@@AdamWelchUK thank you for the info, the interface is looking really good.
Presumably you will want all the cell groups to have as near as feasible the same capacity so I presume you will want to check the 1/2 cell groups that discharged first. You can always add an extra 18650 or two to bring up the capacity.
Yeah, I’m happy they are now all fuses are operational I’ll have to put it down to IR or overall capacity. Some more tests required.
What BMS you usin?
Are you certain that those fuses broke due to mishandling and not due to small long term overloads? Because in these cases the fuse wires will actually look as if they have merely broken rather than "blown" as such. You may well have a few cells that aren't performing, and it could well be that the better cells are the ones passing more current. On the other hand, 3.6MJ is not too difficult to extract from thee or four secondhand car batteries. ;)
Fairly sure yes. I checked the voltage of the cells which were disconnected from the pack and none of them were significantly different from the pack as a whole. So I don’t believe any of them are either significantly self discharging or dead. All the breaks were at the soldering connection. For the fuse to blow via an over current situation I’d have expected that they got hot - and each soldered end is a heat sink so I’d expect the fuse to blow in the middle. That’s my theory anyway!
No I’ve been trough them all and done a reasonable discharge and recharge I will pull it apart again and double check all the fuses. If I’m careful I can be fairly sure that any broken fuses would be due to an over current situation now. Cheers Xan
Your setup is closer than mine and your inverter is far more efficient, if you look at my solar monitor there are now some nice interactive graphs on the last month history page. I have an expected 6kwh of storage but the best i can get on a discharge test is 5kwh so far.
Cheers. I’ll take a look.
Time to check the batteries in cell 4 again since it performed much worse than the others. If cell 4 was in line with the others, you might have hit 1kwhr!
Greg B I think this should be the primary reason.
Time to put in more cells, why tease them on the edge, more amp, longer life.
Nice video as usual. Could you please provide a link to the watt meter? I want to buy one.
Two ticks...
ebay.to/2KDm2WT That's the one I used.
@@AdamWelchUK thanks Adam.
If your bms balanced on discharge you should pick up some more capacity.
It stops balancing at 4volts - so sadly I don’t think that would have helped much.
What's the full trip efficiency? For example, now that you're in a discharged state, if you put 500 watt hours in, how many can you subsequently get back out?
Hi guys that looks really good I’m brand new to this at moment I’ve got 20 110 amph 12 battery bank with 12 volt 3000wat inverter being charged at min with 400 watt wind turbine and 400 wat solor how do you work out the size ov these battery wall I would need like say new to these 18650 battery’s any help much appreciated thanks
So a 12v inverter would pull 2x as many amps to run 100W load?
Yes. That’s correct. Power = current * voltage
@@AdamWelchUK ok thanks. That's why I see people running smaller wire gauges on 24V (and higher) systems.
whats the point in having a bms if it doesn't protect your cells. 2.8v is a bit low, it should of cut off way before that or before you switching the load off.
You’re absolutely right that that bit is missing on this version of the diyBMS. However the designer has just released a new version which does include the ability to control relays based on various situations. I need to upgrade :-)
i certainly wouldn't use the clamp meter to measure amps for calculating efficiency!
Next time bud.... lol load tests are great..
Just getting into battery packs. I have played around with the TP4054 modules and they work well. What I am having trouble understanding regarding the diyBMS modules is how the modules/batteries can be placed in series. Looking at the schematic it looks like the ground of the module is connected to the negative terminal of the next higher potential battery pack and the Batt+ is just the positive terminal of the battery. Can someone explain how this works as the voltage measured is not referenced the cells negative side.
Have you got a link to that inverter, thanks.
You’ll find one in the description.
@@AdamWelchUK thank you.
What is the capacity of your 18650 cells?
Well I tested them all and got an average of 2.5 amp hours. But I tested them at 500mA discharge current on my LiitoKala charger. I know most people test at 1 amp - so mine may give a slightly higher result to others.
Adam Welch ah, ok 👍
Interesting to follow your lithium battery adventures. I'm still waiting for the lithium price vs capacity to reach a level I'm comfortable with.
I just replaced my almost 9 year old AGM battery bank with some AGMs that are only 15 months old. Two 100AH, two 110AH, total cost of $305US. 420AH * 12 volts = 5040WH, which I never use more than half of (50%DOD) but 50% is 2520WH and there's an inline V/A/W/AH/WH meter to track instantaneous and total power usage - when it hits 210AH, it's time to powerdown. My inverter isn't as efficient as yours - just over 90% vs 96% ;-)
Lithium-Ion cell price to manufacturers is expected to drop to below $100 U/S per kWh within 2 years, and is forecast to bottom out at around $75 after that.
Adam only checked his inverter efficiency at one point. It will have varied as input voltage dropped, and as it’s electronics warmed up. Perhaps he should have checked efficiency every hour or so. My guess is that it could easily have fallen to below 90% towards the end of the process. He did mention that the box felt quite warm after several hours of running.
I just ordered the diyBMS controller boards and the bms boards, now waiting 💯🤪
*what software is that you’re using on your tablet that you just showed??*
hello can you help me with bms?
I've put the EpEver inverter GUI on my workshop PC using another usb to RS485 adapter. I already have the 4210A GUI on my indoors PC. Can I parallel the two RS 485 's to be able to look at both units on my indoors PC?
Interesting test, I like your dashboard, it looks similar to the results I was getting with node-red. Personally I'd have password protected it - it's open to the web. I'll have to see if you have made vids explaining the software setup :)
Yes I know it’s open. It’s read only without a login. The grafana dashboard is behind a reverse proxy, and protected by cloud flare. Never say never, but I think it’s reasonably well protected and if not, it’d take me about an hour to rebuild.
@@AdamWelchUK I wasn't trying to hack it or anything, it's a great project mate :) Thank you for the great vids :)
No I didn’t think you were. Actually I’ve been waiting for that very comment from someone because it’s exactly the sort of thing I’d say to someone else! :-)
@@AdamWelchUK Glad I'm not the only weirdo then :) If it wasn't for you and other people pioneering new ideas for putting a system together, researching and testing to the geekiest level and sharing here, I wouldn't have started down the wormhole into lifepo4 in my camper.
Then there is the 3 months of testing the diesel heater that'll be the primary use for the 120AH ..... at 24W per hour. Making sure that I have enough solar to ensure I can run that heater in winter. 350W should do it....... Don't want to get cold. OTT I know
These diy BMS have no low voltage protection
This version doesn’t no. But v4 does.
Shame the second attempt didn’t produce the 1kw, but at least it’s showing up a struggling pack. More to investigate! Btw, can you show us what your actual load this set up is powering? Is it just a few garden lights?
Inverter was ONLY 96% efficient... What do you want?... 100% or more perhaps? :-)
good one :D
96% is actually pretty good
@@3dmixer552 It is crazy good if you ask me. I have 2 UPS's that I use for my desktop computer. Got curious and checked their efficiency. One is doing 88% and the other one is 90%.
@@borayurt66 same here I get 92 on low end and about 86 on full. So 96 is amazing but I doubt it anyway. I think he was just guessing.
isnt this a bit outdated technology now they have mpp inverters with built in batteries and bms..?you simple plug in the solar and away it goes..then the grid makes up the remaining power if the solar falls short or during dark hours..
Well you won't have that thing (you are talking about) for less then 60£. Thats the cost of the Epever MPPT charge controller here. + the batteries are free (salvaged from old laptops) and then a DIY BMS for 5£ each. Whats the price of your device ?
@@timm3802 your setup is cool but im not savvy enough or patient enough to build such a rig..and i love a bargain..for me paying 2k is worth the hassle..if i had that talent i would be making a business out of it..
@@charlie123866 Yeah some due business out of everything, and others sharing their thoughts and ideas. Its called "open source" :D and it's quite fun.
how about adding another pack?,,maybe that would make up the deference?
No you can't do that, as the pack would be 4.2V higher and the inverter might fail on overvoltage.
I charge mine up to 4.25V ... but its also Samsung batteries ICR-18650 -30B
They can handle the high voltage. Maybe you should aim for them too ?
Mine out a mixture of two types some LG some Sanyo. Both datasheets say they have a maximum charge voltage of 4.3volts. As I said though, I want to treat these nicely and don’t typically need to extract even half a kilowatt hour.
It's best to charge to as low a voltage as you can get away with based on your power needs. Charging to 80% capacity can significantly extend the number of cycles they can take.
@@keitho3133 Actually every vendor advice you to charge them full before using them for anything. If your statement was true (and in some cases it is), you would only get 80% out of new batteries.
"We're pulling 3.87 amps currently..."
groan...
Guess they don't apply to the discussion above.