diy solar

diy solar

BYD BMS

The guy in the video is still only playing around in the extreme bottom end of the charge of the cells.

He ought to put the packs on a float charge and just get every cell to 3.2V. Until that happens those bottom feeder cells are way past the point of usability.

He isn't really giving the float a chance to absorb when the cells are lower than 2.9V.

This is where taking the cover off and putting a single cell charger directly on the bottom cells will help.

The other way to get those very low cells up with the rest of the good ones is to do what DavePoz did and put every cell in parallel and passively just wait for a couple weeks until every cell is the same level.

Bulk charge to 53.3, float it at 52.9. And just sit there and wait while in float.

The overall voltage read at the big terminals is the AVERAGE of the cells. Those extremely low cells will take the voltage eventually. But don't forget, when there is a cell way lower than the average, there has to be some way more than the average.

If his(YouTube video) "dead cells" truly were dead they would not charge at all. They are taking charge. The cells just need a lot more juice and time.

52.9V = 3.309V per cell.

Okay, I trust you.
So, in my 24v case, I'm going to set Bulk to 26.6v (or 26.7?) and Float to 26.4 (or 26.5?), and wait. Let the sun rise, let the sun set. For a week, for a month, for a year... just wait.
Do I have that right?
What will be my sign when all cells are finally balanced and the battery is functioning as it can and should?
 
Okay, I trust you.
So, in my 24v case, I'm going to set Bulk to 26.6v (or 26.7?) and Float to 26.4 (or 26.5?), and wait. Let the sun rise, let the sun set. For a week, for a month, for a year... just wait.
Do I have that right?
What will be my sign when all cells are finally balanced and the battery is functioning as it can and should?
Do you have any way to see each cell voltage?

This is one of my BMS, it has a Bluetooth app to display each cell voltage.

2020-03-08.png

If your BMS doesn't show you what the cell diff is, you really have no "absolute" way of knowing unless you can put a voltmeter4 on each cell.

Before I had the BMS installed I used an ISDT Battgo. (link) and it showed each cell voltage, I would just go record the voltage to a spreadsheet.)
My BMS also has an LCD display that I can write down the voltage of each cell.

After a couple of hours or days of doing this, you will probably appreciate looking at your Bluetooth app and seeing the cell voltage.

Once I understood and accepted the upper and lower "flat curve" the cells come into alignment/balance all on their own. Unfortunately with these used cells, yes you still have to float them.

The screenshot shown is of my worst two packs in the battery system. Right now the two worst cells in all 64 that I have are hiding in plain sight.

Cell 6 and Cell 16.
If I discharge the battery below 51.1V, Cell 6 will plummet very fast to less than 2.9V.
If I charge to more than 53.8V Cell 16 will overcharge and go up to 3.60 nearly instantly.

So, to answer your specific question on how long it takes... Once I set the float charge to 53.3V, just one day and all is happy. The higher the bulk charge you set the more of a chance of finding a high voltage dud cell. I have my bulk set at 56.5V ( 3.531V) and the MPP Solar LV5048 is doing great with these settings.

For a week I would set the float to 53.1 for a day, then 52.9, then 53.2, then 52.9, then 53.3, then 52.9. Now many times in the morning. I see 0.002V Cell Difference, usually, 3 of the 8 packs are in perfect balance with their pack cells, which is 0.000V cell difference.

3.316V per cell is about where my cells love to be.

I can run a full 10 hours without solar assistance at that starting voltage, but I have 4 sets of 8 BYD's to go with that load. I have but only one full week of pure solar recharging on these.

So far this last week, by 10:00 am the cells are fully charged to run my house from 12:00 to 8:00 PM.
 
Okay, I trust you.
So, in my 24v case, I'm going to set Bulk to 26.6v (or 26.7?) and Float to 26.4 (or 26.5?), and wait. Let the sun rise, let the sun set. For a week, for a month, for a year... just wait.
Do I have that right?
What will be my sign when all cells are finally balanced and the battery is functioning as it can and should?
Equilibrium: I understand your frustration. I have 4 of these pre-built units. Not having a way to easily see individual cell voltages is truly frustrating. But short of replacing the BMS - (or designing/building some sort of "Y" cable setup with an ISTDGo - @jasonhc73 help! ) there is no way to know exactly what is going on the cell level. So I have compromised - watched carefully for BMS cutoffs and measured discharge watt/amp hours against recharge watt/amp hours. That gave me my baseline and allowed me to set charge/discharge routines that made sense. A cheap meter with a proper shunt can do that. Here's what I used:
Meanwhile, just so I know exactly what you are up against - could you post a picture of your wiring setup - and the charging parameters/discharge loads and times you are using - so maybe we can be of more assistance? You may have a bad unit - but it's unlikely you have 3 clunkers - so odds are something else is awry in setup or metering or..? Hard to say, but keep plugging away, the answers will come. You might post another thread as well - with a carefully written summary - for @Will Prowse. Best...
 
Equal: By the way, here's part of my confusion. I have no idea what this means:
"No matter the settings or usage load (almost zero, a 30 watt LED bulb and the controller itself), it just shuts down to like 30w on my 2kw of available power. The battery doesn't ever charge at 30w, not past 30 to 50%. Who knows how accurate that even is."
In order for this to make sense, we have to know if you are in Bulk, absorption or float, what the charge controller is inputting vs. what the load is - whether the load is coming from inverter AC or DC direct, what you mean by "30 to 50%" - How do you know you have 2kw available? State of charge derived from voltage readings is a moving target - and has to be defined based on careful repeatable readings. Sorry for my lack of understanding - but consider these as some readings;
(these are totally bogus numbers - just given as an example of the kind of valuable information to have at hand for evaluation)
24 hour cycle: Begin after a full charge (Charge to 27.4 or BMS disconnect)
Built in meter (at rest 24 hours): 26.5v (start)
Inverter output: 130w avg. for 3 hours (390wh)
Solar CC charge: 200w avg. for 7 hours (1400wh)
Built in meter (finish): 26.8
Built in meter (at rest 1 hour): 26.6
Built in meter (at rest 24 hour)s: 26.5
Start new cycle (adjust or keep charge time/voltage) and note measurements.
Only by measuring what goes in and what comes out will you have a sense of the state/capacity of your battery. Eventually you will create a set of charge/discharge parameters which will give you a steady reliable state.
Also, are your 3 units hooked up in parallel? Or are you doing these tests one battery at a time?
Best...
 
Do you have any way to see each cell voltage?

This is one of my BMS, it has a Bluetooth app to display each cell voltage.

View attachment 8753

If your BMS doesn't show you what the cell diff is, you really have no "absolute" way of knowing unless you can put a voltmeter4 on each cell.

Before I had the BMS installed I used an ISDT Battgo. (link) and it showed each cell voltage, I would just go record the voltage to a spreadsheet.)
My BMS also has an LCD display that I can write down the voltage of each cell.

After a couple of hours or days of doing this, you will probably appreciate looking at your Bluetooth app and seeing the cell voltage.

Once I understood and accepted the upper and lower "flat curve" the cells come into alignment/balance all on their own. Unfortunately with these used cells, yes you still have to float them.

The screenshot shown is of my worst two packs in the battery system. Right now the two worst cells in all 64 that I have are hiding in plain sight.

Cell 6 and Cell 16.
If I discharge the battery below 51.1V, Cell 6 will plummet very fast to less than 2.9V.
If I charge to more than 53.8V Cell 16 will overcharge and go up to 3.60 nearly instantly.

So, to answer your specific question on how long it takes... Once I set the float charge to 53.3V, just one day and all is happy. The higher the bulk charge you set the more of a chance of finding a high voltage dud cell. I have my bulk set at 56.5V ( 3.531V) and the MPP Solar LV5048 is doing great with these settings.

For a week I would set the float to 53.1 for a day, then 52.9, then 53.2, then 52.9, then 53.3, then 52.9. Now many times in the morning. I see 0.002V Cell Difference, usually, 3 of the 8 packs are in perfect balance with their pack cells, which is 0.000V cell difference.

3.316V per cell is about where my cells love to be.

I can run a full 10 hours without solar assistance at that starting voltage, but I have 4 sets of 8 BYD's to go with that load. I have but only one full week of pure solar recharging on these.

So far this last week, by 10:00 am the cells are fully charged to run my house from 12:00 to 8:00 PM.
No, I don't have the ability to see individual cells...I bought the BYD in the big wheeled case with the BMS. The one will recently reviewed and then connected my same LV2424 to in another video shortly thereafter. This BMS cleary does something, but as far as I am aware, it doesn't give me a visual of what it is doing other than the sorta useless voltmeter on the box itself.

I have 3 of these for three separate LV2424 arrays:
I think this is what's inside:

The only monitoring I can do is through Watchpower, which I am grateful for and this guy (which hasn't been incredibly helpful yet):

I can order the ISDT, but before I do I want to make sure it is not too impossible for me to actually connect it to my BYD. My batteries are not as accessible as yours because they are crammed into this box and closed up. I'm sure if I start tearing it down it won't go well and that I'll void whatever warranty I do have. I am thinking I should write Tech Direct and ask them if they can send me batteries that are balanced instead of whatever they sent me.
 
No, I don't have the ability to see individual cells...I bought the BYD in the big wheeled case with the BMS. The one will recently reviewed and then connected my same LV2424 to in another video shortly thereafter. This BMS cleary does something, but as far as I am aware, it doesn't give me a visual of what it is doing other than the sorta useless voltmeter on the box itself.

I have 3 of these for three separate LV2424 arrays:
I think this is what's inside:

The only monitoring I can do is through Watchpower, which I am grateful for and this guy (which hasn't been incredibly helpful yet):

I can order the ISDT, but before I do I want to make sure it is not too impossible for me to actually connect it to my BYD. My batteries are not as accessible as yours because they are crammed into this box and closed up. I'm sure if I start tearing it down it won't go well and that I'll void whatever warranty I do have. I am thinking I should write Tech Direct and ask them if they can send me batteries that are balanced instead of whatever they sent me.
If you start at the first post in this entire thread, you will have listed out all the pieces you need to get at the wire connector.
TE made a very nice connector that connects to their black box of a BMS.

 
Equal: By the way, here's part of my confusion. I have no idea what this means:
"No matter the settings or usage load (almost zero, a 30 watt LED bulb and the controller itself), it just shuts down to like 30w on my 2kw of available power. The battery doesn't ever charge at 30w, not past 30 to 50%. Who knows how accurate that even is."
In order for this to make sense, we have to know if you are in Bulk, absorption or float, what the charge controller is inputting vs. what the load is - whether the load is coming from inverter AC or DC direct, what you mean by "30 to 50%" - How do you know you have 2kw available? State of charge derived from voltage readings is a moving target - and has to be defined based on careful repeatable readings. Sorry for my lack of understanding - but consider these as some readings;
(these are totally bogus numbers - just given as an example of the kind of valuable information to have at hand for evaluation)
24 hour cycle: Begin after a full charge (Charge to 27.4 or BMS disconnect)
Built in meter (at rest 24 hours): 26.5v (start)
Inverter output: 130w avg. for 3 hours (390wh)
Solar CC charge: 200w avg. for 7 hours (1400wh)
Built in meter (finish): 26.8
Built in meter (at rest 1 hour): 26.6
Built in meter (at rest 24 hour)s: 26.5
Start new cycle (adjust or keep charge time/voltage) and note measurements.
Only by measuring what goes in and what comes out will you have a sense of the state/capacity of your battery. Eventually you will create a set of charge/discharge parameters which will give you a steady reliable state.
Also, are your 3 units hooked up in parallel? Or are you doing these tests one battery at a time?
Best...

I'm sorry @GVSolar , I'm speaking in layman's terms and not using proper terminology, I'm sure. I was saying this: I had almost no output load from the array or the battery, only the LV2424 and a 4.5w LED bulb (equal to 30w incandescent) is using any power whatsoever. It was midday at the time, so I could test my 9 panel array and it was getting over 2kwh at that time. However, Watchpower was telling me that only 30 to 40 of those watts were being used to charge the BYD. 30 watts is, of course, almost nothing. So it was not even making a dent in the BYDs charging. Watch power would tell me that the battery was at 50% or less, and would even decline. That's all I was saying. It was before I clearly understood what you had diagnosed: That my cells were limiting what the BMS was allowing.

I have enough stuff here for 3 separate systems (for 3 micro cabins), but I've been doing these tests one battery at a time, individually.

But tomorrow I am going to do my best to set up a table and all three LV2424 units right next to each other so that I can show you guys, and monitor them simultaneously. That seems like it will save me a ton of time in the long run. Each LV2424 will be connected to its own Tech Direct BYD (with their installed BMS) and each will be fed by it's own 1500w 112.8v 24.8a solar panel array. I'll post a few pics of the layout when I have it looking nice enough.

Your sample data will prove to be very helpful. I will no doubt go back to your and @jasonhc73 recent instruction a hundred times. But I will do my best to follow, where my experience and common sense let me do, well, I don't know what I can do about that. Maybe @Will Prowse will even see something that is relevant enough to the masses to address.
It still bugs me a little that Tech Direct would not do active cell balancing before sending them out, as they just connected and charged for their BMS, and boxed these things up. If the customer already had their BMS installed when it arrived, why would they expect to have to tear what was just installed apart in order to connect things that would help them to force their new battery to work right. It is bizarre to me. But maybe I'm missing something or just don't know enough about this field yet.
 
If you start at the first post in this entire thread, you will have listed out all the pieces you need to get at the wire connector.
TE made a very nice connector that connects to their black box of a BMS.

I'll open it up in the morning. Are you saying you would recommend I order the new BMS, take out Tech Direct's, and follow your instructions to make and connect the new one?
 
I'll open it up in the morning. Are you saying you would recommend I order the new BMS, take out Tech Direct's, and follow your instructions to make and connect the new one?
Or splice into their wire connector.

I would use their connector with a M/F Dupont jumper wires. Then wire in your ISDT or other BMS to the jumper wires. They already made the connector, it works fine.

 
Or splice into their wire connector.

I would use their connector with a M/F Dupont jumper wires. Then wire in your ISDT or other BMS to the jumper wires. They already made the connector, it works fine.

Okay. Do I select 40Pcs, 120Pcs or 240Pcs?
 
I'm sorry @GVSolar , I'm speaking in layman's terms and not using proper terminology, I'm sure. I was saying this: I had almost no output load from the array or the battery, only the LV2424 and a 4.5w LED bulb (equal to 30w incandescent) is using any power whatsoever. It was midday at the time, so I could test my 9 panel array and it was getting over 2kwh at that time. However, Watchpower was telling me that only 30 to 40 of those watts were being used to charge the BYD. 30 watts is, of course, almost nothing. So it was not even making a dent in the BYDs charging. Watch power would tell me that the battery was at 50% or less, and would even decline. That's all I was saying. It was before I clearly understood what you had diagnosed: That my cells were limiting what the BMS was allowing.

I have enough stuff here for 3 separate systems (for 3 micro cabins), but I've been doing these tests one battery at a time, individually.

But tomorrow I am going to do my best to set up a table and all three LV2424 units right next to each other so that I can show you guys, and monitor them simultaneously. That seems like it will save me a ton of time in the long run. Each LV2424 will be connected to its own Tech Direct BYD (with their installed BMS) and each will be fed by it's own 1500w 112.8v 24.8a solar panel array. I'll post a few pics of the layout when I have it looking nice enough.

Your sample data will prove to be very helpful. I will no doubt go back to your and @jasonhc73 recent instruction a hundred times. But I will do my best to follow, where my experience and common sense let me do, well, I don't know what I can do about that. Maybe @Will Prowse will even see something that is relevant enough to the masses to address.
It still bugs me a little that Tech Direct would not do active cell balancing before sending them out, as they just connected and charged for their BMS, and boxed these things up. If the customer already had their BMS installed when it arrived, why would they expect to have to tear what was just installed apart in order to connect things that would help them to force their new battery to work right. It is bizarre to me. But maybe I'm missing something or just don't know enough about this field yet.

@jasonhc73 is way smarter than I am - but - I think you need to slow down.
From my reading of your explanation above I'm not sure you ever discharged enough amps/energy to make a dent in your battery. (4.5w is nothing up against one of these BYDs) And the entire charging process depends on the charge parameters that have been set on your charge controller and your inverter. Bulk=? Absorb =? (for how long) Float = ? Battery AH = ? Example: Many charge controllers stop charging as soon as the charging amps reach 2% of the battery rating. So if you rated the battery at 200ah the CC would stop charging as soon as the amps necessary to hold absorb reached 4 amps - which would happen in no time at all if you only discharged 4.5 watts. A 2kw array in full sun would do that in less than a minute. As to the "Watchpower" - it's analysis of your state of charge is only as good as the numbers you put in to program it... and those may be suspect at this point.

You can tear deep into the case if you want to - and I intend to do so myself at some point (with ONE of my units) but only after I have them settled into a conservative charging regimen that meets my needs.

I don't mean to discount your abilities at all - but I suspect that building out a custom interface at this point - when you are still learning the basics - is not in your best interest. Just my 2 cents.

Meanwhile, I suggest taking a picture of how you have one system wired - also of your charge controller settings - and any other settings you have programmed into the inverter - and post them for analysis. @jasonhc73 will have a good idea about the MPP - I do not. I highly respect his judgement. But I truly suspect something else is going on here.

As always, my advice is free - and worth what you paid for it. Good thoughts, sir.
 
Last edited:
Don't open the case, the cells need compression. Don't do it.

The cells do not need active balancing. Hook up a LiFePO4 8S or 16S BMS (1/2 packs), and cycle it. It will balance it as much as it can in a few cycles, and you are done.

No need to see individual cell voltages. The BMS that comes with the ready built unit works great. Just connect BMS and cycle it.

If you want to see the condition of your pack, build your entire system and do a single capacity test. Cycle it a few times and test again. If the capacity results look good, no need to check anything. Keep using it.
 
These packs won't balance until you hit the balancing voltage threshold. So be sure to charge your BYD packs to 29 volts(3.625V for each cell). And cycle down to 23-24V. Repeat this over and over.

If you want to cycle up to 90% SOC, use a shunt and set absorption manually. Thats the best method. Dont set absorption manually till you cycle it a few times.
 
These packs won't balance until you hit the balancing voltage threshold. So be sure to charge your BYD packs to 29 volts(3.625V for each cell). And cycle down to 23-24V. Repeat this over and over.

If you want to cycle up to 90% SOC, use a shunt and set absorption manually. Thats the best method. Dont set absorption manually till you cycle it a few times.
Okay. Are you saying I can use Watchpower to try setting "Bulk" to 29v and the "Float" to 23 or 24v?
 
Don't open the case, the cells need compression. Don't do it.

The cells do not need active balancing. Hook up a LiFePO4 8S or 16S BMS (1/2 packs), and cycle it. It will balance it as much as it can in a few cycles, and you are done.

No need to see individual cell voltages. The BMS that comes with the ready built unit works great. Just connect BMS and cycle it.

If you want to see the condition of your pack, build your entire system and do a single capacity test. Cycle it a few times and test again. If the capacity results look good, no need to check anything. Keep using it.
When I first got them I took my first ready built unit and charged it until full, I believed it was full anyway. The LV2424 light had been solid green for days. I tried to do everything just like in your video. I used a heat gun to discharge and measured. Only got 2.2kwh. I was disappointed, but wanted to try again.
That's when I started noticing the low input rate, built-in BMS was slowing the CC's roll. It seems it won't charge ever. The second battery with built-in BMS behaves the same. I have a third I haven't cracked yet. I have been using the same LV2424 all along (I have 3).
So, I want to keep charging and discharging, testing (before I complain to Tech Direct) but it seems I can't. I've tried lots of the guy's settings, don't change much.
I don't really want to see the condition of the packs or monitor cells, I just want them to work, honestly Will. If in 3 to 5 years I need to upgrade to higher tech more pricey storage at that time, fine and good.
I am confused though. You say "just connect a BMS and cycle" but I thought I had a connected BMS in my built-ins. Are you saying add a different one? Or am I not connected to and using the one I have? Whatever the case, cycling is what isn't happening cause they aren't really charging.
 
When I first got them I took my first ready built unit and charged it until full, I believed it was full anyway. The LV2424 light had been solid green for days. I tried to do everything just like in your video. I used a heat gun to discharge and measured. Only got 2.2kwh. I was disappointed, but wanted to try again.
That's when I started noticing the low input rate, built-in BMS was slowing the CC's roll. It seems it won't charge ever. The second battery with built-in BMS behaves the same. I have a third I haven't cracked yet. I have been using the same LV2424 all along (I have 3).
So, I want to keep charging and discharging, testing (before I complain to Tech Direct) but it seems I can't. I've tried lots of the guy's settings, don't change much.
I don't really want to see the condition of the packs or monitor cells, I just want them to work, honestly Will. If in 3 to 5 years I need to upgrade to higher tech more pricey storage at that time, fine and good.
I am confused though. You say "just connect a BMS and cycle" but I thought I had a connected BMS in my built-ins. Are you saying add a different one? Or am I not connected to and using the one I have? Whatever the case, cycling is what isn't happening cause they aren't really charging.
I used my packs without any BMS at all for about two weeks.

I set the charge on the MPP Solar to a voltage and let it charge, then used the battery.

BMS can only do two things, open the circuit (disconnect the battery) and add a small resistive load to try to discharge the highest voltage cell.

Set the MPP Solar to bulk charge.
Set the MPP Solar to float charge at the nominal voltage.

These packs do not behave anywhere near like NEW cells. These are VERY USED.

If you don't have a kWh meter showing you from full charge to where ever your battery disconnects (FULL to EMPTY), the real kWh capacity is merely a guess.
 
I used my packs without any BMS at all for about two weeks.

I set the charge on the MPP Solar to a voltage and let it charge, then used the battery.

BMS can only do two things, open the circuit (disconnect the battery) and add a small resistive load to try to discharge the highest voltage cell.

Set the MPP Solar to bulk charge.
Set the MPP Solar to float charge at the nominal voltage.

These packs do not behave anywhere near like NEW cells. These are VERY USED.

If you don't have a kWh meter showing you from full charge to where ever your battery disconnects (FULL to EMPTY), the real kWh capacity is merely a guess.
I do have a kwh meter. The problem is, when I implement all these ideas, I don't see a change in behavior. By behavior, I mean the CC or the BMS is not allowing the battery to charge at any reasonable input. No matter what, within minutes, the batteries (any and all of the 3) are getting 30watts from my panels and aren't doing jack squat. They don't charge up. Says they are, but hours and days pass and still are the same.
 
Back
Top