Seeking Normally Closed Latching Contactor and Q&A about balancing my Li Ion battery bank

124678

Comments

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited January 2018 #92
    So, just to wrap up, I go the cells top balanced successfully. The array is all buttoned up with the cell covers and the spaghetti wiring is as cleaned up as possible. System is up and running.

    Rick, your point about the cells staying in balance once the top balance is completed is well taken. As much as I would prefer to think that the balancers I purchased are a worthwhile investment, and in spite of how elated my battery vendor was to find these devices, and although the balancer maker has years worth of testimonies, and although the balancers may help at the bottom of the range if one is looking to squeeze every last bit of charge out of the array (depending on load and how quickly the devices can balance those cells with a bit less capacity, I would find it hard to justify the expenditure when I expand the array.

    For anyone reading through this meandering thread I would summarize my take on the balancers as follows:

    1) The balancers seem to balance voltage, but they do not balance state of charge (SOC). At least not over the 2-3 day period I left them on the pack after inserting the replacement cell.

    2) The vendors website gives the impression the balancers will balance SOC, although it does not specifically state that. And it certainly does not qualify that the balancers only balance voltage.

    3) The balancers do not prevent over charging or over draining of a cell. A BMS of some sort is still required (even if it is simply tapping into the celllog output as karrak suggested.

    ------------------------------------------------------
    This project has been partially designed to be a learning experience so I have no regrets. In fact there was one other failed experiment I will mention. I chose to use Unistrut as a lower cost way to ground mount the solar panels. I think that goal was achieved (I ran the numbers on several engineered ground mount solutions and the unistrut always came out more competitive). However I pushed it bit to see if I could pull of an adjustable mount that could increase the angle of the panels in winter and decrease the angle in the summer.

    Toward that end I used "Telestrut" which is simply a strut inside a strut. I deployed the telestrut on the angled struts the panels mount to (although there are other approaches). The basic idea works, BUT it turned out to be a major PITA to make the adjustments, as I designed the system anyway).AND, telestrut is insanely expensive. It diluted all cost savings the unistrut approach had achieved, making the mount as costly or more costly than the engineered mounting systems. But again, not a headache as this was an experiment. The mount is one of four that will be built (6 panels per mount), and now I know what works and doesn't work.









    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    Oh, and I did measure the cellogs with a Fluke.... 4MV difference.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    One other note for anyone considering Unistrut (not telestrut). DON'T use the unistrut solar end clamps, mid-clamps, or angle brackets. I tried them, but found third party solutions better engineered and less expensive. For example the Unistrut mid-clamps did not have as much meat to overlap onto the panel as say, the ProSolar clamps.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited February 2018 #95
    UPDATE: At this stage of the project I am now finding it difficult to successfully perform a top balance before placing the array back online. I once again tried to get the third party balancers to balance the cells, but even after leaving the balancers on for 4-5 days, all the balancers did was balance cell voltage, not SOC. In trying to get the array up 3.4 volts or higher -out of the flat part of the curve- a few cells will bang up against 3.65v, while many cells are still at, say, 3.38x -- and thus pack charging must be discontinued. I am now rating the balancers as "less than worthless".

    I tried placing a DC lamp on individual cells to discharge the full cells, but finding that is taking foooorever. So I have taken Karrak's que and purchased a hobby charger. Simon, you had expressed curiosity about the capacity of the damaged cells. Now that I have a cell charger I'll let you know. I was going to buy a Turnigy Reaktor 300W, but found a bruising review of the latest version (claiming product has gone downhill) from someone who seems to know what they are talking about. I ended up purchasing two different units for comparison; an HTRC B6 Mini V2 ($25 + $16 for power supply), and an iMax B6AC V2 ($57).

    So now I am essentially starting over. I don't think I ever payed close enough attention to top balancing the cells when first purchased. Now I'm going to try and do it right. My intention is to cycle each of the 32 cells, charge them fully, and measure capacity and internal resistance. Will post results comparing the good to the overcharged cells when done.

    Cheers








    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    edited February 2018 #96
    >  a few cells will bang up against 3.65v, while many cells are still at, say, 3.38x -- and thus pack charging must be discontinued.

    I haven't followed this thread, but isn't the idea of a balancer (bypass or active) to allow exactly this?   Ie, you can safely continue charging because no more current will be applied to the cells that are at 3.65V

    I expect that if you hold all the cells at 3.65V long enough (which implies a very low current), they will all be pretty close in terms of SOC.

    I agree that once one cell reaches max voltage, charging should only continue at an amp rate that the balancer can handle.  Fail to do this and a cell would see excessive voltage.

    I am available for custom hardware/firmware development

  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    Thanks for the update.

    I am not sure that your balancers are useless. It looks like your battery is badly out of balance and it will take a while for the balancers to balance it.

    The way to stop the situation where you are having some cells at 3.38V and some going over 3.65V is to reduce the SCC (charge controller) bulk voltage and float voltage to around 54.1V (3.38*16) and wait for the balancers to balance all the cells to within 0.05V. When this has occurred increase the bulk and float voltage by 0.2V and wait for the balancers to do their work. Keep slowly increasing the bulk and float voltage until you reach 57.6V(3.6*16). This can be done over several days. I hope you have your Cellog 8 connected to your SCC to stop it charging if any cell voltage goes higher than 3.65V.

    It is a shame that Hobbyking is no longer selling the version of the Turnigy 300W that I bought. My charger is functionally the same as the Junsi iCharger 206b which is still available. There is a 10A version, the 106b and a 30A version, the 306A. The two units you bought will be just about useless for discharging the cells as the maximum discharge current is ~1.5A-2.0A at 3.2V.

    Simon
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • cow_rancher
    cow_rancher Solar Expert Posts: 117 ✭✭✭✭
    One other note for anyone considering Unistrut (not telestrut). DON'T use the unistrut solar end clamps, mid-clamps, or angle brackets. I tried them, but found third party solutions better engineered and less expensive. For example the Unistrut mid-clamps did not have as much meat to overlap onto the panel as say, the ProSolar clamps.
    How about listing your sources of supply for your Unistrut 3rd party parts?

    Is telestrut just not Unistrut  with sizes that telescope into one another?

    Rancher
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    jonr said:

    I haven't followed this thread, but isn't the idea of a balancer (bypass or active) to allow exactly this?   Ie, you can safely continue charging because no more current will be applied to the cells that are at 3.65V

    Well, not exactly. The balancers I have do not stop current from flowing into a cell once max voltage is reached. They are not that type. But, they are daisy chained to each other and are supposed to equalize voltage among the cells over time. So far though I have found that although they equalize voltage -over time- they don't seem to be equalizing SOC. See my reply to Karrak also.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    karrak said:
    Thanks for the update.

    I am not sure that your balancers are useless. It looks like your battery is badly out of balance and it will take a while for the balancers to balance it.

    The way to stop the situation where you are having some cells at 3.38V and some going over 3.65V is to reduce the SCC (charge controller) bulk voltage and float voltage to around 54.1V (3.38*16) and wait for the balancers to balance all the cells to within 0.05V. When this has occurred increase the bulk and float voltage by 0.2V and wait for the balancers to do their work. Keep slowly increasing the bulk and float voltage until you reach 57.6V(3.6*16). This can be done over several days. I hope you have your Cellog 8 connected to your SCC to stop it charging if any cell voltage goes higher than 3.65V.

    It is a shame that Hobbyking is no longer selling the version of the Turnigy 300W that I bought. My charger is functionally the same as the Junsi iCharger 206b which is still available. There is a 10A version, the 106b and a 30A version, the 306A. The two units you bought will be just about useless for discharging the cells as the maximum discharge current is ~1.5A-2.0A at 3.2V.

    Simon

    I gave the balancers 4 days to balance the pack with no charge coming in. Maybe I should have waited longer, I don't know.

    I had first tried reducing SCC on this latest foray to get the cells balanced as you are suggesting. I can't recall exactly but I think I did start at around 54.1. Then when any cell went full I stopped and let the balancers do their work (did not wait anything near 4 days though). I was able to slowly increase the voltage, and did this several times. But finally it reached a point where some cells would immediately hit 3.65 while others lagged in the 3.3x range. Again, resting the pack with the balancers still not equalize the charge.

    I now watch the meter through all these exercises and NEVER let a cell go over 3.65. ;>

    I am strongly disposed toward charging and testing each cell individually at this point. So if the chargers I bought don't cut it I will swap them for a Junsi unit.

    By the way, when measuring internal resistance of a cell, is that performed at full charge? Documentation is not clear on that.

    As always, thank you.
    jim


    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    cow_rancher said: How about listing your sources of supply for your Unistrut 3rd party parts?

    Is telestrut just not Unistrut  with sizes that telescope into one another?

    Rancher
    There is photo of the balancers and the vendor listed earlier in this thread. But let me say that in spite of my earlier comment that the balancers are less than worthless, I don't want to trash the vendor at this point, because I am one person, on a first time system, and you don't want to draw conclusions from that. I have spoken with the vendor and the maker in China - very cooperative, and swearing that thousands of balancers sold with few issues. So I could just be being idiotic on this. As I alluded to, knowing now what I have learned from Karrak and others, I may not have have been fastidious enough at the very beginning in getting the cells top balanced properly. That is why I am starting over from scratch and charging each cell individually, as well as testing it. Maybe down the road the balancers will prove effective if I get everything setup properly from the getgo. ;>

    Yes, the telestrut is just a 1/8" oversized piece of unistrut. And for that they charge you about 10X over unistrut. Gotta love it.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    edited February 2018 #102
    There are various games a manufacturer can play with specs for a charge shuffling (aka active) balancer.   For example, say it's rated at 6A.  But 6A at what differential between two cells (charge moves faster with a higher V differential in capacitor type designs).  Continuously or intermittent?  And of course if the low cell is several hops away from the high cell, then the effective rate is much lower.     Once one cell reaches max voltage, charging needs to be limited to this unknown effective rate (not the rated amp rate).  Do that and no cell will over-voltage up to the point where all cells are at max voltage.  A cell at 3.65V being charged at 5A with charge being shuffled away at an actual 6A *will* decrease in voltage.

    If a charge shuffling balancer doesn't have cells voltage balanced after no load or charging for 4 days, then I think it is defective.  It would only take a failure of one unit.

    Someone should put these balancers on the bench and get some real specs.

    I am available for custom hardware/firmware development

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    jonr said:
    There are various games a manufacturer can play with specs for a charge shuffling (aka active) balancer.   For example, say it's rated at 6A.  But 6A at what differential between two cells (charge moves faster with a higher V differential in capacitor type designs).  Continuously or intermittent?  And of course if the low cell is several hops away from the high cell, then the effective rate is much lower.     Once one cell reaches max voltage, charging needs to be limited to this unknown effective rate (not the rated amp rate).  Do that and no cell will over-voltage up to the point where all cells are at max voltage.  A cell at 3.65V being charged at 5A with charge being shuffled away at an actual 6A *will* decrease in voltage.

    If a charge shuffling balancer doesn't have cells voltage balanced after no load or charging for 4 days, then I think it is defective.  It would only take a failure of one unit.

    Someone should put these balancers on the bench and get some real specs.

    I'd be happy to mail you one or two if you want to volunteer. This whole project at this point is just an expensive hobby/experiment. I have grid power.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    LostinSpace said:
    I gave the balancers 4 days to balance the pack with no charge coming in. Maybe I should have waited longer, I don't know.

    I had first tried reducing SCC on this latest foray to get the cells balanced as you are suggesting. I can't recall exactly but I think I did start at around 54.1. Then when any cell went full I stopped and let the balancers do their work (did not wait anything near 4 days though). I was able to slowly increase the voltage, and did this several times. But finally it reached a point where some cells would immediately hit 3.65 while others lagged in the 3.3x range. Again, resting the pack with the balancers still not equalize the charge.
    Are you letting the battery charge to say 54.1V then terminating the charge or leaving it charging at 54.1V?

    If you terminate the charge the individual cell voltages will all drop and converge so that the balancers will do nothing.

    Below is a graph showing how the individual cell voltages diverge while the battery is being charged and come back together when the charging stops. This battery doesn't have any automatic cell balancing.


    I now watch the meter through all these exercises and NEVER let a cell go over 3.65. ;>
    It is really easy to connect the cellog 8 up to your Midnite.
    I am strongly disposed toward charging and testing each cell individually at this point. So if the chargers I bought don't cut it I will swap them for a Junsi unit.
    This is the best way to get your system up and running. How were you planning to do the capacity test and balance the battery using the Junsi unit?

    By the way, when measuring internal resistance of a cell, is that performed at full charge? Documentation is not clear on that.
    I wouldn't use the Junsi unit to measure the internal resistance. The apparent resistance of the battery is affected by how much current you are drawing/supplying, the temperature and the SOC. A better test is to take a reading of the battery or cell voltages and current when it is in use with a small load of around C/20 and then apply a large load > C/2 and take the voltage and current readings. The resistance is
    Rbat=(Vc/2-Vc/20)/(Ic/2-Ic/20)

    If you want to see how much this charges over time you should make the measurements under the same conditions with the same current draw.

    Simon
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited February 2018 #105

    Wow. Another very helpful graph. Nice to see that that sort of temporary deviation is to be expected.

    It is really easy to connect the cellog 8 up to your Midnite.
    Thanx for holding my feet to the fire on this one. Today I programmed my cellogs to close the output relay if any cell exceeds 3.60, or if pack voltage exceeds 28.5v (half of 57v). The trick will be to get my Midnite Classic charge controller to stop charging on the proper external input. The Classic has two Aux ports, but only one of the two can be programmed for INPUT, and I'm already using it for a WizzBang Jr.; a device to track power usage and SOC. So I am going to take this up with the Midnite people. Ha. This topic almost wraps back to the original reason I made the post

    This is the best way to get your system up and running. How were you planning to do the capacity test and balance the battery using the Junsi unit?
    First, slight change of plans. The Junsi + power supply is ~$200. I have other places for that cash to go for the moment. So I am keeping the two lesser units for now (30-day return window), foregoing capacity testing at this point, and yesterday I individually charged 16 of the 32 cells to 3.60. Today I put the array back together, and this afternoon will put a moderate load on (standup freezer) and see what happens. No balancers. I did measure the cell voltages after sitting overnight and they are running from 3.487-3.564 (77mv dif).

    I was actually going to ask you how your charger does capacity testing. Does it have a dedicated feature that will cycle the battery and report an AH number? If not it occurred to me that you could simply time how long it takes each cell to go from, say, 3.45 to 2.8, and compare the results across all cells to at least get an idea about relative capacity. Or you could discharge all cells at a fixed current rate and calculate AH?

    I wouldn't use the Junsi unit to measure the internal resistance. The apparent resistance of the battery is affected by how much current you are drawing/supplying, the temperature and the SOC.
    Yeah, for giggles I used the internal resistance function the iMAx unit (clearly the better of the two I bought), and readings were all over the map, even though temp was somewhat uniform and I tested just after completion of charge to 3.60.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • Raj174
    Raj174 Solar Expert Posts: 795 ✭✭✭✭
    @LostinSpace

    Even though all the cells are 3.6 volts, they are not balanced. But I suspect that you know that. If I were to use the bank that way, I would use the balance boards because as the cell voltages diverge the boards will protect the cells with slightly lower capacity from getting over charged. Then the bank can be charged until the individual cells reach 3.6 volts.

    If the cells were balanced at 3.55 to 3.6 volts then no balance boards needed and absorb could be set to 55.2. Have you considered charging all the cells in parallel?

    Rick



     
    4480W PV, MNE175DR-TR, MN Classic 150, Outback Radian GS4048A, Mate3, 51.2V 360AH nominal LiFePO4, Kohler Pro 5.2E genset.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited February 2018 #107
    Raj174 said:
    Even though all the cells are 3.6 volts, they are not balanced. But I suspect that you know that. If I were to use the bank that way, I would use the balance boards because as the cell voltages diverge the boards will protect the cells with slightly lower capacity from getting over charged. Then the bank can be charged until the individual cells reach 3.6 volts.

    If the cells were balanced at 3.55 to 3.6 volts then no balance boards needed and absorb could be set to 55.2. Have you considered charging all the cells in parallel?

    Rick



     

    I actually let the cells rest in parallel for two days before putting each cell on the rc chargers I bought.

    I'm a little uncertain as how to proceed from here to get to the 3.55-3.6 target. Not sure if putting the balancers back on and slowly incrementing pack voltage upward will reach the target. That strategy failed before. But perhaps will work this time. I can try and see.

    I could certainly take your hint and first give the pack another charge with the cells wired in parallel. I'm a little surprised to see a few cells drop into the 3.48x range after sitting just one night. I don't have a reference to know if that is normal.


    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    Thanx for holding my feet to the fire on this one. Today I programmed my cellogs to close the output relay if any cell exceeds 3.60, or if pack voltage exceeds 28.5v (half of 57v). The trick will be to get my Midnite Classic charge controller to stop charging on the proper external input. The Classic has two Aux ports, but only one of the two can be programmed for INPUT, and I'm already using it for a WizzBang Jr.; a device to track power usage and SOC. So I am going to take this up with the Midnite people. Ha. This topic almost wraps back to the original reason I made the post

    I had forgotten about the Wizzbang Jr, very annoying. A normally closed (closed without power) relay driven by the Cellog 8 between the solar panels and the Midnite would be one way to go. A slightly more complex approach is to use the remote battery temperature sensor input to the Midnite to simulate a battery temperature fault. I can give you some details if you are interested.
    ... yesterday I individually charged 16 of the 32 cells to 3.60. Today I put the array back together, and this afternoon will put a moderate load on (standup freezer) and see what happens. No balancers. I did measure the cell voltages after sitting overnight and they are running from 3.487-3.564 (77mv dif).
    <

    If cells are left overnight with no load on them and the resting voltage of the cells is > 3.4V they are balanced to better than ~0.5%. If you had a scenario where the voltage range was 3.45V-3.38V they would be slightly out of balance. If the voltage range was 3.40V-3.32V they would be out of balance.

    When charging, the cell voltage will be around 50mV higher at low charge rates than after the cell has been allowed to rest for a while. If the cells are charged to > 3.45V they drop down to around 3.4V over a few days.
    I was actually going to ask you how your charger does capacity testing. Does it have a dedicated feature that will cycle the battery and report an AH number? If not it occurred to me that you could simply time how long it takes each cell to go from, say, 3.45 to 2.8, and compare the results across all cells to at least get an idea about relative capacity. Or you could discharge all cells at a fixed current rate and calculate AH?
    Yes my charger tells you the total Ah on the display and also logs the cell voltages and Ah every two seconds. I would have thought that the units you bought would have showed the total Ah on the display?

    Discharging at a fixed current rate and measuring the time it took for the battery to drop to 2.5V would be  one way to calculate the battery capacity. A word of warning, you must have some way of terminating the discharge when the battery voltage gets to 2.5V otherwise the cell voltage will keep going down to 0V which will damage the cell.

    Simon
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    karrak said:

    Yes my charger tells you the total Ah on the display and also logs the cell voltages and Ah every two seconds. I would have thought that the units you bought would have showed the total Ah on the display?

    Discharging at a fixed current rate and measuring the time it took for the battery to drop to 2.5V would be  one way to calculate the battery capacity. A word of warning, you must have some way of terminating the discharge when the battery voltage gets to 2.5V otherwise the cell voltage will keep going down to 0V which will damage the cell.

    Simon
    So far neither unit has displayed AH. However, I have not attempted to put them in a discharge/charge cycle yet (which I assume is needed to determine AH) due to the fact that with these units it is so time consuming.

    Both chargers have the ability to set the voltage at which discharge terminates.

    -------------------------------
    Having said that, I did connect the pack to a charger today, just to see how the pack behaved before considering other steps to further balance. What happened surprised me. I want to video the results to show you guys, so will repeat my experiment tomorrow and post the video. All I will say for now is that the shadow of suspicion has now moved from the balancers to the batteries themselves.

    jim

    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • Raj174
    Raj174 Solar Expert Posts: 795 ✭✭✭✭
    edited February 2018 #110
    @LostinSpace

    I think the best way to go would be to discharge the full bank connected in series/parallel about 15 amp hours. Then check cell voltages under charge. You want to see them around 3.4 volts without the highest going over 3.55. Then connect them all in parallel. Let them balance for a few hours, then use a 3.65 volt charger connected positive on one end to negative on the other to slowly bring them up to 3.6 volts. Be sure to check cell voltage while charging to make sure the cells with the charger clips are not rising too fast. If they, are then move the clips to lower cells. If you have buss bars, it may not be an issue. I am sure there are other ways to do this. I really can only tell you what has worked for me. Simon may have a better approach, maybe he will chime in.

    Rick
    4480W PV, MNE175DR-TR, MN Classic 150, Outback Radian GS4048A, Mate3, 51.2V 360AH nominal LiFePO4, Kohler Pro 5.2E genset.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited February 2018 #111
    Raj174 said:
    @LostinSpace

    I think the best way to go would be to discharge the full bank connected in series/parallel about 15 amp hours. Then check cell voltages under charge.

    Rick

    You know that is probably what I should have done today. I did connect the charger to what was ostensibly a fully charged pack... even though some cells had drifted down to the 3.48x area. That was silly. So all that happened that I was going to video is that two cells immediately went to full charge (3.60) within 10 seconds under a 13A charge while others were still in the low 3.5's. But now I am thinking this is just a normal reflection of variations in cell capacity. Naturally I stopped the charger after that 10 second period.

    So given what Simon said about my cells being in balance now, it seems the next thing to do is as you suggest and drain some charge from the pack and proceed from there.


    -------------------------------------
    NOTE: this evening I turned on the inverter and am running my freezer overnight (~200W). Pack voltage just prior to switching on the inverter was notably down to 3.331-3.351 from the earlier 3.487-3.564 reading, although the difference in cell voltage narrowed from the earlier 77mv reading to a 20mv difference.




    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • Raj174
    Raj174 Solar Expert Posts: 795 ✭✭✭✭
    edited February 2018 #112
    @LostinSpace

    I agree with your appraisal of what is happening with the cell voltage. A reasonable analogy might be to imagine each cell as a long neck wine bottle, all with a little different capacity, all proportional in shape but varying in size ranging from 30 to 34 ounces. When charging in series, all receive almost exactly the same amount of liquid at the same rate, and as the liquid  reaches the tapered neck of the smaller bottles the level will rise faster than the others. So obviously the smaller bottles will overflow first. To balance the liquid level a tube is connected between the bottles allowing flow from bottle to bottle. This is connecting them in parallel. If this is done with all the bottles full, the smaller bottles will overflow.  :*   Not good. So you really want to fill the bottles until the smallest bottle is almost full while in series, then connect them in parallel and wait for the fluid levels to even out. Once they are level, you'll want to feed one bottle slowly until the even level reaches the top of the smallest bottle and congratulations, your case of wine is now balanced.   :)

    The thing is, you have a lot of cells to monitor. And as you know the higher the voltage gets the faster it will rise. So you may get busier than a one armed wallpaper hanger.  :)  

    Rick

    4480W PV, MNE175DR-TR, MN Classic 150, Outback Radian GS4048A, Mate3, 51.2V 360AH nominal LiFePO4, Kohler Pro 5.2E genset.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    I think I will switch to balancing wine bottles. They don't make power but at least one can drink the overflow.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    edited February 2018 #114
    The wine bottle analogy is good for a system with a capacitor based active balancer.  The balancing flow rate between bottles/cells depends on the difference in height/voltage.    A inductor based active balancer is different - more like someone moving a fixed flow (say 1 cc/sec) between bottles.   A top balancing bleed/shunt balancer is like a small hole near the top of each bottle - dumping a little wine when the level is too high.  Any of the balancing methods can be overwhelmed by filling too fast - you need to slow down when any bottle is close to overflowing.

    I'm surprised that people still have to worry about cell balancing.   Should be able to just buy a combined charger/balancer/BMS that handles everything.  And can be powered from a variety of sources (say 30-150V AC or DC  with optional MPPT).

    I am available for custom hardware/firmware development

  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    Raj174 said:
    I think the best way to go would be to discharge the full bank connected in series/parallel about 15 amp hours. Then check cell voltages under charge. You want to see them around 3.4 volts without the highest going over 3.55. Then connect them all in parallel. Let them balance for a few hours, then use a 3.65 volt charger connected positive on one end to negative on the other to slowly bring them up to 3.6 volts. Be sure to check cell voltage while charging to make sure the cells with the charger clips are not rising too fast. If they, are then move the clips to lower cells. If you have buss bars, it may not be an issue. I am sure there are other ways to do this. I really can only tell you what has worked for me. Simon may have a better approach, maybe he will chime in.

    I agree with this for the first commission charge before you have put the battery together but don't think it is worth the effort and risk of damaging the cell terminal threads disassembling and reassembling the battery.

    If you have an assembled battery and a BMS that can shut down the charge controller I would just monitor the cell voltages when the charge controller is about to go to float and adjust the high and low cells with a load or constant current power supply placed on the individual cells for a fixed amount of time to remove a set % of the cell capacity. This can be done over a number of days or weeks. It is made really easy if you have a system to log the individual cell voltages. That is one reason why I built my BMS. As @jonr has said it is a real pain that no one has made a DIY charge controller with inbuilt BMS. At least now most good charge controllers and inverters can be shutdown via an external signal from a BMS.
     
    There are now some cheap Chinese BMS units which will monitor the cell voltages. I have reverse engineered the communication protocol of these units and have written code to interface these units to my BMS.

    Simon

    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    Well here is today's news. After leaving a load on overnight I found the cells within a 20mv range this morning, centering around 3.35x.

    I disconnected the load, then charged all day from the solar, going through a bulk, absorb, and float cycle. At the end of the cycle cell voltages were more than 100mv different, from 3.401-3.531 (see photos).

    The charge controller was set to go absorb at 55.2 (3.45v). One nice thing is that at float no cell was in the 3.6+ danger area.


    Sorry for the crappy shots.

    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    edited February 2018 #117
    Well here is today's news. After leaving a load on overnight I found the cells within a 20mv range this morning, centering around 3.35x.

    I disconnected the load, then charged all day from the solar, going through a bulk, absorb, and float cycle. At the end of the cycle cell voltages were more than 100mv different, from 3.401-3.531 (see photos).

    The charge controller was set to go absorb at 55.2 (3.45v). One nice thing is that at float no cell was in the 3.6+ danger area.

    Great!

    What conditions was your charge controller programmed to change to Float charging, end amps?, time etc.

    4.2 volts + is the danger zone. It won't do any harm taking the cell voltages above 3.6V for a small amount of time. It will reduce the battery lifespan holding the voltage above 3.6V for any length of time

    Simon
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    edited February 2018 #118
    The wiz bang has not been operating lately for some reason. I'm afraid I have to replace it. So absorb transitions to float after 60 min at present.


    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    BTW - I have been talking to boB at Midnite about updating the controller to handle two inputs. Doesn't look like it's going to happen soon, but there is a solution developing using a Rasberry Pi and talking to the Classic via the Ethernet jack and MODBUS. Here is the thread if you want to have a look: http://midniteftp.com/forum/index.php?topic=3973.0. i am '5 minutes to midnite'.

    So I don't know if your temperature sensor trick or the Pi would be the quickest path. Need to do something though.


    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.
  • karrak
    karrak Solar Expert Posts: 326 ✭✭✭✭
    Such a long absorb time at 54.8V (3.425/cell) Will get you very close to 100%. I would change the time to 30 minutes.

    If you now hook your balancers up they should fine balance your pack. It may take a few days or longer. I wouldn't leave the balancers on permanently.

    Simon
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • LostinSpace
    LostinSpace Registered Users Posts: 97 ✭✭
    Check.
    Qty 24 Sharp 153W poly panels, Midnite MNPV6 combiner, Magnum MS-4448-PAE inverter with RC50 remote, Magnum MMP Mini panel, Midnite Classic 150 charge controller, Qty 32 LiFePO4 72AH cells in 2 string 48V array. BMS.