diy solar

diy solar

16s chargery BMS numbers not matching reality / active balancer

dudeinthewoods

New Member
Joined
Apr 7, 2020
Messages
111
I've got 16s LiFePo4 (48v) that I built using a chargery BMS.

I also thought I'd try out an active balancer. I picked up a 2a Chinese one and after some work tracked down the software.


My issue is the BMS and the AB are showing different numbers. The AB consistently reads .4v lower (total, not per cell).

Also the AB balanced it within 5mv but the BMS thinks it's off by 180mv.

I checked the voltage with my meter and it's matching the AB not the chargery.

I'm using the factory shunt so not sure what the deal is if I should calibrate the chargery or what.

I'll keep messing with the AB and post on it individually.
 
I'm suspecting the Chargery is reading 1 cell wrong. Either by calibration or a bad connection. But it should have nothing to do with shunt. Also with the external AB it could be drawing 1 cell down while balancing. Try checking voltages on each cell with AB turned off. These are all just ideas to find source of problem
 
I've got 16s LiFePo4 (48v) that I built using a chargery BMS.

I also thought I'd try out an active balancer. I picked up a 2a Chinese one and after some work tracked down the software.


My issue is the BMS and the AB are showing different numbers. The AB consistently reads .4v lower (total, not per cell).

Also the AB balanced it within 5mv but the BMS thinks it's off by 180mv.

I checked the voltage with my meter and it's matching the AB not the chargery.

I'm using the factory shunt so not sure what the deal is if I should calibrate the chargery or what.

I'll keep messing with the AB and post on it individually.
For My further BMS Education plus my Full Charge on with Top Balancing Objective:

Fishing for further Chargery BMS8T Specifics here: What does AB mean? or what is an AB? Auto Balance? Is that the one that shows up in the Chargery BMS8T or 16T's configuration section labeled: "Balance Perameters" / within that section/ one can turn such perameter OFF or ON for charge cycles, discharge cycles, or for in storage cycles ??? and also adjust "Balance Start Voltage" to a figure between 3.00V to 3.40V . IS that what AB is referring to?

I am revisiting specific on my Chargery BMS8T with a few new questions as I seek to properly set one replacement cell into my first DIY 24v 280 Ah LiFePO4 build; and getting closer to my second physical Top Balance, ... hopefully eventually getting to a 3.65V cut off cell reading (briefly) for a full charge. I will then further Test the AH abilities of all 8 of LiFePO4 cells.

Wondering about: Which "start voltage" selection for individual cells gets the most balancing adjustments (that perameter adjusts between 3.00V to 3.40V)? Does "Start" @ ?V mean; Start Balancing when you charge a cell UP to that voltage, and then the balancing starts, or Start Balancing when Voltage gets Down to that lower Voltage on a discharge cycle? or maybe both if you select balance on charge and discharge ? and in Storage.

My curious thoughts re matching reality. I figure measuring cell voltage with a triple digit capable DMM at each LiFePO4 cell is the best indicator of reality on individual cell voltages (???); as compared to what I see in the Chargery BMS8T info pages on individual cell voltages. I just Noticed from a written down comparison: (and was not surprised) that all my cell voltage reading at my BMS were slightly lower than my DMM (digital multi meter) Voltage readings, ... AND (I was surprised) that the range of those cell voltage reading varied from - 0.005V to - 0.016V with some -0.008V and -0.010V contrasts in the spread of 8 cells. This is on the stock Chargery supplied wiring harness going between my BMS8T and each battery cell. Is that normal?

... Since I have a couple of other stock Chargery wiring harness from a BMS16T here, I wondering if a new set of harness wires with newly crimped connectors might make any difference ???, or if Chargery accepts such variations in Cell Voltage readings ??? ... Any thoughts or feedback will be appreciated :+) Bill
 
Last edited:
For My further BMS Education plus my Full Charge on with Top Balancing Objective:

Fishing for further Chargery BMS8T Specifics here: What does AB mean? or what is an AB? Auto Balance? Is that the one that shows up in the Chargery BMS8T or 16T's configuration section labeled: "Balance Perameters" / within that section/ one can turn such perameter OFF or ON for charge cycles, discharge cycles, or for in storage cycles ??? and also adjust "Balance Start Voltage" to a figure between 3.00V to 3.40V . IS that what AB is referring to?

I am revisiting specific on my Chargery BMS8T with a few new questions as I seek to properly set one replacement cell into my first DIY 24v 280 Ah LiFePO4 build; and getting closer to my second physical Top Balance, ... hopefully eventually getting to a 3.65V cut off cell reading (briefly) for a full charge. I will then further Test the AH abilities of all 8 of LiFePO4 cells.

Wondering about: Which "start voltage" selection for individual cells gets the most balancing adjustments (that perameter adjusts between 3.00V to 3.40V)? Does "Start" @ ?V mean; Start Balancing when you charge a cell UP to that voltage, and then the balancing starts, or Start Balancing when Voltage gets Down to that lower Voltage on a discharge cycle? or maybe both if you select balance on charge and discharge ? and in Storage.

My curious thoughts re matching reality. I figure measuring cell voltage with a triple digit capable DMM at each LiFePO4 cell is the best indicator of reality on individual cell voltages (???); as compared to what I see in the Chargery BMS8T info pages on individual cell voltages. I just Noticed from a written down comparison: (and was not surprised) that all my cell voltage reading at my BMS were slightly lower than my DMM (digital multi meter) Voltage readings, ... AND (I was surprised) that the range of those cell voltage reading varied from - 0.005V to - 0.016V with some -0.008V and -0.010V contrasts in the spread of 8 cells. This is on the stock Chargery supplied wiring harness going between my BMS8T and each battery cell. Is that normal?

... Since I have a couple of other stock Chargery wiring harness from a BMS16T here, I wondering if a new set of harness wires with newly crimped connectors might make any difference ???, or if Chargery accepts such variations in Cell Voltage readings ??? ... Any thoughts or feedback will be appreciated :+) Bill
AB is active balancer. From my experiences the Chargery will vary from actual DVM measurements in the range which you indicate. I think it has to do with the unit having only 12 bits of precision instead of 16. For my purposes that margin is acceptable but I am not as anal as some people who want it exact. I do not think you need to change the leads that is "normal" from my experience. weather that is acceptable is up to the end user. I recommend only balancing during charge and maybe storage. I have actually found that charging cells till one reaches max value then not restarting charging until all cells are below the natural sinking voltage works better than passive balancing.

The balance start at parameter means it will balance only when cell voltages are above that set number I am not sure weather it means all cells need to be above or just 1 cell, I think just one cell though.
 
AB is active balancer. From my experiences the Chargery will vary from actual DVM measurements in the range which you indicate. I think it has to do with the unit having only 12 bits of precision instead of 16. For my purposes that margin is acceptable but I am not as anal as some people who want it exact. I do not think you need to change the leads that is "normal" from my experience. weather that is acceptable is up to the end user. I recommend only balancing during charge and maybe storage. I have actually found that charging cells till one reaches max value then not restarting charging until all cells are below the natural sinking voltage works better than passive balancing.

The balance start at parameter means it will balance only when cell voltages are above that set number I am not sure weather it means all cells need to be above or just 1 cell, I think just one cell though.
just the lowest cell voltage over Start Voltage setting, BMS start to balance, it is effective in charge, in discharge or in Storage.
 
I've got 16s LiFePo4 (48v) that I built using a chargery BMS.

I also thought I'd try out an active balancer. I picked up a 2a Chinese one and after some work tracked down the software.


My issue is the BMS and the AB are showing different numbers. The AB consistently reads .4v lower (total, not per cell).

Also the AB balanced it within 5mv but the BMS thinks it's off by 180mv.

I checked the voltage with my meter and it's matching the AB not the chargery.

I'm using the factory shunt so not sure what the deal is if I should calibrate the chargery or what.

I'll keep messing with the AB and post on it individually.
disconnect AB BMS and other BMS, don't charge and don't discharge, only conenct all cells to BMS, turn off balance on Chargery BMS16T, measure all cell voltage with DMM, and compare DMM reading with cell voltage reading on BMS. if each cell voltage difference is within 10mV, means the BMS reading is normal.
during passive balancing or Active Balancing, the cell vorlage always changed, the sample time on different BMS is different, the cell voltage reading is different too.
so when balancing stop, then check cell voltage reading is correct or not.
 
And remember that specificity does not equate to accuracy.

Is your DMM a Fluke brand or a Harbor Freight freebie?

Or as they say, a man with two watches never knows what time it really is ?
 
AB is active balancer. From my experiences the Chargery will vary from actual DVM measurements in the range which you indicate. I think it has to do with the unit having only 12 bits of precision instead of 16. For my purposes that margin is acceptable but I am not as anal as some people who want it exact. I do not think you need to change the leads that is "normal" from my experience. weather that is acceptable is up to the end user. I recommend only balancing during charge and maybe storage. I have actually found that charging cells till one reaches max value then not restarting charging until all cells are below the natural sinking voltage works better than passive balancing.

The balance start at parameter means it will balance only when cell voltages are above that set number I am not sure weather it means all cells need to be above or just 1 cell, I think just one cell though.
Thanks Craig for your more experienced perspective. Taking such in with my current view, I just readjusted my "Balance Perameters" to include: start at 3.0V when I have a 25mv diff (had previous changed default diff to 15mv / now thinking with my noticing a 16mv max diff between my DMM reading at Battery vs at reading in my BMS, why have BMS in balance mode at 15mv diffs) ... and I turned balance function on for charge and storage only, and off for discharge (previously chose to turn all three on). Not a giant deal changer, but those were my most recent Chargery BMS8T configuration choices after learning a little bit more.
 
From my experiences the Chargery will vary from actual DVM measurements in the range which you indicate. I think it has to do with the unit having only 12 bits of precision instead of 16. For my purposes that margin is acceptable but I am not as anal as some people who want it exact.

12bits is way more than enough bits, but says nothing at all about accuracy. I just built a battery with a Chargery BMS and rather appalled at the lack of precision and filtering.

For reference, I design power electronics for industrial applications and very familiar with what it takes to sense voltage precisely. The analog to digital converter itself certainly has importance but the more important elements are the front-end filtering and voltage reference to get precision results. For a BMS application, 12bits is really more than you need already - but in the case of Chargery - it needs some filtering/averaging along with a better precision voltage reference. The voltages in real life are very stable, yet the numbers on the display are jumping all over which is a good indication that the hardware is not up to the task and the software is doing nothing to overcome the bad hardware.

Putting a 16bit A/D into a crappy analog system just captures the noise and errors more precisely - useless.

I am about to throw the Chargery unit into the e-cycle bin.

NOTE: battery system are not my specialty at all, just the electronics and software that manages power systems.
 
12bits is way more than enough bits, but says nothing at all about accuracy. I just built a battery with a Chargery BMS and rather appalled at the lack of precision and filtering.

For reference, I design power electronics for industrial applications and very familiar with what it takes to sense voltage precisely. The analog to digital converter itself certainly has importance but the more important elements are the front-end filtering and voltage reference to get precision results. For a BMS application, 12bits is really more than you need already - but in the case of Chargery - it needs some filtering/averaging along with a better precision voltage reference. The voltages in real life are very stable, yet the numbers on the display are jumping all over which is a good indication that the hardware is not up to the task and the software is doing nothing to overcome the bad hardware.

Putting a 16bit A/D into a crappy analog system just captures the noise and errors more precisely - useless.

I am about to throw the Chargery unit into the e-cycle bin.

NOTE: battery system are not my specialty at all, just the electronics and software that manages power systems.
Wonder what BMS in similar price or bit more price range you are considering next? Seems like more smart / configurable options are coming on line. One plus aspect I like about Chargery is its' ability to run high current through an external shunt (w 3 select options for your needs), vs being limit by questionable "amp ratings" for running big amps through the BMS itself. I could see liking a different BMS, but just on my 1st try. Glad we got this forum to compare notes :+)
 
I don't have an immediate alternative to the Chargery at the moment. I purchased it as an experiment to learn more about large battery systems. In that regard, it is rather nice in concept. Very programmable with a dedicate pendant. All that is missing is the accuracy in general.

I would be happy with a solution that has higher performing hardware and some option to program, monitor, log in detail.

Before I connected the Chargery, I had the pack connected to my test bench with has 20 channels of 6.5 digit DMM's, 6 channels of precsion power supplies, 8 channels of precision DC electroncic loads all automated through software that I have written.

Before I go much further, I plan to examine each cell in the pack to get detailed specs as a baseline to compare results with other BMS's. That way, I have a realistic way to grade the various BMS's in terms of functional performance. The jury is out on the reliability question, so I plan to take my Chargery apart and critique the hardware design.
 
I don't have an immediate alternative to the Chargery at the moment. I purchased it as an experiment to learn more about large battery systems. In that regard, it is rather nice in concept. Very programmable with a dedicate pendant. All that is missing is the accuracy in general.

I would be happy with a solution that has higher performing hardware and some option to program, monitor, log in detail.

Before I connected the Chargery, I had the pack connected to my test bench with has 20 channels of 6.5 digit DMM's, 6 channels of precsion power supplies, 8 channels of precision DC electroncic loads all automated through software that I have written.

Before I go much further, I plan to examine each cell in the pack to get detailed specs as a baseline to compare results with other BMS's. That way, I have a realistic way to grade the various BMS's in terms of functional performance. The jury is out on the reliability question, so I plan to take my Chargery apart and critique the hardware design.
I'd be interested in your diagnostics of the Chargery and potential to improve upon it. I know Jason is open to hearing good critical analysis and potential applicable solutions / suggestions. They are working on a new line and so, such could be most timely.
 
I've got 16s LiFePo4 (48v) that I built using a chargery BMS.

I also thought I'd try out an active balancer. I picked up a 2a Chinese one and after some work tracked down the software.


My issue is the BMS and the AB are showing different numbers. The AB consistently reads .4v lower (total, not per cell).

Also the AB balanced it within 5mv but the BMS thinks it's off by 180mv.

I checked the voltage with my meter and it's matching the AB not the chargery.

I'm using the factory shunt so not sure what the deal is if I should calibrate the chargery or what.

I'll keep messing with the AB and post on it individually.
AB and chargery BMS connected same battery bank at the same time?
if not, AB can balance all cell voltage within 5mV, but chargery BMS don't balance it to 5mV and the cell voltage difference is 180mv, The fact is AB continue to balance till cell voltage difference is under 5mV, Chargery BMS is passilve balancing, due to internal temperauter protection, balancing will stop automatically and start balancing again, so the BMS don't stop balancing unless the cell volage difference under setting. if add cooling fan, the balancing can continue to do, the cell voltage difference will be under setting finally.
As we know, the active balancing is faster than passive balancing, especially for large balancing current with large cell volage difference and large battery capacity. we are developing new BMS with active balancing, the balancing is up to 10A and it is programmable.
thanks.
 
AB and chargery BMS connected same battery bank at the same time?
if not, AB can balance all cell voltage within 5mV, but chargery BMS don't balance it to 5mV and the cell voltage difference is 180mv, The fact is AB continue to balance till cell voltage difference is under 5mV, Chargery BMS is passilve balancing, due to internal temperauter protection, balancing will stop automatically and start balancing again, so the BMS don't stop balancing unless the cell volage difference under setting. if add cooling fan, the balancing can continue to do, the cell voltage difference will be under setting finally.
As we know, the active balancing is faster than passive balancing, especially for large balancing current with large cell volage difference and large battery capacity. we are developing new BMS with active balancing, the balancing is up to 10A and it is programmable.
thanks.
Are you able to speak to a time frame on this new product?

Make me a guinea pig. :cool:
 
12bits is way more than enough bits, but says nothing at all about accuracy. I just built a battery with a Chargery BMS and rather appalled at the lack of precision and filtering.

For reference, I design power electronics for industrial applications and very familiar with what it takes to sense voltage precisely. The analog to digital converter itself certainly has importance but the more important elements are the front-end filtering and voltage reference to get precision results. For a BMS application, 12bits is really more than you need already - but in the case of Chargery - it needs some filtering/averaging along with a better precision voltage reference. The voltages in real life are very stable, yet the numbers on the display are jumping all over which is a good indication that the hardware is not up to the task and the software is doing nothing to overcome the bad hardware.

Putting a 16bit A/D into a crappy analog system just captures the noise and errors more precisely - useless.

I am about to throw the Chargery unit into the e-cycle bin.

NOTE: battery system are not my specialty at all, just the electronics and software that manages power systems.
we just released new firmware for BMS8T, BMS16T and BMS24T. both main unit version and LCD version is V4.02

the update are as below,

On new version V4.02, new update:

  • Optimize current measurement, with 100A shunt, the BMS can detect 0.3A minimum, with 300A shunt, the BMS can detect 0.5A minimum. The AH and WH accuracy is improved at the same time.
  • Optimize cell voltage measurement, improve the cell voltage accuracy
  • Modify over current protection resume from automatically to by manual.
new manual is as below, more details please check the final page of manual.
New Firmware is here, https://chargery.com/update.asp

our testing result state clearly the current is very stable and accurate, if curent reading on BMS still have fluctuations, please email me to jasonwang3a@163.com, i will try my best to fix it, maybe need modify hardware.

new BMS8P, BMS16P, BMS24P will be with RS485, and CAN communication, the cooling fan will be added, new hardware and new program, continue to balance will be realized, for battery with large capacity difference, large impedance difference and large cell volage difference will be resumed at the shortest time, 1A balancing current will remove 10AH difference within 10 hours.

the cell IR measurement is still an importang tool, it can indicate cell actual difference, it is easy to tell you which cell is bad (lower capacity and higher impedance), and need to exchange.

about parameter setting, don't setup the protection point to the highest or the lowest as descriped in cell/battery datasheet, it is abosulutely maximum or minimum value, and don't reach it abolutely, that is to say, if over it or under it, the cell or battery is dangeous, so it is not suitable in actual using, unless you need test cell/battery actual capacity, because the capacity listed on cell / battery datasheet is tested according these maximum or minimum conditions.

good luck and thanks for support.
 
I don't have an immediate alternative to the Chargery at the moment. I purchased it as an experiment to learn more about large battery systems. In that regard, it is rather nice in concept. Very programmable with a dedicate pendant. All that is missing is the accuracy in general.

I would be happy with a solution that has higher performing hardware and some option to program, monitor, log in detail.

Before I connected the Chargery, I had the pack connected to my test bench with has 20 channels of 6.5 digit DMM's, 6 channels of precsion power supplies, 8 channels of precision DC electroncic loads all automated through software that I have written.

Before I go much further, I plan to examine each cell in the pack to get detailed specs as a baseline to compare results with other BMS's. That way, I have a realistic way to grade the various BMS's in terms of functional performance. The jury is out on the reliability question, so I plan to take my Chargery apart and critique the hardware design.
I am curious; did you switch wire set from BMS to battery cells to rule that out a possible? are you on update v 4.1? or know update v 4.2 just came out? Here's a link I read just 10 minutes ago. Already downloaded the new v4.2 manual (Thanks Steve S and Jason), plus just download the new firmware files I will use to flash update my Chargery BMS8T. I will be flashing my main unit with 300amp shunt, and the LCD unit to v 4.2 soon to see some improvements. I know this BMS is not totally accurate to the 0.000 it shows on volt readings, but has been good enough for me, plus appears to be improving with firmware updates. Food for Thoughts, and will welcome pictures of you tear down or insight info. if yo go that route wo further testing. I just read your recent post so do not know full story. I once compared voltage of my Chargery BMS v 4.0 to voltages measure via a DMM, and then a BattGo cell balancer quick look see v readout; and only two cell reading matched my DDM double digit to 0.00 DMM on the BattGo, w reads varied higher by .008v or lower by .007v / and for differences between my BMS and my DMM, nothing matched exactly, while varying by up to -0.016v to + 0.005v ... For me, after comparing note with Craig, I decided that was accurate enough for my needs. Glade to have the Chargery BMS option, and to also see their efforts at ongoing improvements. ... My two cents :+) Bill
 
we just released new firmware for BMS8T, BMS16T and BMS24T. both main unit version and LCD version is V4.02

the update are as below,

On new version V4.02, new update:

  • Optimize current measurement, with 100A shunt, the BMS can detect 0.3A minimum, with 300A shunt, the BMS can detect 0.5A minimum. The AH and WH accuracy is improved at the same time.
  • Optimize cell voltage measurement, improve the cell voltage accuracy
  • Modify over current protection resume from automatically to by manual.
new manual is as below, more details please check the final page of manual.
New Firmware is here, https://chargery.com/update.asp
our testing result state clearly the current is very stable and accurate, if curent reading on BMS still have fluctuations, please email me to jasonwang3a@163.com, i will try my best to fix it, maybe need modify hardware.

new BMS8P, BMS16P, BMS24P will be with RS485, and CAN communication, the cooling fan will be added, new hardware and new program, continue to balance will be realized, for battery with large capacity difference, large impedance difference and large cell volage difference will be resumed at the shortest time, 1A balancing current will remove 10AH difference within 10 hours.

the cell IR measurement is still an importang tool, it can indicate cell actual difference, it is easy to tell you which cell is bad (lower capacity and higher impedance), and need to exchange.

about parameter setting, don't setup the protection point to the highest or the lowest as descriped in cell/battery datasheet, it is abosulutely maximum or minimum value, and don't reach it abolutely, that is to say, if over it or under it, the cell or battery is dangeous, so it is not suitable in actual using, unless you need test cell/battery actual capacity, because the capacity listed on cell / battery datasheet is tested according these maximum or minimum conditions.

good luck and thanks for support.

Great to see some fresh firmware for my existing unit - I will flash the unit at the end of this week when I tear down my test setup.

The P models with 485 and CAN communication sounds great!
 
Back
Top