XiaoXiang BMS Spontaneously resets Remaining Capacity display

[OnTheRoadAgain]

I have a XiaoXiang BMS which is basically identical to the one OverkillSolar sells but you're better off paying the small difference for the OKS BMS imo for the warranty.

What happened was I fully charged the battery to 3.65v per cell then reset the Remaining Capacity to full, Then discharged the battery to storage voltage of 3.300v (the voltage it arrived at new at and the voltage at which the parameters view says 39% SOC)and it said 54Ah remaining.

Today I turned on the BMS (having not handled it since then) and I was surprised to see that the Remaining Capacity had reset to the full capacity, even though the battery was sitting at 39% SOC. I don't know of any way to get it back right other than to fully charge the battery and reset it again at full SOC. Big pain.

I also have a very hard time getting it connect to Bluetooth and have to use the "FIX" feature (which usually does not work) so I end up "refreshing" the list over and over and over and over until usually I finally get it listed. Location is on. Sometimes it can take twenty refreshes and or a power down. I checked the connections.

I can't see this behavior as normal and just wanted to run this by before it goes back.

Thanks

 

[HRTKD]

That doesn't sound right to me. If you run the battery through a few charge/discharge cycles the BMS usually figures out the 100% state of charge. You don't need a deep discharge to make this happen.

 

[OnTheRoadAgain]

That doesn't sound right to me. If you run the battery through a few charge/discharge cycles the BMS usually figures out the 100% state of charge. You don't need a deep discharge to make this happen.

Thanks Jim. So even if you only use a few Ah and then just replace the used Ah it will recalculate ?
I haven't seen my Xiaoxiang BMS do that yet, but I haven't cycled it much either.

Trying not to run them through too may unnecessary cycles.
Still not exactly sure why the remaining capacity reset itself. Also very difficult to connect to BT at times. was thinking it could be a BMS defect.

The other solution is to get the USB communications dongle so it can be connected to a PC and set manually through the PC software since the app doesn't have that capability AFAIK.

It's possible that installing the app on another device will automatically reset everything to factory default.
Trying to remember if I added the xiaoxing BMS app to another device before or after the reset took place.

 

[HRTKD]

Thanks Jim. So even if you only use a few Ah and then just replace the used Ah it will recalculate ?

That's how it worked on mine.

 

[OnTheRoadAgain]

Did it again, overnight.
Went to bed with it showing 60.63Ah remaining capacity. Woke up it says full capacity without charging.
I do not believe that the BMS is supposed to lose the Remaining Capacity count unless powered down?
This one was connected to the battery overnight, wasn't touched.

I think this means it's defective.

 

[curiouscarbon]

Dear @OnTheRoadAgain after having the Ah remaining value jump around on the JBD BMS over the course of a discharge, I grew bitterly distrustful of its veracity. Especially with my 8Ah 3S JBD pack, it struggles to keep a stable hold on Ah remaining figure.

I’m in the process of putting together a DIY shunt monitor with a cheap 500A 75mV shunt and doing arduino instead of using the LCD display. Using 24-bit voltage amplifier to read the voltage drop, it has +/- 40mV range. So at 40mV max it ought to register up to 266A in either direction before maxing out.

Anyways, just thought to share. The pack has not had any other issues like voltage dropping out, which lead me to conclude the SoC monitoring is just not that precise.. my 0.02usd

To be clear you are having other issues it sounds like, besides the Ah remaining jumping around, so maybe that unit is busted. I just also wanted to share that I trust my victron shunt a heck of a lot more than the Ah figure in xiaoxiangbms app… this hasn’t really changed even after multiple cycles. Bluetooth communication worked perfectly fine on those units in that case.

 

[OnTheRoadAgain]

Thank you.
I'm an Arduino fan myself. I have quite a bit of Arduino related hardware on hand and have worked with programming it a bit.
I would be very interested in your project if you cared to share.

Are you developing your own programming scripts to do this?

On my unit, the Remaining Capacity doesn't just jump around, it resets to full capacity, as if there was an internal reset.
My thoughts were that maybe it's best to buy these from OverkillSolar because he tests them and stands behind them.

 

[HRTKD]

Just for grins, check all your connections. A loose connection can cause all sorts of issues. However, most of these issues are seen when under load. You're seeing problems at night with no load so that would lead me to suspect that it's not a connection issue.

Are the safety features of the BMS working? If you set the High Voltage Disconnect to X volts and your charger to X + .1 volts, will the BMS disable charging?

Where I'm going with this is if everything but the state of charge is working, do you really want to spend the money to replace the BMS? Most of us don't rely on the stage of charge as reported by the BMS. We may look at it every once in a while, but we use a separate battery monitor for reporting the "official" state of charge.

 

[curiouscarbon]

my fuzzy memory says it was a jumping around in an unexpectedly downward fashion, made me think it was relying on a linear interpolation between the configured 20/40/60/80 voltage points or whatever thresholds.. maybe it does snap into a better coulomb counting mode after a few cycles and i was merely impatient.

random thought: is the cycle count resetting too? if so then it sounds broken or maybe faulty lead in harness.

the JST-XH plug sketches me out like molex, but we are not here to fix that.. anyways i’ve seen examples of people here noticing that one or more leads going into the bms at the balance lead harness thing.. that one jiggled and needed to be re seated or something. more random thoughts 0.02usd
?

Most of us don't rely on the stage of charge as reported by the BMS. We may look at it every once in a while, but we use a separate battery monitor for reporting the "official" state of charge.

edit:

Arduino fan

woohoo! will make a separate thread eventually and post info about results. short mode: (ab)using HX711 load cell amplifier(sparkfun) green and white pins are just inputs into a differential 24-bit ADC with at least 80 samples per second. still figuring out capacitor parameter to lower bandwidth for slow 80 samples per second rate. hookup guide in link has the lowdown for pins and stuff. “cheap” 500A 75mV shunt(amazon) it’s listed as 350A but it has 500A 75mV printed on the shunt itself. convenient spots to attach two sides of sense leads separately from high amp lugs. samd21 mcu (adafruit qt py) reads HX711 repeatedly and does time window averaging or whatever is appropriate. from there easy to Serial.println() to victory for computer logging on raspberry pi or something. Planning on using teensy4.1 for big stuff to have direct logging to SDHC card and other stuff. i’ve rambled enough,

cheers

 

[ereams65]

Have you confirmed all of your other settings to make sure they stored? May just be an app issue.

 

[OnTheRoadAgain]

Just for grins, check all your connections. A loose connection can cause all sorts of issues. However, most of these issues are seen when under load. You're seeing problems at night with no load so that would lead me to suspect that it's not a connection issue.

Are the safety features of the BMS working? If you set the High Voltage Disconnect to X volts and your charger to X + .1 volts, will the BMS disable charging?

Where I'm going with this is if everything but the state of charge is working, do you really want to spend the money to replace the BMS? Most of us don't rely on the stage of charge as reported by the BMS. We may look at it every once in a while, but we use a separate battery monitor for reporting the "official" state of charge.

Hello Jim,

The problem is that if it resets everything, then I cannot set parameters specific to my pack.
It keeps defaulting back to default settings.

The safety features do seem to work. It properly stop charging when it detected any cell at 3.65v
It does seem to be working for safety features, but seems to totally reset the stored memory randomly.

I agree a bad connection could do this. All connections seem good. Nothing suspicious found.

I'm talking with another member about alternative monitoring as well as thinking about alternatives myself.

What I didn't know (until CuriousCarbon mentioned it) was that other people are having this same problem. No one else has mentioned it yet so maybe it's rare and most people never have this issue?

What do you use to store your battery info?

Would these be good for this?

 

[OnTheRoadAgain]

my fuzzy memory says it was a jumping around in an unexpectedly downward fashion, made me think it was relying on a linear interpolation between the configured 20/40/60/80 voltage points or whatever thresholds.. maybe it does snap into a better coulomb counting mode after a few cycles and i was merely impatient.

random thought: is the cycle count resetting too? if so then it sounds broken or maybe faulty lead in harness.

the JST-XH plug sketches me out like molex, but we are not here to fix that.. anyways i’ve seen examples of people here noticing that one or more leads going into the bms at the balance lead harness thing.. that one jiggled and needed to be re seated or something. more random thoughts 0.02usd
?

edit:

woohoo! will make a separate thread eventually and post info about results. short mode: (ab)using HX711 load cell amplifier(sparkfun) green and white pins are just inputs into a differential 24-bit ADC with at least 80 samples per second. still figuring out capacitor parameter to lower bandwidth for slow 80 samples per second rate. hookup guide in link has the lowdown for pins and stuff. “cheap” 500A 75mV shunt(amazon) it’s listed as 350A but it has 500A 75mV printed on the shunt itself. convenient spots to attach two sides of sense leads separately from high amp lugs. samd21 mcu (adafruit qt py) reads HX711 repeatedly and does time window averaging or whatever is appropriate. from there easy to Serial.println() to victory for computer logging on raspberry pi or something. Planning on using teensy4.1 for big stuff to have direct logging to SDHC card and other stuff. i’ve rambled enough,

cheers

I will look for a thread you start on this

 

[OnTheRoadAgain]

Have you confirmed all of your other settings to make sure they stored? May just be an app issue.

Yes. The BMS basically does a full reset as if it was powered down completely.
I'm using 3.1.1026 from OverkillSolar's website which I think most people are also using?

The problem is that if it resets everything, then I cannot set parameters specific to my pack.

 

[OnTheRoadAgain]

The BMS that is resetting is not from OverkillSolar. It is a separate one I bought.

But I do have one from OverkillSolar on hand and I'm going to remove the suspect faulty one and install the overkill BMS in it's place and see if that resolves the issue.

I would very much like to ask Steve at Overkill if he has observed this glitch

 

[HRTKD]

What do you use to store your battery info?

Technically, I don't "store" my battery history. My Victron BMV-712 doesn't have a memory. If I wanted memory I would step up to the Cerbo GX. My brain is the memory at this time, which means I have no history or a very spotty history.

 

[OnTheRoadAgain]

Replaced the unknown vendor Xiaoxiang BMS with OverkillSolar BMS to see if this issue is specific to one BMS or both that I have.

 

[SV Angular Momentum]

I'm having the same problem, new pack of Eve cells from Shenzen Basen that pulled 290 ah on test. The bms from overkill works perfectly for protection but the capacity is all over the place. Just looked here for a solution but after reading will probably get a powermon shunt for monitoring. Also discovered the app has to come from the overkill website the playstore version doesn't work. Any advice is welcome.

 

[HRTKD]

The version of the app on the Google Play store is really old. The version that most had been using is 1015 I think. I had no issues with that version, but still upgraded to the newer one from the Overkill Solar site.

If it's an app issue, then consider rolling back to the 1015 version of the app.

 

[OnTheRoadAgain]

I was thinking about this one too, maybe even better?......

Charge-Discharge Monitor, DROK 0-90V 100A DC Ammeter Voltmeter, Battery Capacity Amp-Hour Watt-Hour Power Time Multimeter, LCD

It's probably overkill...but to be honest, I'm a tinkerer and enjoy it, so it's a hobby.

 

[Gazoo]

Did it again, overnight.
Went to bed with it showing 60.63Ah remaining capacity. Woke up it says full capacity without charging.
I do not believe that the BMS is supposed to lose the Remaining Capacity count unless powered down?
This one was connected to the battery overnight, wasn't touched.

I think this means it's defective.

I have never had that problem. Capacity % and ah's are fairly accurate on my Overkill. The only time things get thrown off is when I change a parameter setting. I never turn discharge off. I have turned off charging and have no problems with that.

I have had bluetooth connection issues after upgrading the app. Sometimes while charging or discharging and sometimes both. When I go to read the parameters it takes a long time. I am not sure why this happens. But after awhile the problem magically goes away. My guess is it is taking awhile for the app to read or sync with the BMS. Once it's working properly, the bluetooth connection remains stable.