4S BMS cutoff conditions and "reset"

bmcmullin

New member
Joined
Jun 12, 2019
Messages
18
I have constructed my first pack, 4S8Pwith a basic 4S BMS,the $5 type from eBay - Model:HX-4S-F30A(couldn't post a link). The batteries are Li-Ion18650s. All the voltages are proper. ~16.4V on B+/B- as well as P+/P-. B1,B2 and B3 are ~4.17V.

I built the pack to replace a 12V SLA battery used in a golf trolley. In hindsight, I should have used Li-Fe 3.2V cells.

Accepting that my pack is out of spec with the intended load, I performed some tests with two different 12V loads. The first was with the golf trolley which has a 200W motor, an LCD display, a switch and a speed control. The cart worked "on the bench" just fine during a short few minute test. As you might expect, the motor spins faster if I turn up the potentiometer. As for the LCD and electronics (the cart has a USB port on it to charge a GPS), there must be a voltage regulator that drops the voltage to 5V. My research says such regulators are designed to accept input voltages well over 14V, so the output of my pack at ~16V doesn't pose a problem.

The second load I tested is where I encountered something that I need help with. I connected a 12V 400W DC to AC inverter and plugged a 75W light into the AC side of the inverter. My pack powered the inverter briefly for about 15 seconds then shut down. What I can't figure out is whether it was the inverter shutting down because of overvoltage or the BMS of the pack. After the shutdown, the voltage on P+/P- was <1V. Even after disconnecting the leads to the load the P+/P- voltage remained <1V. The only way I could reset the BMS was to connect it to the charger. Adding to the confusion, I just completed the same test again but the shutdown didn't happen. It's been running for over 5 minutes. The voltage on P+/P- while running declined from ~16V at start to ~15.5V around the 3 minute mark. If it was the inverter that shut down on overvoltage, how is it that the BMS is triggered into a state that partially shuts down the voltage to P+/P-?

Appreciate any guidance or insight.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,399
The inverter wouldn't have caused the BMS to trigger. Most likely the BMS sensed and overcurrent event and shutdown. I would assume this was because the inverter was initially pulling more current than what the BMS was "actually" built as (not what the label states). It may have been really close to the max, and then at the 15 second mark (or there abouts), the inverter pulled a little more and went over the threshold of the BMS.

Putting a charger on the BMS is to reset it is normal. You could also have just connected the battery terminals to the P-/+ as well. Altho, I do not know if you could do this "while" the battery was connected to B-/+ at the same time.

What's the link the BMS you are using and more suggestions can be given.
 

bmcmullin

New member
Joined
Jun 12, 2019
Messages
18
Korishan said:
The inverter wouldn't have caused the BMS to trigger. Most likely the BMS sensed and overcurrent event and shutdown. I would assume this was because the inverter was initially pulling more current than what the BMS was "actually" built as (not what the label states). It may have been really close to the max, and then at the 15 second mark (or there abouts), the inverter pulled a little more and went over the threshold of the BMS.

Putting a charger on the BMS is to reset it is normal. You could also have just connected the battery terminals to the P-/+ as well. Altho, I do not know if you could do this "while" the battery was connected to B-/+ at the same time.

What's the link the BMS you are using and more suggestions can be given.

Korishan. Here's the link to BMS:https://www.ebay.com/itm/30A-4S-14-...alance-BMS-PCB-Protection-Board-/322790729476

Thank you for your assistance. Onthe cause of the shutdown, given that it was only a 75W load (light bulb) on the inverter that would be less than an amp of currentat 115V. Not sure what that translates to on the 12V side of the inverter drawing from the pack. I reallywonder if there could be a problem with the inverter. I should try and measure the amperage on the input side of the inverter to see exactly what it's drawing. Unfortunately, my multimeter only goes up to 10amps.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,399
75W / 115V = 0.65A on the AC side. However, we're needing to know the draw on the DC side. So, we still have 75W for the bulb, plus the load of the inverter itself. I will just assume that load is around 10W. This gives us a full load consumption of 85W.

85W / 12V = 7.08A. This is, of course, the static load. This is not including the draw required to charge the inverter to begin with. The capacitors were probably fully discharged. This could cause the inverter to draw 20W or more on initial power up. This is a guess as we don't know what inverter you are using.

For the BMS you have, it is "rated" at 30A. Because it is a cheap chinesium version, we will assume that 15A is more likely the constant load draw and 30A is surge.

So, now we have a possible of 95W - 100W or so. 100W / 12V = 8.3A. This is still lower than the supposed current limit of the BMS. Unless, the inverter actually pulls far more power than our current assumption, which is quite possible.

What inverter are you using?

If your DMM goes to 10A, you should be able to still test. You could use 2 DMM's in parallel possibly (I'm actually not sure if this would work as I don't know if the DMM's would interfere with each other)
 

bmcmullin

New member
Joined
Jun 12, 2019
Messages
18
Korishan said:
75W / 115V = 0.65A on the AC side. However, we're needing to know the draw on the DC side. So, we still have 75W for the bulb, plus the load of the inverter itself. I will just assume that load is around 10W. This gives us a full load consumption of 85W.

85W / 12V = 7.08A. This is, of course, the static load. This is not including the draw required to charge the inverter to begin with. The capacitors were probably fully discharged. This could cause the inverter to draw 20W or more on initial power up. This is a guess as we don't know what inverter you are using.

For the BMS you have, it is "rated" at 30A. Because it is a cheap chinesium version, we will assume that 15A is more likely the constant load draw and 30A is surge.

So, now we have a possible of 95W - 100W or so. 100W / 12V = 8.3A. This is still lower than the supposed current limit of the BMS. Unless, the inverter actually pulls far more power than our current assumption, which is quite possible.

What inverter are you using?

If your DMM goes to 10A, you should be able to still test. You could use 2 DMM's in parallel possibly (I'm actually not sure if this would work as I don't know if the DMM's would interfere with each other)

I definitely need to find a way to accurately measure the current draw. I have a Hitec X4 AC multi-charger which does discharge testing.https://hitecrcd.com/products/charg.../x4-ac-plus-4-port-acdc-multi-charger/product I'm going to see if I can control the discharge current draw. I'm sure there are countless ways to put and measure a load on a battery. Thanks again!
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,399
bmcmullin said:
I'm sure there are countless ways to put and measure a load on a battery. Thanks again!

Yes, indeed there are. If you don't have any devices that can handle the higher current, you can build one fairly easily.

Take a piece of copper wire about 2 inches long & measure its resistance. The small wire can be anything larger than 18 gauge (due to higher amp draw). You then measure the voltage drop across the wire to determine how much current is flowing. This is what is also called a current shunt.

Then you use the I = V/R formula. I = Current, V = Voltage, R = Resistance (in Ohms)
So if the wire is 10-Ohm resistance, and you have a voltage drop of .1V, then the current transferred would be 0.01A = .1V / 10
 

bmcmullin

New member
Joined
Jun 12, 2019
Messages
18
Korishan said:
bmcmullin said:
I'm sure there are countless ways to put and measure a load on a battery. Thanks again!

Yes, indeed there are. If you don't have any devices that can handle the higher current, you can build one fairly easily.

Take a piece of copper wire about 2 inches long & measure its resistance. The small wire can be anything larger than 18 gauge (due to higher amp draw). You then measure the voltage drop across the wire to determine how much current is flowing. This is what is also called a current shunt.

Then you use the I = V/R formula. I = Current, V = Voltage, R = Resistance (in Ohms)
So if the wire is 10-Ohm resistance, and you have a voltage drop of .1V, then the current transferred would be 0.01A = .1V / 10

Korishan. I got good news and a new issue. First the good news. I could not replicate the problem where the BMS cutoff was getting triggered by the inverter. I didn't yet complete the test you suggested because I don't fully understand what I'm doing...yet.

The new issue I've discovered is that three of the series in the 8P pack have significant voltage variance between the cells. I created a spreadsheet and took a snapshot of it here.https://www.awesomescreenshot.com/image/4087015/210564f2fc53ea89cf706f183c84d3ad
As you'll see, series 3,4,5 have cells with high voltages. Actually, 5 of 12 cells are above 4.20V.One cell is 4.48V! The pack was fully charged this morning and off the charger since a short while ago.

I don't know what to make of or do about this. How concerning is the voltage variance? I can rip the pack apart and individually test and charge the suspect cells. Any guidance appreciated.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,399
Firstly, you need to bring those high voltage cells down asap! The longer the cells are at that high voltage, the shorter their life span. Not to mention the more risky of them going off becomes.

Secondly, if the packs are not of equal voltage "and" the bms is connected correctly, then the bms is toast. Maybe that's why you couldn't replicate the issue you initially had. It's also possible that the original issue was caused by the bms failing (this is a guess at this point).
If the bms isn't balancing, then "something" is wrong. You really need to be scrutinizing the whole setup and double or triple check all connections to the bms to make sure everything is good.
Time to narrow down where the problem is.
 

jonyjoe505

Active member
Joined
Feb 28, 2018
Messages
225
The bms is bad. No bms should ever let a cell get to 4.48. If anything a bms will trip early as soon as a cell reaches 4.20 and leave you with an undercharge battery pack. It should protect against an unbalance battery, thats its most important job. I would get rid of the bms.

Before I trust a bms I always monitor its first full charge to make sure it cutsout at 4.20 volts. I have encounter some bms that will not cutout at 4.20 and if I hadn't been monitoring would have overcharge the pack. Its rare to get a bad bms but it does happen.

They sell battery checkers that will sound an alarm if the pack goes above 4.22 volts, I use them. They cost about 10 dollars. Its good when your testing a bms. But even with an alarm I like toeyeball the voltages as the battery voltage gets close to the upper limit. Lots of things can fail, bms, alarm.


Tenergy Intelligent Cell Meter Alarm Digital Battery Checker

image_opgsic.jpg
 

bmcmullin

New member
Joined
Jun 12, 2019
Messages
18
jonyjoe505 said:
The bms is bad. No bms should ever let a cell get to 4.48. If anything a bms will trip early as soon as a cell reaches 4.20 and leave you with an undercharge battery pack. It should protect against an unbalance battery, thats its most important job. I would get rid of the bms.

Before I trust a bms I always monitor its first full charge to make sure it cutsout at 4.20 volts. I have encounter some bms that will not cutout at 4.20 and if I hadn't been monitoring would have overcharge the pack. Its rare to get a bad bms but it does happen.

They sell battery checkers that will sound an alarm if the pack goes above 4.22 volts, I use them. They cost about 10 dollars. Its good when your testing a bms. But even with an alarm I like toeyeball the voltages as the battery voltage gets close to the upper limit. Lots of things can fail, bms, alarm.


Tenergy Intelligent Cell Meter Alarm Digital Battery Checker

image_opgsic.jpg


Hmm. Now I'm really confused. As this is a 4S8P pack with a single BMS, I don't see how the BMS can balance individual cells in any one series. Everything I read says it's not possible to balance individual cells within a series of a parallel pack. I took great care to connect the BMS correctly. I have it attached to one end of the pack and all the connections B+/-, B1,B2,B3 wireles connected to the series at the same end of the pack. All voltages from the BMS pads are reading within spec. I just checked again and B and P read the same at 16.63V and B- to B1 is 4.14V, B1 to B2 4.15, B2 to B3 4.16 and B3 to B+ 4.15. I do have a balance charger (Hitec X4) and I did for the hell of it (a few days ago)connect the balance connector and charge wires to my pack through the BMS. I should point out that all the cells were capacity, voltage tested and charged individually before I assembled the pack.As you might expect, the X4 topped up the pack and "balanced" but, again, the balance leads of the BMS are connected physically tothe first series in the pack so the voltages are going to be the average of the 8 cells in the string or just the four cells?

I'm going to be embarrased if something is connected wrong and/or I have a gross misunderstanding of this domain but I'll happily admit failure if it leadsme to enlightenment and a proper understanding of this science. I am very grateful for your assistance. I wish I could return the favor.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,564
Most cheap bms trigger at around 4.25-4.3v and within a second
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,399
bmcmullin said:
I'm going to be embarrased if something is connected wrong and/or I have a gross misunderstanding of this domain but I'll happily admit failure if it leadsme to enlightenment and a proper understanding of this science. I am very grateful for your assistance. I wish I could return the favor.

I think we need a drawing/schematic of how 'you' have everything hooked up. This might help us clarify things up.

So for added info:
Voltages between 'parallel' cells can not have different voltages, not possible. Voltages between 'series' connected cells/packs can have different voltages, and this is what I was assuming when you stated different voltage readings earlier.

It is still possible the bms can be wired up exactly as it's supposed to be, and still not function as intended. Unfortunately, they can't be all be tested before they leave the factory, especially chinesium brands.
 

Alexb

New member
Joined
Mar 21, 2021
Messages
1
I appreciate the post is almost two years old, but wanted to reply for anyone else like me that stumbles across it... that pack is wired up wrong. You have 8 cells in parallel, each cell with 4 batteries wired in series (8P4S). That’s not what you want. 4S8P means you want 4 cells in series, each cell with 8 batteries wired in parallel.


start by making a single cell, with 8 batteries wired together in parallel (all lined up in the same direction, with one single strip of nickel running across all of their tops (+ve), and another strip across all of their bottoms ( -ve ).

that’s 1S8P

make 3 more of those cells, with each cell completely seperate from the others, and with each cell containing 8 batteries wired in parallel.

each cell should be 4.2v (assuming lithium ion), despite having 8 batteries, because the batteries are connected in parallel.

you now have four packs, each 1S8P, that aren’t connected to each other at all.

then connect those 4 cells in series to make a single pack (16.8v). That’s 4S8P.
Viewed from side on it should look like an M (three series connections between 4 cells).


The 3 balancing wires would then connect to the tabs between each of the 4 cells (the tabs that join them in series - the top two points on the M, and the point at the bottom / middle of the M)
 
Last edited:
Top