Signalab only hits 4.16v on homemade 18650 pack

[flyingbeekeeper]

I've got 3 sub par cell blocks on a BMS, I'm confused why it can go up to 4.16 but not 4.20v. Any ideas?

First, I know that 3 cell banks are weak, and replacement of said 3 will likely fix the issue. I'm not wanting to know why my pack is not hitting 4.2v, I want to know the controller logic that won't get it up to 4.2
Second, the homemade pack is a new build with old laptop cells (18650) that passed (13S2P) and the BMS hasn't proved to me it can hit 4.20.
Third, I can take the pack to 50% DOD and plug it in and it will fairly quickly rise to 4.16v. BUt I can leave it charging for 3 days and still be at 4.16. So I'm confused.
Fourth, I can charge the piss out of a cell group and take it majorly out of ballance, like 4.26 on one group and the BMS bleeds it down until all are 4.16 again.
Fifth, I'm an EE and have a clue. But explain to me what you think the BMS may be doing, as if I were five.


The question is not "what is wrong with my pack?", I'm pretty sure I know the cause. The real question is what is the BMS doing and why can't it get to 4.20?
I've bypassed the BMS charging cable and put on balancing leads and balanced the pack to 4.20 using an IMax B6. It gets to 4.2v and holds with the BMS connected. The weak 3 banks decay to 4.19 but stay alright there. So the cells can get there, but the BMS can't. Any ideas what the BMS is doing?

Specs on the Signalab
13S
Model: BMS13LI
Protection Circuit Module Specifications For 46.8V Li-ion Battery Pack, Different with normal PCB, it provides equilibrium function after battery pack is fully charged, e.g. PCM will detect each cell's voltage and trim higher voltage until other cell's voltage reach same,therefore allow Li-ion cell has longer service life. Also the PCM will provide the following protections:
* Balance each cell at Max. 4.2/cell ( required voltage tolerance within 0.2V for each cell)
* Overcharge protection voltage for single cell: 4.35V +/-0.025V
* Overdischarge protection voltage for single cell: 2.40V +/-0.08V
* Continus working current: 30A
* Overcurrent detection protection: 50A ~ 60A
* Supply current: Max 30uA
* Short circuit protection
* Protection circuitry resistance: <=50mohms
Charge current : <=5A

 

[Nobuo]

what kind of charger are you using and voltage?

 

[flyingbeekeeper]

The Sunthing 60v 5a charger and I also tried the Satiator. The BMS begins to cycle the charger on/off at 4.14 and by 4.16 it is mostly off then blinked on once every ten minutes or so.

 

[Nobuo]

Basically the BMS can't avoid the charger to being charging a paralleled group until 4.20V or more.

Depending on how exactly are the parameters of the BMS it should starting to drain paralleled groups from somewhere between 4.16V to 4.20V while, if the max charger voltage has not yet risen, that draining paralleled gropus at ~0.05A discharging will be indeed still rising their voltages (because they are still charging at charger amps)

Then if any paralleled group hits somewhere (depending on exactly how is configured) between 4.20V to 4.22V BMS only can switch off the charge mosfet and stop the charge.

So if when your cells reaches 4.16V the charger still in CC phase (max current) so the max voltage has not yet reached, then only the cell chemical properties could avoid rising over 4.16V, making probably considerable heat.

Another reason would be that your cells are balanced, and when one paralleled group and followers reaches BMS draining voltage, your BMS is in CV stage with an enough low current to be equal than draining current. So there must be anywhere a wrong voltage read value

It could also be possible that the BMS has the voltage cut per paralleled group along with the draining voltage hit configured too close each other at ~4.16V, but this is less Probably

 

[dmwahl]

Could just be that the component tolerances on the BMS that are causing the charger to shut off early. I'd have to see a picture of it to be sure, but I'd guess that each channel on your signalab BMS is based around a TL431 configured as a shunt regulator. If the reference voltage divider resistors are not precise enough or the TL431 itself is at one end of the normal distribution, you could get one channel shutting off the charger at a lower voltage than another. The fact that they state the supply current as "Max 30uA" means that they clearly don't understand the parts on the board too well and may not have taken the care required to meet all their published specs (see the graph titled "Cathode Current vs Cathode Voltage" in TI's datasheet for the part, actual current draw is around 200uA).

 

[999zip999]

What's missing is the chargers voltage could you post here ?

 

[flyingbeekeeper]

What's missing is the chargers voltage could you post here ?

The Sunthing charger is a 48v nominal that hits 60.2v. The cycle Satiator can do 60 as well, but the bms turn on/off cycling shuts the satiator down in mid 50s

 

[999zip999]

13 x 4.2volts = 54.6volt

 

[flyingbeekeeper]

Yes....and?

13 x 4.2volts = 54.6volt

 

[999zip999]

The magic has left those old cells.
13s li-ion pack charger is set to 54.6 volt maybe a little higher 55v ?
Maybe it's the way you prepared the cells to make a pact you can't just connect cells together to make a pack, there's a process

 

[flyingbeekeeper]

The magic has left those old cells.
13s li-ion pack charger is set to 54.6 volt maybe a little higher 55v ?
Maybe it's the way you prepared the cells to make a pact you can't just connect cells together to make a pack, there's a process

No, it is set to 60v, there is plenty of headroom for the BMS to balance cells above the pack level.

 

[nukezero]

I have the same SignalLabs BMS as you do and I bought it from eBay as well from this same guy (victorstudio). It is for a 13s cell.

basically, I have two different 48v chargers, all outputs 54.4-54.6V. but my 18650-25r 13s4p pack with this signalabs BMS ALWAYS only lets the charger charge to 54.1V (which is exactly 4.16V per cell bank).

Glad to know I'm not the only guy with this problem but it is definitely a BMS issue and not a battery issue because my cells were brand new from day 1 and it was already behaving like this.

What you are seeing is exactly what I'm seeing so clearly, this IS BMS related. I have not cared to overcome this problem because 54.1V charge is good enough for me.

 

[flyingbeekeeper]

I have the same SignalLabs BMS as you do and I bought it from eBay as well from this same guy (victorstudio). It is for a 13s cell.

basically, I have two different 48v chargers, all outputs 54.4-54.6V. but my 18650-25r 13s4p pack with this signalabs BMS ALWAYS only lets the charger charge to 54.1V (which is exactly 4.16V per cell bank).

Glad to know I'm not the only guy with this problem but it is definitely a BMS issue and not a battery issue because my cells were brand new from day 1 and it was already behaving like this.

What you are seeing is exactly what I'm seeing so clearly, this IS BMS related. I have not cared to overcome this problem because 54.1V charge is good enough for me.

OK, then. Mystery solved!

 

[nukezero]

I have the same SignalLabs BMS as you do and I bought it from eBay as well from this same guy (victorstudio). It is for a 13s cell.

basically, I have two different 48v chargers, all outputs 54.4-54.6V. but my 18650-25r 13s4p pack with this signalabs BMS ALWAYS only lets the charger charge to 54.1V (which is exactly 4.16V per cell bank).

Glad to know I'm not the only guy with this problem but it is definitely a BMS issue and not a battery issue because my cells were brand new from day 1 and it was already behaving like this.

What you are seeing is exactly what I'm seeing so clearly, this IS BMS related. I have not cared to overcome this problem because 54.1V charge is good enough for me.

OK, then. Mystery solved!

let me know how you intend to solve this problem. I'm sure that the seller made it this way to protect the cells from possibly overcharging.

Btw, I had a suspicion that the BMS may or may not be balancing so what I did was took a USB port (5V) and attached it to one of the cell group. Sure enough, an LED lit up (indicating overvoltage) and so it works.


this is the seller that i bought it from: devicepro, a.k.a. victorstudio

he's some random chinese guy, with not so good english grammar.

http://www.ebay.com/itm/BMS-for-48V-13-cells-Li-Ion-Battery-Pack-50A-10A-limit-With-Balance-Function-/151607642588?hash=item234c8509dc

 

[dmwahl]

Any ideas what the BMS is doing?

They probably ignored the current into the TL431 reference pin when calculating the voltage divider values. As I recall, it's about 2uA, which could easily change the point that it starts to turn on. If you want to raise the terminal cell voltage, you could tack some high value parallel resistors onto the bottom (R2) half of the reference divider to lower it's effective value. Or you could remove them and replace with slightly smaller parts.

Personally, I would just use it as is. The amount of energy stored between 4.16V and 4.20V is pretty small relative to the total pack capacity. If you do want to fix it, and as a fellow engineer I understand completely , figure out what the resistor values are that set the terminal voltage and change them. They're probably around 47k and 75k for R1 and R2 respectively, or something close to them with the same R1/R2 ratio. Whether or not your cells are old and tired doesn't seem to be the issue here.

 

[flyingbeekeeper]

Any ideas what the BMS is doing?

They probably ignored the current into the TL431 reference pin when calculating the voltage divider values. As I recall, it's about 2uA, which could easily change the point that it starts to turn on. If you want to raise the terminal cell voltage, you could tack some high value parallel resistors onto the bottom (R2) half of the reference divider to lower it's effective value. Or you could remove them and replace with slightly smaller parts.

Personally, I would just use it as is. The amount of energy stored between 4.16V and 4.20V is pretty small relative to the total pack capacity. If you do want to fix it, and as a fellow engineer I understand completely , figure out what the resistor values are that set the terminal voltage and change them. They're probably around 47k and 75k for R1 and R2 respectively, or something close to them with the same R1/R2 ratio. Whether or not your cells are old and tired doesn't seem to be the issue here.

If it is the BMS logic that simply stops at 4.16v, then I won't fix it. I am a little dissapointed that the lights won't go lit when it is done charging, but I can live with it. I actually would prefer 4.16 to 4.20 anyway to increase cycle life.

 

[dmwahl]

You could always check by disconnecting the BMS and applying 4.5V or so to each channel. If you have a bench power supply and can current limit to 50mA or so that would tell you exactly what the BMS is set to.

 

[nukezero]

Any ideas what the BMS is doing?

They probably ignored the current into the TL431 reference pin when calculating the voltage divider values. As I recall, it's about 2uA, which could easily change the point that it starts to turn on. If you want to raise the terminal cell voltage, you could tack some high value parallel resistors onto the bottom (R2) half of the reference divider to lower it's effective value. Or you could remove them and replace with slightly smaller parts.

Personally, I would just use it as is. The amount of energy stored between 4.16V and 4.20V is pretty small relative to the total pack capacity. If you do want to fix it, and as a fellow engineer I understand completely , figure out what the resistor values are that set the terminal voltage and change them. They're probably around 47k and 75k for R1 and R2 respectively, or something close to them with the same R1/R2 ratio. Whether or not your cells are old and tired doesn't seem to be the issue here.

If it is the BMS logic that simply stops at 4.16v, then I won't fix it. I am a little dissapointed that the lights won't go lit when it is done charging, but I can live with it. I actually would prefer 4.16 to 4.20 anyway to increase cycle life.

I have NEVER seen those LEDs light up, when I charge it, EVER. that's what concerned me, but I guess the cells never really went out of balance. I have never balance charged my 18650-25R samsungs and depended on the BMS to do its job. When I bought all 52 cells, they were perfect at 3.63Volts. Today, after 200+ cycles, using this "interesting" Signalabs BMS, they still charge to 54.1V. Although the only thing I can say is that capacity has gone down ~20%, but that is due to wear and tear and I discharge to empty quite a bit (42-41V). Overall, I would buy it again and re-use it because it has not caused me any problems.

On the other hand, my gf has a 36v 10s4p (18650-26F) with an unknown BMS. this pack is problematic now after 2 years holding 39.8V. I will tear this sucker apart tonight to figure out what's wrong. Most likely a cell imbalance , or a crappy BMS not doing it's job. I would like to stick a 10s4p signalabs in there though but I'm almost positive it is not a signalabs one in there.

 

[nukezero]

I spoke too soon having too much faith in this "Signalabs" BMS. Check out my other threaad about my 36V 18650 10s4p which also runs a signalabs 36v BMS. I bought the entire battery pack (18650-26F) and BMS assembled from VictorStudio (a.k.a. DevicePro).

After two years, half the cell bank finally went a bit out of balance, which is the reason why it only charged up to 39.8V (instead of the proper 42.0V)

https://endless-sphere.com/forums/viewtopic.php?f=14&t=73288

 

[999zip999]

I put a set of sense wires for cellog and balancing, yes with a bms. Time saver.

 

[Ykick]

I put a set of sense wires for cellog and balancing, yes with a bms. Time saver.

Yep, parallel JST dongle is invaluable for illustrating what the BMS is doing.

I grabbed this discharge curve using CellLog 8S

As you can see at the end BMS disconnects the load. Prevents cell damages. I've gotta charge curve someplace, can't find right now but petty much the same illustration in the other direction.

BestechPower BMS do seem to operate closer to published OVD specs (4.22V/cell) thereabouts...

 

[flyingbeekeeper]

New clue today after a 130 mile weekend ride. Charging FETS melted my wrapping.

The chips are CS48N88, specs at:
http://www.alldatasheet.com/datasheet-pdf/pdf/665699/THINKISEMI/CS48N88.html

Essentially, 70v 92amp N-channel mosfet. So why are they getting hotter than 160F with just charging with 4amps. Any ideas?

I saw a thread here: https://endless-sphere.com/forums/viewtopic.php?f=14&t=8841&start=75
that kinda addressed hot Signalab chips, with some replacement ideas.

 

[flyingbeekeeper]

Measured the Gate to Source (two outer legs) on the mosfets. All were 16.02v, except that bottom one in the picture. And it had the most severe heat damage, melted solder, and its' Gate to Source voltage was 0.12 V. Hmmm... doesn't seem right. It is either not fully on, or it is totally blown. Digikey doesn't have a replacement, so I'll have to spot an equivalent.

 

[nukezero]

Hey guys, so I have some update for you guys regarding "victor-studio" or "devicespro" Signalab 48v BMS. So I bought two BMSs' from this guy.

The first BMS is wired to my 48v 10Ah, Samsung 25R in 13S4P. I never saw the red led lights, light up. They always charge to 4.16V and stop. Total voltage is 54.1V hot off charger. I guess the pack never goes out of balance after 1.5 years and 200+ cycles

Today, I used the spare BMS and built another 25R pack but in 13S2P only. Here's what happened! The first 4 pairs of 2P, all hit RED led light on this spare BMS! I measured the voltage and they were all at 4.19 and 4.20 volt. The little optoresistors thing got a bit warm/hot and I can tell they were burning the energy off. Within about 10 minutes, the LEDs will shut off and I think that means it has bled the cells to about 4.19-4.18V. However, when I leave the charger on, it get's shut off I guess until the first few cells bleed off to let the rest of the other cells catch up. Now, when I built this pack, sure there were at 0.05-0.08V out of balance. Not too bad. But ultimately, I dont know why this second Signalab BMS seems like is functioning properly. It allows itself to be charged to 4.20V ! and then the red LEDs turn on.

So something must be different between these two BMSs that I bought from devicespro.