What's the point of having a coulometer when you have a BMS

[Abyss]

Quick skim and I see no one has mentioned that a hall sensor isn't as accurate as a shunt at lower amperages and voltages, thought that was worth a mention since the OP was looking at this option.

 

[Supervstech]

Probly this. Yes, if I cycle it then bms keeps up.
No, I’m not dead wrong. ?
And I have a shunt, so don’t ask me if I know what one is.

Pull the cover off thw bms...
Look near the p- terminal...

See those bars on thw bms... thats a shunt.
A crappy one, but it is a shunt.

Hardly any electronics operate without a shunt of some fashion in them.

 

[Supervstech]

Quick skim and I see no one has mentioned that a hall sensor isn't as accurate as a shunt at lower amperages and voltages, thought that was worth a mention since the OP was looking at this option.

Hall sensors can be accurate, but they alsoget thrown off by earth and any other magnetic fields, so shielding and cable length, and orientation etc all add up to inaccuracies with them.
A shunt is a specific value easy to calculate off of, and not thrown off by random environmental effects.

 

[Browneye]

To wrap it up, you and I fundamentally agree. A programmable coulomb-counting-based battery monitor is a very useful addition to a system even with a BMS, and I would almost never personally choose to not have that redundancy - even for more $. I have a BMV-702 in series with my Batrium (which actually has a separate shunt), and it tickles me to no end when I see those two devices reporting within 0.1A of each other at all levels. It also reminds me that neither system is completely accurate for my use case because I frequently see their SoC values drift:

Thanks for clarifying. I don't have the same technical ability as you do to explain why and how SOC is calculated with these devices. Bottomline is a meter is advantageous over having just a BMS to measure battery condition, which was the subject of the thread, and confirmed by pretty much everyone here, and why I got one in the first place. That was the question by the OP.

And yes, both my meter (which won't work without a shunt, hence my calling it a shunt) and BMS give whonky values when the charge level of the battery is allowed to drift for weeks or months at 13.25 volts. When SCC is reprogrammed to fully charge it then both arrive at 100%, and both accurately count percent of charge, going up and down in use. At idle seems to be a 'worst case' as you say.

Thank you for apologizing, I've edited my post. IMO a technical explanation has more value than telling people they're wrong. The technical nature of these devices, varied terminology, and the confusion these systems create, brings a lot of users into the forum. Everyone comes here with a different level of understanding and technical ability.

 

[sunshine_eggo]

Thanks for clarifying. I don't have the same technical ability as you do to explain why and how SOC is calculated with these devices. Bottomline is a meter is advantageous over having just a BMS to measure battery condition, which was the subject of the thread, and confirmed by pretty much everyone here, and why I got one in the first place. That was the question by the OP.

And yes, both my meter (which won't work without a shunt, hence my calling it a shunt) and BMS give whonky values when the charge level of the battery is allowed to drift for weeks or months at 13.25 volts. When SCC is reprogrammed to fully charge it then both arrive at 100%, and both accurately count percent of charge, going up and down in use. At idle seems to be a 'worst case' as you say.

Had you included that bit with your original, I would have totally understood why you came to your conclusion.

Thank you for apologizing, I've edited my post. IMO a technical explanation has more value than telling people they're wrong. The technical nature of these devices, varied terminology, and the confusion these systems create, brings a lot of users into the forum. Everyone comes here with a different level of understanding and technical ability.

 

[sunshine_eggo]

Thanks for clarifying. I don't have the same technical ability as you do to explain why and how SOC is calculated with these devices. Bottomline is a meter is advantageous over having just a BMS to measure battery condition, which was the subject of the thread, and confirmed by pretty much everyone here, and why I got one in the first place. That was the question by the OP.

And yes, both my meter (which won't work without a shunt, hence my calling it a shunt) and BMS give whonky values when the charge level of the battery is allowed to drift for weeks or months at 13.25 volts. When SCC is reprogrammed to fully charge it then both arrive at 100%, and both accurately count percent of charge, going up and down in use. At idle seems to be a 'worst case' as you say.

Thank you for apologizing, I've edited my post. IMO a technical explanation has more value than telling people they're wrong. The technical nature of these devices, varied terminology, and the confusion these systems create, brings a lot of users into the forum. Everyone comes here with a different level of understanding and technical ability.

I've been playing around with my Sassy_Navi pack (4S 25Ah Navitas cells pack).

I tested these cells for capacity with several cycles @ 20A, and they were just a hair over 25Ah.

I let it sit fully charged for X months. I'm really not sure how long. Guessing at least 3 months, probably less than 6.

All cells were 3.325±.002V.

SoC was 69.2% (17.3Ah remaining).

That was pretty consistent with the ~70% SoC indicated visually and with the voltage to SoC values in settings.

Conclusion: Reporting SoC based on cell voltage.

I raised the full charge voltage to 3.85V to try and make sure I can capture what it gained before it indicated 100%.

30V/10A set to 14.8V and 10A. Charged at 10A. Indicated 19.3Ah remaining before it triggered full simultaneous with the 3.65V OVP cut-off. All cells were above 3.57V.

so with 70% indicated, it took on 2Ah and synched to 100%.

I held it at 14.8V and allowed it to balance for an hour bouncing off the 3.65V limit. It probably engaged protection every 20 seconds. All cells were above 3.60V.

Conclusion: Even after an extended sit, the cells remained extremely well balanced. Happy.

Every 1-2 days for the last 10 days, I have been accessing the BMS with my phone. It continues to show 100% even though cells have dropped to 3.35V.

Conclusion: Reporting SoC based on the last known synchronization.

I'm going to continue to check it on an increasing delay in frequency. What I expect to find is that at some point, the BMS goes into a standby or "off" state, and the last sync SoC is lost, OR it's some timed arbitrary value, OR it's some voltage threshold. When that condition is met, it reports SoC based on cell voltage.

Conclusion: we're both right depending on usage.

 

[Cal]

Hall sensors can be accurate, but they alsoget thrown off by earth and any other magnetic fields, so shielding and cable length, and orientation etc all add up to inaccuracies with them.
A shunt is a specific value easy to calculate off of, and not thrown off by random environmental effects.

Shunt accuracy is affected by self heating and ambient temperature.

It's been a long time since looking at shunt specifications, but I believe current should be derated by 50%. That means a 100A shunt should conduct 50 A max.

 

[scubadoo]

My series connected shunt based Victron and Junctek battery monitors always agree within 1 or 2% SOC.
Either they are both fairly accurate or much less likely they are inaccurate by a similar amount.

9 year old 4 cell Sinopoly LiFePO4 battery with no internal "BMS" circuitry involved.

Click to enlarge image.

 

[kundip]

Hi,
What's the point of having a coulometer if the BMS already gives you all the informations you need?
I mean, my JBD displays the amps when charging, and when discharging too, it also gives the battery's voltage (I am using a lifepo4). These informations are all you need, or am I missing something ?
Thanks

I have never seen any use for them in my system.
They just add more complexity to the wiring.

 

[629658]

I suspect somewhere inside the BMS there is some device that measures current. Otherwise the BMS would have no way to calculate any of the parameters related to current. Even my 4S Overkill reports consumption.

Yes most BMS have a shunt of some sort to measure the current

 

[629658]

Having a big shunt is easier to monitor your batteries as a pack instead of adding up all the individual BMS data

 

[Ampster]

Having a big shunt is easier to monitor your batteries as a pack instead of adding up all the individual BMS data

That is important with a pack that consists of multiple batteries, each with their own BMSs. A single source to measure Coulombs in and out of the entire pack is important.

 

[Browneye]

I found my soc voltage values in bms we’re off. With corrected values it reports much more accurately. How did that happen??

Still, after sitting for a month at a low float charge, meter says 89% and bms says 55%. If I put a full charge on it, bms will start counting up, but will suddenly jump to full when voltage comes up. Meter says it’s already near full, so it counts up to 100% and stays there till charging terminates.

My Overkill BT app quit working - would not connect. Reinstalled it a couple times, still nothing. Reinstalled the Xiaoziang app and it works perfectly.

Damn electronics. Now if I could just get my Cadillac CUE system to quit dropping out I’d be golden. Lol

 

[curiouscarbon]

for JBD/overkill BMS, the behavior documented on this forum suggests that it will switch between (voltage based) and (coulomb counting) state of charge estimations, and perhaps even mix the two.

 

[Browneye]

Today OK bms counted up from 53% to 89% before it jumped to full as the battery reached 14.2V.

The shunt meter works best counting from full. It seems to get more confused than the bms.

 

[sunshine_eggo]

Today OK bms counted up from 53% to 89% before it jumped to full as the battery reached 14.2V.

The shunt meter works best counting from full. It seems to get more confused than the bms.

Do me a favor and check the BMS every few days to confirm you see the same thing I do. I don't remember the last time I checked, but it's probably been at least 3 days:

3.333-3.335V, which would be something between 60% (3.300) and 80% (3.350), but it's still showing 100%.

 

[brb58]

I have two batteries with a JBD BMS in parallel. Currently, My Victron BMV-71 says 80%...one of the JBD BMS says 81% and the other says 60%. Needless to say, I don't trust the JBD ones.

BMS reports 13.3V, BMV-712 reports 13.27V

 

[sunshine_eggo]

I have two batteries with a JBD BMS in parallel. Currently, My Victron BMV-71 says 80%...one of the JBD BMS says 81% and the other says 60%. Needless to say, I don't rust the JBD ones.

I'm pretty convinced the JBD reports SoC based on coulomb counting and regular syncs. After some period of inactivity, the BMS goes to sleep or resets and will show SoC based on voltage at that point... until it syncs to a full charge.

 

[brb58]

I'm pretty convinced the JBD reports SoC based on coulomb counting and regular syncs. After some period of inactivity, the BMS goes to sleep or resets and will show SoC based on voltage at that point... until it syncs to a full charge.

That's possible. I always wondered about the 80%, 60%, 40% settings and what they were used for. I saw this weird phenomena months ago when I built the batteries. It pretty much doesn't concern me since I rely on the BMV-712.

 

[Browneye]

Do me a favor and check the BMS every few days to confirm you see the same thing I do. I don't remember the last time I checked, but it's probably been at least 3 days:

3.333-3.335V, which would be something between 60% (3.300) and 80% (3.350), but it's still showing 100%.

Oh it settles back right away at rest, but no amps are going out so it still shows 100%. Currently reporting 13.29V - 3.22Vpc and 100%, meter is the same.
But bms didn’t report full during charge till it hit 14V - 3.5Vpc - the value in the settings. Then it jumped from 93% to 100.

So yeah, uses both. Some fuzzy logic there. I still have never done a full discharge to 10% and count amp hours - actual capacity. From full the meter and bms will report the same usage, very close.