reg
Winter: New York City Summer: Atlantic Canada
- Joined
- Mar 24, 2021
- Messages
- 336
Thanks for the video. It's always good to hear from someone who actually knows what he's talking about.
Coulomb counter not strictly compatible with LiFePo4?
- Thread starter Freep
- Start date
Thanks for the video. It's always good to hear from someone who actually knows what he's talking about.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,696
I agree, the video is good information. When I was in a power failure, and my battery bank was running down, I do remember seeing it drop faster and faster as the voltage fell. If I was just looking at SoC% I might not have realized that I really had 2 hours less time than doing the quick amp hours / amps at the beginning. But I knew this was going to happen, and I kept an eye on it. I watched my amp hours remaining fall, and turned off a few more things to stretch it through the night.
Bob B
Emperor Of Solar
- Joined
- Sep 21, 2019
- Messages
- 8,576
I guess I agree that there is more to accurate tracking than just tracking amps in and amps out .... but, the idea of intermittently checking voltage and then comparing that to an open circuit voltage chart is just as far off the mark as just tracking amps.
As I said earlier .... if you want accuracy, just track watts. Pretty simple.
My Chargery BMS tracks amps and watts ... so take your choice. Pretty sure Victron does it the easy way and tracks watts.
As I said earlier .... if you want accuracy, just track watts. Pretty simple.
My Chargery BMS tracks amps and watts ... so take your choice. Pretty sure Victron does it the easy way and tracks watts.
Diysolar123
Solar Addict
- Joined
- Feb 28, 2021
- Messages
- 675
well, only an old retired engineer here, so I will take a shot at making it "simple" 
(said absolutely every engineer as they start their presentation)
An amp is just a coulomb/sec
or
1coulomb = 1amp * 1sec.
the rate of flow of the electron does not give a hoot about the voltage so it literaly is the number of electrons that have flowed.
energy is the relationship between voltage and current.
if you know how many amp-seconds worth of electrons you put into something, well, when you have started to pull that many out you are running on empty. Sounds like people are trying to compare "current" and "energy"...they are related but different.
the "energy" is a function of voltage * coulombs used; and therefore the potential energy in the cell is dropping as the voltage drops even with a constant coulomb draw... the flat discharge nature of the lifepo4 means that the window of constant energy draw (consistant voltage and current) is also mostly flat until the end.
Once you have hit the point of rapid voltage drop you ARE getting less "energy" for the same coulomb draw; but the end is near anyway due to the sharp dropoff. Your coulomb counter is still valid, if you want to maintain a specific power output, you will draw more coulombs as the voltage drops...shunts measure coulombs and you will not get out more than you put in(if you do, quick, get a patent!!!).
you can tell how long you have left before dead just by coulomb counting...
If I have 200coulombs left and I am drawing them thru the shunt at some rate, then its easy to solve for 0.
You could probably take the standard discharge curve of the end of the lifepo4 and create a simple two point equation from it once you hit the drop-off to actually estimate the amount of energy left. The voltage will keep changing and coulombs will increase to sustain the energy demand.
Or, you can just count coulombs, when you hit zero, game over hehe
You are using the energy but its really the flow of electrons you are controlling.
That said, its way simpler to just accept that once you drop under 3volts, you are done. If you have been counting coulombs all along, you know how many amp-seconds you have left if you want to predict exactly how many watt-seconds remain.
(said absolutely every engineer as they start their presentation)An amp is just a coulomb/sec
or
1coulomb = 1amp * 1sec.
the rate of flow of the electron does not give a hoot about the voltage so it literaly is the number of electrons that have flowed.
energy is the relationship between voltage and current.
if you know how many amp-seconds worth of electrons you put into something, well, when you have started to pull that many out you are running on empty. Sounds like people are trying to compare "current" and "energy"...they are related but different.
the "energy" is a function of voltage * coulombs used; and therefore the potential energy in the cell is dropping as the voltage drops even with a constant coulomb draw... the flat discharge nature of the lifepo4 means that the window of constant energy draw (consistant voltage and current) is also mostly flat until the end.
Once you have hit the point of rapid voltage drop you ARE getting less "energy" for the same coulomb draw; but the end is near anyway due to the sharp dropoff. Your coulomb counter is still valid, if you want to maintain a specific power output, you will draw more coulombs as the voltage drops...shunts measure coulombs and you will not get out more than you put in(if you do, quick, get a patent!!!).
you can tell how long you have left before dead just by coulomb counting...
If I have 200coulombs left and I am drawing them thru the shunt at some rate, then its easy to solve for 0.
You could probably take the standard discharge curve of the end of the lifepo4 and create a simple two point equation from it once you hit the drop-off to actually estimate the amount of energy left. The voltage will keep changing and coulombs will increase to sustain the energy demand.
Or, you can just count coulombs, when you hit zero, game over hehe
You are using the energy but its really the flow of electrons you are controlling.
That said, its way simpler to just accept that once you drop under 3volts, you are done. If you have been counting coulombs all along, you know how many amp-seconds you have left if you want to predict exactly how many watt-seconds remain.
Fun fact - One coulomb is 6,241,509,074,460,762,607.776 electrons, give or take.
You can't actually count coulombs, but you can measure voltage over a known calibrated resistance (shunt) to calculate amps and measure time to calculate Amp-hours/coulombs. Yes, a bit nit-picky.
Also, the shunt is placed on the negative terminal because that's where the electrons flow from/to. Electron flow is opposite from current flow.
You can't actually count coulombs, but you can measure voltage over a known calibrated resistance (shunt) to calculate amps and measure time to calculate Amp-hours/coulombs. Yes, a bit nit-picky.
Also, the shunt is placed on the negative terminal because that's where the electrons flow from/to. Electron flow is opposite from current flow.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,696
Technically, the shunt can be anywhere in the circuit, as the current flow is the same anywhere in the entire loop. But most systems place it on the negative side to keep the voltage across the shunt close to the negative rail so the electronics do not need to be isolated and floating to measure the millivolts across the shunt while at 58 volts on the battery positive. There are single chip ic's that can handle up to 100 volts of input offset an still measure a few millivolts, but they are still not as accurate as directly measuring within the power supply range and they can get expensive as well.
My BMS appears to measure the current about every 0.5 seconds. It adds it up every sample and calculated the amp hours used. I am fine with this and understand the limitation. All it would need to do it also sample the voltage at the same time and multiply to get the watt hours used. That part is easy, but to get a very accurate watt hours left, you would need to fully cycle the battery down to cut off a few times to log a trend so it knows the accurate true watt hours total of the battery to begin with. If the battery measured at 15,000 watt hours at 20 amp load, and 14,800 watt hours at a 100 amp load, it would need to calculate a correction factor as well, based on the rate of current consumption. Modern lithium batteries are much better than lead acid, but there can still be some of this effect. So even counting the watt hours used, and subtracting from the total watt hours is not perfect, but it will be closer. This will give a better warning when you are at 50% energy left vs 50% amp hours left, but that second half will still likely not last as long as the first half. As the voltage falls, the current will increase to keep the power up. So you are pulling more current during the second half. Using the watt hour calculation does correct for a lot of it, but not all of it. This will matter if I am driving an EV uphill into AZ, but it is not a problem with my solar storage. The only time it matters to me at all is when I am in a power failure and I use my amp hours remaining to decide if I need to turn off more things to make it through the night. As a backup, I set my inverter shut down well above the safe bottom limit of my battery, so when I need power again, I can log in and lower the cutoff and get the system running when I have sun to charge again. Or like I did last time, decide to fire up my generator to get a little more back into the battery before the system shut down. Knowing the approximate Amp Hours left is good enough in my case. But if I was selling a system like this to a customer, it might be worth having a better Watt Hour fuel gauge to give a better warning of how much time they have left during a power outage. It can be difficult to explain to a customer that 50% amp hours remaining is only going to be 40% of the energy left, and it will keep falling faster and faster as the voltage drops..
My BMS appears to measure the current about every 0.5 seconds. It adds it up every sample and calculated the amp hours used. I am fine with this and understand the limitation. All it would need to do it also sample the voltage at the same time and multiply to get the watt hours used. That part is easy, but to get a very accurate watt hours left, you would need to fully cycle the battery down to cut off a few times to log a trend so it knows the accurate true watt hours total of the battery to begin with. If the battery measured at 15,000 watt hours at 20 amp load, and 14,800 watt hours at a 100 amp load, it would need to calculate a correction factor as well, based on the rate of current consumption. Modern lithium batteries are much better than lead acid, but there can still be some of this effect. So even counting the watt hours used, and subtracting from the total watt hours is not perfect, but it will be closer. This will give a better warning when you are at 50% energy left vs 50% amp hours left, but that second half will still likely not last as long as the first half. As the voltage falls, the current will increase to keep the power up. So you are pulling more current during the second half. Using the watt hour calculation does correct for a lot of it, but not all of it. This will matter if I am driving an EV uphill into AZ, but it is not a problem with my solar storage. The only time it matters to me at all is when I am in a power failure and I use my amp hours remaining to decide if I need to turn off more things to make it through the night. As a backup, I set my inverter shut down well above the safe bottom limit of my battery, so when I need power again, I can log in and lower the cutoff and get the system running when I have sun to charge again. Or like I did last time, decide to fire up my generator to get a little more back into the battery before the system shut down. Knowing the approximate Amp Hours left is good enough in my case. But if I was selling a system like this to a customer, it might be worth having a better Watt Hour fuel gauge to give a better warning of how much time they have left during a power outage. It can be difficult to explain to a customer that 50% amp hours remaining is only going to be 40% of the energy left, and it will keep falling faster and faster as the voltage drops..
Well, if anyone knows whether the Victron Smart Shunt uses a method for determining the SOC of a battery that is different than the AiLi or the Xantrex LinkLite SOC meter I would like to know what it is for future reference. From what I'm gathering here even if the Smart Shunt is using the method described in the video you're not really going to be able to notice a difference until you get down lower than 20% of SOC anyway. Especially in a 272ah 12V pack used in a camper.
I shared the link to this discussion with the guy who got me interested enough to ask you guys, hoping he could shed some light and maybe teach us something but the more I asked, the more snarky and argumentative he got.
At least now I know a little more about how SOC is measured. Thanks for that.
I shared the link to this discussion with the guy who got me interested enough to ask you guys, hoping he could shed some light and maybe teach us something but the more I asked, the more snarky and argumentative he got.
At least now I know a little more about how SOC is measured. Thanks for that.
Last edited:
Johncfii
Solar Enthusiast
- Joined
- Nov 24, 2020
- Messages
- 265
The Victron units are not more sophisticated, at least as “sophisticated” is being discussed in this thread. They express state of charge as directly related to amp-hours drawn, compared to what you specify as the full AH capacity of your battery bank. If you tell it that you have a 100 AH battery, when it measures 50 AH discharged out of the battery, it will read 50.0% state-of-charge. It is not more sophisticated than that, except that your set Peukert exponent, and “Charge Efficiency Factor“ will also be reflected in the displayed SOC, over cumulative discharge and charge cycles.
Ampster
Renewable Energy Hobbyist
That is my sense as well. Both my Outback Skybox and my Orion BMS use Coulomb counters and I find them both accurate enough for my purposes.
The guy in the video is a supplier to the EV market and his point was partly to say why his system had a better algorithm. That may be an important issue for EV drivers who might find themselves some distance from a charging station. This thing he did not mention, and something my wife figured out was that just driving slower can get you home even if the EV told you you did not have enough range to get home.
Most stationary storage users would rarely go to 20% before charging their batteries or turning off some loads.
