Coulomb counter not strictly compatible with LiFePo4?

[Freep]

I was having a discussion with someone on an RV forum and I pointed out that I like the AiLi meter that Will reviewed and it's fine for me because my BMS can give me SOC from my phone and the Victron Smart Shunt is just too pricey for me for what it does.

I got the following response:

Ok, but for clarity it should be noted here that a meter that only counts coulombs is not strictly compatible with LFP, though it may be fine above 50% SOC - in practical terms with LFP, SOC only begins to get important below 50%...

As has been already noted here several times, computing SOC with LFP requires a sophisticated algorithm including a ‘look-up’ ledger which is necessary when below a 50% SOC...This is due to the extremely flat voltage curve characteristics of LFP

Is this right? Have I missed this detail on this forum?

 

[MisterSandals]

Something seems wrong with those quoted sentences. First, counting coulombs is counting coulombs, and what the source/destination is makes no difference.
Secondly, "computing SOC with LFP..." this sounds like its based on voltage instead of coulomb counting... the non-counting meters have only the current voltage to look up (estimate) the SoC. And this of course differs based on type of battery chemistry.

 

[Bob B]

I agree with Mr Sandals .... that statement you quoted would only be correct if trying to determine SOC via voltage .... counting current in and out .... coulombs...... is the remedy for that, not the problem.

Maybe we are missing some context with just a single statement?

 

[Ampster]

Is this right?

Ask your friend if he thinks it would be more accurate if you counted Ahrs. I agree withj @MrNatural22 and @Bob B. it sounds like there is some missing context.

 

[JeepDaddy]

^^^^^ x3

They are referring to measuring SOC by voltage which is not accurate with LFP.

For what it's worth, the Smartshunt is much more accurate than a BMS-based SOC. Pricey - yes. Too pricey - subjective opinion. To each their own.

 

[Freep]

Thanks, everyone. I thought it was wrong too. But I could have missed something so I thought I'd check with the brain trust.


More context. I had recommended the AiLi and this guy was saying why risk your expensive LiFePo4 battery with a Chinese meter(which I dispute) and was touting the superiority of the Victron Smart Shunt(which I don't dispute) and then went on to say what I quoted in the OP and "My point on meters was for awareness to the broader audience that not all SOC meters are LFP compatible - this is why I had to replace my former Xantrex LinkLite SOC meter.."

I just looked up that meter and it looks like it's also using a shunt. So that's really confusing. Also the AiLi looks suspiciously similar to the LinkLite.

 

[Freep]

^^^^^ x3

They are referring to measuring SOC by voltage which is not accurate with LFP.

For what it's worth, the Smartshunt is much more accurate than a BMS-based SOC. Pricey - yes. Too pricey - subjective opinion. To each their own.

Yeah I use the AiLi and the app with my OverkillSolar BMS. The AiLi is when I'm in the camper and just want to glance at the current coming in, going out, or the state of charge and I use the app for checking on the battery from the house when the camper is parked outside or when I want more detailed information about the state of the battery. With unlimited funds, I'd probably go Victron but for what it does, I'd rather buy the AiLi and save the cash for something else.

 

[Freep]

Here is the response:

For LFP, Coulomb counting is just one of several parameters used - coulomb counting alone is insufficient at accurately determining LFP SOC, particularly below 50%...I know nothing about the AiLi, but this issue is why the Victron should not be compared to a simple coulomb counter...

 

[Bob B]

For the first part of that video .... a LiFeP04 pack has a much flatter discharge curve than what he represents .... so, while it is true that from a WH perspective, there is more energy in the first 50% of amps used than the last 50% ...it is less of a factor than his representation makes it appear.
Just a note .... Some BMS .... the Chargery being 1, both SOC and WH are represented on the display.

I disagree with the second part of the video. I don't think a voltage comparison is going to increase accuracy. After all, the reason we don't use voltage to represent SOC is that it is not accurate for LiFeP04 .... so, while the WH representation would require a voltage reference, I don't think it would be of much value for our purposes.

And .... for some reason, I just didn't like that guy. LOL

 

[Zwy]

I was having a discussion with someone on an RV forum and I pointed out that I like the AiLi meter that Will reviewed and it's fine for me because my BMS can give me SOC from my phone and the Victron Smart Shunt is just too pricey for me for what it does.

I got the following response:


Is this right? Have I missed this detail on this forum?

Last I checked, electrons pass thru the shunt and are counted. Battery type has nothing to do with it.

He must be confused with using a voltmeter.

 

[GXMnow]

I looked at the video also, and he is basically right, if you are worried about the Kilo Watt Hours remaining vs just the Amp Hours remaining, a coulomb counter only is going to fall a bit short. But I also agree that what the OP said is backwards. The much flatter discharge curve of LFP makes a coulomb counter actually MORE accurate to the energy remaining. I know my NMC pack is constantly dropping voltage as it discharges, so my first 50% is a lot more power than my last 50%, but if you know you are pulling 10 amps, and you have 50 amp hours remaining, you should make close to 5 hours, and keep an eye on it as you get under 10 amp hours remaining. For the vast majority of us with solar storage, just knowing amp hours is enough data. For an electric car, I can sure see why that might not be good enough. To get the same power as the voltage drops, you need to pull more current, so the amp hours remaining will start falling faster. You will see this on the monitor as higher current so it is telling you it won't last as long. But you have to do some quick math in your head. A true "Fuel Gauge" measures the voltage and after a few cycles, or with a look up table, it can do that math for you, but do we really need that for solar storage?

 

[Freep]

A true "Fuel Gauge" measures the voltage and after a few cycles, or with a look up table, it can do that math for you, but do we really need that for solar storage?

And does the Victron actually do that whereas the others don't? And if it does, what is the accuracy advantage? And to get that accuracy wouldn't you have to enter a lot of data from your battery spec sheet to make it work right?

 

[Zwy]

Here is the response:

For LFP, Coulomb counting is just one of several parameters used - coulomb counting alone is insufficient at accurately determining LFP SOC, particularly below 50%...I know nothing about the AiLi, but this issue is why the Victron should not be compared to a simple coulomb counter...

I watched the video. Basically, he is stating to use the shunt to measure AH (either discharged or remaining). Yes, there will be drift under high voltage or high amperage conditions. This is why a shunt is usually installed on the grounded (P-) side of the circuit. Do we even approach the amperage draws to create the second condition? I say we don't, it would have to be a very large draw unless the shunt is undersized for the task at hand. Remember the resistor in the shunt determines it's capacity. Too small and it will heat up. Too large and it will be inaccurate for small amperage draws.

 

[Zwy]

I looked at the video also, and he is basically right, if you are worried about the Kilo Watt Hours remaining vs just the Amp Hours remaining, a coulomb counter only is going to fall a bit short. But I also agree that what the OP said is backwards. The much flatter discharge curve of LFP makes a coulomb counter actually MORE accurate to the energy remaining. I know my NMC pack is constantly dropping voltage as it discharges, so my first 50% is a lot more power than my last 50%, but if you know you are pulling 10 amps, and you have 50 amp hours remaining, you should make close to 5 hours, and keep an eye on it as you get under 10 amp hours remaining. For the vast majority of us with solar storage, just knowing amp hours is enough data.

That is correct, we do not have the load seen in other applications.

For an electric car, I can sure see why that might not be good enough. To get the same power as the voltage drops, you need to pull more current, so the amp hours remaining will start falling faster.

What happens is the shunt resistor heats up and thus causes the drift under high amperage draws. In our case with small solar systems, we will never approach such a scenario.

You will see this on the monitor as higher current so it is telling you it won't last as long. But you have to do some quick math in your head. A true "Fuel Gauge" measures the voltage and after a few cycles, or with a look up table, it can do that math for you, but do we really need that for solar storage?

I could see it for EV's, not for solar storage unless we start getting high voltages and high amperage.

 

[Bob B]

And does the Victron actually do that whereas the others don't? And if it does, what is the accuracy advantage? And to get that accuracy wouldn't you have to enter a lot of data from your battery spec sheet to make it work right?

I don't have Victron stuff ..... but from what I remember of some of the videos I have seen, things are represented in Watts and Watt Hours .... which WILL be an accurate representation of SOC

 

[reg]

Notwithstanding reservations expressed above, Erik Stafl, who made the video in post #8, would appear to know what he's talking about.

He owns a company that makes battery management systems: Stafl Systems

He's a graduate of MIT, and his Chief Electrical Engineer graduated from Stanford.

Pretty impressive staff: World Class Backgrounds

 

[Bob B]

Notwithstanding reservations expressed above, Erik Stafl, who made the video in post #8, would appear to know what he's talking about.

He owns a company that makes battery management systems: Stafl Systems

He's a graduate of MIT, and his Chief Electrical Engineer graduated from Stanford.

Pretty impressive staff: World Class Backgrounds

He's making it more difficult than it is .... It's really very simple ... P=IE
So, if you want very accurate tracking .... just track watts and watt hours.

 

[reg]

He's making it more difficult than it is .... It's really very simple ... P=IE
So, if you want very accurate tracking .... just track watts and watt hours.

Ahh, in post #9 you suggested that he's simply wrong, and you ended your post with "I just didn't like that guy".

Sometimes it's a good idea to find out who someone is and what their credentials are

 

[Ampster]

I am sure the manufacturers are already ahead of him with their fuel gauges that have to factor in consumption to come up with a miles of range on an EV. I agree with @GXMnow that we don't his complicated math to calculate energy storage.

 

[Bob B]

Ahh, in post #9 you suggested that he's simply wrong, and you ended your post with "I just didn't like that guy".

Sometimes it's a good idea to find out who someone is and what their credentials are

I didn't say he is simply wrong .... I don't agree that intermittent voltage sampling is a good solution .... I think constand voltage sampling and calculating watts is the simple answer .... and whether I like him or not has nothing to do with it. But ... I do think his head is toooooo pointy. Sometimes it is necessary to see the forest instead of the trees.