PDA

View Full Version : Battery side current draw / State of Charge measurement



DaveAK
03 December 2010, 1143
First off a question. What kind of maximum current draw are you guys seeing from your batteries? (From a typical street bike perspective, not from a 1/4 miler!)

I've been looking at all the offerings that various chip manufactures have and it seems like Coulomb counting is the way to go. From what I can gather they achieve this by sampling the current at a given period and basically summing as they go. Well, I can do that easily enough, right? I can think of several ways to do it with varying degrees of accuracy.

So, next question. Let's say that I have 8 10Ah LiFePO4 cells in series at a manufacturers suggested charge level of 3.65V, with a LVC of 2.7V. This gives me a 8x3.65x10=292Wh pack at full charge. When my cells have depleted to 2.7V, has my Ah also been lowered? And if so by how much? If I say that the battery is now empty can I consider my Ah as zero, (or thereabouts)? (8x2.7x0=0Wh). Voltage drops by 25% from a full charge to what's considered empty but you get more than 500Wh out of a 2kWh pack, don't you?

When I start counting Coulombs I'll start with the assumption that my 10Ah pack is 36000C and I'll count the Coulombs until I hit LVC, so I'm sure I can figure out the answer to my questions above, but it would be nice to know that I'm heading in the right directions, and not missing something crucial. :)

frodus
03 December 2010, 1329
don't use 3.65. Use the cell nominal voltage. This isn't the rest voltage neccessarily (headways rest at 3.3ish volts, but when you lightly load them, they pretty quickly go to 3.2V). Use the manufacturer nominal voltage and multiply to get Wh.

Here's some great reading about calculating SOC:
http://liionbms.com/php/wp_soc_estimate.php

Elithion uses a combination of coulomb counting and voltage as a way to measure SOC.

DaveAK
03 December 2010, 1336
Thanks for the link studying it now. Fair enough on using nominal voltage to calculate pack, but the question remains the same.

DaveAK
03 December 2010, 1340
Well a quick scan through that document makes me think that I'm more cleverer than I thought. I'd already thunk up all that stuff. :)

jdh2550
03 December 2010, 1502
Find the discharge curves for your cells and that will show you the voltage under load at different SoC. You'll see it's very flat. The only use of voltage is at the end points to control over charging and over discharging.

Here's the discharge graph for TS LFP's:http://www.thunder-sky.com/pdf/201072313435.pdf It's the top left graph on the second page.

One other thing to bear in mind is that not only do you not want to drain below 20% SoC you don't want to charge to 100% SoC. They used to have a charge curve on this page as well (it shows the voltage "shooting up" at the end just like it "falls off" at the end of the discharge cycle). The point being you get a rapid change in voltage but little change in stored energy at either end of the range. And it's in those areas that the cell is under most stress - keep out of those areas for extended battery life.

Not really an answer to your question - but I've got to run...

BaldBruce
03 December 2010, 1938
Are we making this problem more complicated than it needs to be? Why not simply measure power out from each cell and sum for pack total???? Why guess or do any fancy algorithims, when you can simply start with a reset based on hitting your HVC. System then measures and subtracts from a programed in starting point. LVC is still necesary in case of error, abuse, or decreasing capacity over time. Maybe I'm just sloooowwwww?

DaveAK
03 December 2010, 2020
Are we making this problem more complicated than it needs to be? Why not simply measure power out from each cell and sum for pack total???? Why guess or do any fancy algorithims, when you can simply start with a reset based on hitting your HVC. System then measures and subtracts from a programed in starting point. LVC is still necesary in case of error, abuse, or decreasing capacity over time. Maybe I'm just sloooowwwww?
That's exactly what I'm talking about isn't it? I wasn't talking about guessing anything was I? My only real question is if my 10Ah pack will give up every one of its 36000 coulombs as voltage goes from HVC to LVC. I'm guessing not, (OK, so I am guessing something), but are we talking in the same ratio as LVC/HVC or not?

As for measuring power vs. coulombs, I'd still use the same method more or less - current x voltage vs. current per second. Which brings me back to the first question, what magnitude of current am I going to have to be able to measure? I know my cells will do 3C which would be 180A, but what are guys with larger packs seeing? Am I likely to see more than 180A when I'm being a little heavy handed on the throttle? :D