RC lipo SOC @3.5V

[wesnewell]

Been a lot of talk about when a lipo cell is empty lately. The SOC @3.5v is 3-4%. AT 3.6V it's ~7%. These are resting voltages, not under load. All indications are it will be completely drained of any reasonable energy usage at 3.4V. From charts posted all over the place, one can see that the cliff effect starts at ~3.62V under a light load. This is why an LVC of 3.5V per cell is the lowest LVC should be set. That's 42V for a 12s pack and 35V for 10s pack. Any lower than this is just asking for trouble and shortening cycle life.
With all the BS about lipo not be fully discharge until you hit 2.7V, here's the real facts.


I hope this settles that. BTW, never run your lipo this low.

 

[Ykick]

Yep, I don't even like 3.5V/cell. As you see 3.65V (or so) is when most will see the cliff rapidly approaching. Without proper BMS I don't ever dream about going under 3.65V.

My Turnigy 20C cells approaching 3 years and unbelieveable charge cycles still doing pretty good in the recent 20F cold spell. I thought for sure they'd need to be refreshed but so far, so good....

 

[spinningmagnets]

I saved this chart from a recent post. Not many miles by going over 4.10V, and then below 3.65V. Also notice one cell is stronger than the rest, one cell is weaker, and four of them are well-matched as far as capacity and resistance. (if I'm reading this right). I believe I saved this from a Neptronix post.

 

[etriker]

Yep, I don't even like 3.5V/cell. As you see 3.65V (or so) is when most will see the cliff rapidly approaching. Without proper BMS I don't ever dream about going under 3.65V.

My Turnigy 20C cells approaching 3 years and unbelieveable charge cycles still doing pretty good in the recent 20F cold spell. I thought for sure they'd need to be refreshed but so far, so good....

I agree. Not running Li-ion over the cliff makes them last longer.

 

[dogman dan]

I agree with all said. It's worth noting that when you go to calculating your potential range before you buy your batteries, that you won't get 5 ah from a 5 ah pack if you treat them nice. Ideally, you'd carry enough battery to avoid deep discharges every cycle. Some do discharge to 3.5 every ride, but I like the luxury of being able to have a cell that is low not go below 3.6v, by stopping when the pack has more left in it.

I like rules of thumb, to save my brain from having to think hard, so I like to think of a 5 ah lipo pack as having 4 ah useable. There is more of course, and sometimes I have used it all. But when my packs were fresh, my 10 ah pack was kept from discharging uncomfortably deep when I stopped by about 8 ah used. I'd be in the 3.75-3.8 ballpark, starting from 4.15 or so, after charging the day before to 4.2v.

So where I dirt ride, I'd tend to head for the house two miles away when I saw about 7 ah discharged. Mostly street to get home, I'd make it home with only 8 ah showing on the CA. If way out in the desert, I'd turn back at 4 ah used.

Now that my lipo is 2.5 years old, My max is about 6 ah, and I am now starting from 4.1v. This will bring me home at 3.6-3.7v.

 

[friendly1uk]

Wes, you have sent me here, after many times of me asking, to see your proof that 3.3v is flat. Where is it?

This looks like you told your device to discharge 5Ah from your pack. Then read the voltages. Which is just to funny for words.

 

[dogman dan]

His wattmeter says so. There is some below that voltage, but not much. Nothing worth the risk of cell damage to get.

So it's effectively at 0%, even if there is a few wh left if you are willing to go to 2.7v.

However, better quality cells might go to 2.7v without damage. But I don't trust HK cells to do that.

 

[Hillhater]

!00 % agree that there is no point in going below 3.6 v
..but the wattmeter readings are meaningless and prove nothing.
The graph says all that needs to be said.

 

[friendly1uk]

This does not prove they are empty at 3.3v as he keeps banging on about, It proves the batteries had more than 5Ah to offer.

 

[johnrobholmes]

But this does not prove they are empty at 3.3v as he keeps banging on about, It proves the batteries had more than 5Ah to offer.

Empty is a relative term. I can eat more watermelon if I eat into the rind, but that doesn't change my mind that the watermelon is done when the red sweetness has been eaten.


You can either pull all the possible watt hours from a battery, or you can discharge them to a level where useful power is gone and long term life will be less affected. In my opinion, Lipos are empty on an ebike when the first cell hits 3.5v. Beyond this point the weakest cells get beaten on and capacity degrades quickly.


I have four years of use out of one 72v10ah brick, charged to 4.1v per and discharged to 3.5v. Still kicking over 8ah.

 

[DrkAngel]

I tend to map cell capacity then keep usage within the beefier areas.

I began testing by mapping mAh /100th V discharge - too time consuming and inaccurate, IMO.
Voltage graph shifted almost .1V lower than actual rest voltage.
See 4th graph ("Recycled Li-ion" - red) for direct comparison.

RC LiPo displayed a noticeably lower area of good capacity

So I decided to remap using a mAh/10thV

Tools -
MeanWell 24V bulk charger (19.8 - 29.8V adjustable)
30V 4 digit volt meter, 100ths capable
Ah meter

I discharged my 25.9V recycled Lipo pack to 24.5V.
Then precisely equalized the cells at 3.50V.
I applied charge with MW (MeanWell) set to:
25.20V (3.60V) - full charge required .27Ah
25.90V (3.70V) - full charge required .53Ah
26.60V (3.80V) - full charge required 3.87Ah
27.30V (3.90V) - full charge required 3.15Ah
28.00V (4.00V) - full charge required 4.60Ah
28.70V (4.10V) - full charge required 4.10Ah
29.40V (4.20V) - full charge required 1.85Ah
(Ah is capacity between each .10V)

Telling, but lacking resolution.
So I mapped in 2/100thV increments. (50th V)
1st discharging to a stable 3.50V per cell
Then setting MeanWell to 2/100thV higher (3.52 etc), then charging through a mAh meter and recording mAh required to stabilize voltage at higher voltage.
Then repeating test in 2/100thV increments.
These non-RC Lipo displayed a strange capacity bulge near 4.12V, so I repeated the same test with another sample ... with very similar results.
(Rated optimal use at >3.75-4.05 to 4.15 for trips.)

Li-ion 18650 (laptop cells) look optimal with a >3.70V to 4.05V for an estimated 300% lifespan.

An alternate source displays very similar results in a more familiar graph style.

Every formulation is different and you should test you own cells.
Discharge to a stable lowest voltage you want to test.

After finding no noticeable capacity at 3.5V, I began starting my tests at 3.6V.

Results show barren and rich regions of energy capacity. At the upper (4.2V) regions 1/10thV reduction in charge voltage can double usable cycles with a ~10% loss in capacity, a similar doubling for each 10thV reduction.
It seems deterioration is more determined by voltage than Ah.

Every 1/10thV of additional charge cuts usable life in half! = reasonably proven and generally accepted.
Every 1/10thV of additional discharge also cuts usable life in half? = reasonable assumption!

"Recycled Li-ion" Red graph, - discharging from 3.72V to 3.70V might have the same available mAh as from 3.70V to 3.00V = 35 times the damage to scrape out the miniscule remaining mAh.

Confirming might take years of testing .... I prefer to just keep my use within the energy rich regions.

Most everyone agrees that deeply discharging is damaging.
Discharging where there is minimal energy = needlessly-wastefully damaging! IMO!

 

[teslanv]

I can eat more watermelon if I eat into the rind, but that doesn't change my mind that the watermelon is done when the red sweetness has been eaten.

Well said JRH.
I'm inclined to leave a little pink on the rind myself.

 

[liveforphysics]

If you call ~3.5v 0% SOC, you get to use your Lipo and get a decent lifespan.

You can actually continue to discharge below 2v/cell. You completely de-lithiated the anode and collapse it's structure, and then become a copper - ion battery as the copper anode foil is the next ion donor source to continue to drive current.

So, you can say true 0% SOC doesn't occur until ~0Vdc if you want to be anal about it.

If you want a good choice for cells you wish to use for more than a 1 time true 100% DOD, 3.5Vdc is where I call them 0% SOC for any performance application using a cell similar to a Nanotech. They gas like little heat-shrink splitting sausages if you try pulling > 30C discharge at lower SOCs.

 

[Punx0r]

Thankyou, Luke.

It would be interesting to know why the manufacturers of Wes's wattmeter programmed it to indicate 0% SOC at 3.5V. It could be a an industry guideline, the results of actual research or just a number they pulled from the air.

 

[dogman dan]

There's a way to solve this argument, but neither of you will like it. You are just going to have to agree to disagree.

We have plenty of Lico is playing with fire warnings on this forum. Readers make their own decisions. There is no one only way to put together a battery and use it.

Thanks for a shot of sanity here Luke. And also JRH. I don't like the white watermelon either, unless it's the only food. If I must eat the rind, I eat it very slow, like 3 amps.

Personally, I would like to start using a bms again. But I'd want the lvc set to about 3.3v, so it would shut me down right about 3.5v.

I continue to run naked, so that 6 different 14s lico packs can be configured in many ways, all 45 ah on one bike, or as little as 5 ah on a bike. I can't afford 6 bms's but I'd like to set up at least 15 ah on a bms. But meanwhile, I am NOT running my battery unsafe. But I DO have to pay close attention to it. It's like a second hobby watching those batteries.

 

[johnrobholmes]

Thankyou, Luke.

It would be interesting to know why the manufacturers of Wes's wattmeter programmed it to indicate 0% SOC at 3.5V. It could be a an industry guideline, the results of actual research or just a number they pulled from the air.

A number pulled from the air, but good enough for estimates. SOC% based on resting voltage alone is close to a wild ass guess. Even though, 4.2v per cell is generally considered 100% but many lipos will take an additional 10% of capacity if you take it to 4.35v. At the bottom you could pull some energy past 3.5v, but as we have covered it is very diminishing returns and it reduces the capacity quicker.


SOC is a frame of reference that isn't rigid. There is no real defined beginning and end, but boundaries can be set depending on the user needs. Like Luke said, technically we could call 0v the true 0% SOC. 100% could be the point right before the cell goes into thermal runaway.

 

[Punx0r]

Completely agree with that.

I would strongly considering using a BMS (subject to reliability, price and discharge capability) if the LVC and HVC was adjustable (the later being more important). It would be nice if there were trimmers to set the HVC and LVC reference voltages.

The BoC configuration (BMS on Charger) of an RC charger permits this to some extent (4.1 vs 4.2V) and also gives you a visual display of each cell voltage. So you can tell at a glance how unbalanced the pack is at start of charge, what state of the balancing is at and which cell, if any, is at issue.

You don't get this with a BoB (BMS on Battery) configuration but you do get cell level over-discharge protection. As non-adjustable feature it will act as an emergency device to prevent unrecoverable over-discharge.

For my needs the latter can be achieved using an inexpensive LV alarm. BoC means two multi-pin connectors for charging, BoB means a single twin-pin connector.

 

[wesnewell]

Thankyou, Luke.

It would be interesting to know why the manufacturers of Wes's wattmeter programmed it to indicate 0% SOC at 3.5V. It could be a an industry guideline, the results of actual research or just a number they pulled from the air.

If you read the original post you will see it doesn't read 0% soc until it gets to a little less than 3.3V per cell average. These percentages were verified on my Hyperion 1420i for confirmation of accuracy. The only manufacturers recommendations I've ever seen said to not go below 3.5V resting, though they recommended not to go below 10% soc for longevity. This was a few years ago and I don't recall which manufacturer it was. I wish I did. That's ~3.7V per cell according to my test. I've got my controllers LVC set to ~88V for my 10ah 24s pack and on the few occasions I've run it down that low, my Hyperion shows it right at 10% soc.

 

[Punx0r]

Oops, yes, 0% SOC @ 3.3V

The point was more about the precedence of the value.