Help determining Charge% for 74V Battery

[Nuckingfuts]

I may not have enough info to go on but I've got a 74V 55Ah battery for my Surron Ultra Bee. When it's 100% charged (according to the bike display and after charging) I get a voltmeter reading of 76.0VDC. As a second data point, when the battery is discharged to 45% I get a reading of 66.5VDC. This doesn't correlate with any of the 74V battery charts I can find. Do these values make sense to you? Is my charger only charging to a much lower percentage than indicated by display??

Searching which cells are used by Surron I get the A-grade Samsung Li-Ion INR 21700 cells, which appear to have a nominal voltage of 3.63V (LiMn ??) but I can't confirm if this is accurate. If this is indeed the case, then I would think this 74V pack should have a voltage north of 80V when charged to 100%???

I tried to study up on the basics but I'm still learning so any help to understand is appreciated.

 

[amberwolf]

Is the battery chemistry the same as that the display meter is calibrated for?

Is the battery new, or used?

Is it the original for that specific bike, or a replacement? (I ask because you said you got the battery *for* the bike, not *from* the bike, which implies a replacement)

 

[Nuckingfuts]

Is the battery chemistry the same as that the display meter is calibrated for?

Is the battery new, or used?

Is it the original for that specific bike, or a replacement? (I ask because you said you got the battery *for* the bike, not *from* the bike, which implies a replacement)

First I want to say I appreciate the attention to syntax. Second, yes the battery is the OEM battery that came equipped with the bike up on original purchase and can only assume that the bike display and controller are calibrated for this battery. Battery has only ~36 miles of use spread across three charges, each taken from 100% down to about 50% (i.e. 12 miles per ride and charge cycle).

 

[amberwolf]

Since it's original, then yes, presumably the system ought to be calibrated for it.

So if it is using the cells you found data for, then they should charge to at least 4.1v, if not 4.2v, but this depends on how they setup the BMS--if they set it up for long pack life then they might've used a lower full voltage (and possibly a higher empty one too), so that you only actually get some large fraction of the cell capacity--this is how large-EV packs are designed, to maximize lifespan, and minimize expensive warranty work.

But we'll assume for this discussion that it's like nearly all the other small-EV packs and maximizes capacity and capability over pack life, and charges to full. So if it charges to 4.1v, then a 20s (72v Li(XX) pack would be 82v. 4.2v would be 84v. If it discharges to say, 3.0v, then that would be 60v empty. If it's 2.8v, that's 56v.

Charge curves aren't linear, being somewhat linear from say, 25% to 75%, with some usually very curved portions at the ends, rapidly falling at empty and rising at full.

Not knowing what actual cell is used in there, can't point you to a tested curve for them, but this site has several Samsung 21700 cells; this is the last one on their list; it shows the discharge curve at several different currents for two samples of the cell.

Battery test-review 26650 comparator

Test of Samsung INR21700-50G 5000mAh (Green)

We don't know what the bike manufacturer considers, voltage wise, for any particular capacity percentage, though, or if they are actually using a "coulomb counter" that measures true capacity in Wh or Ah. (there may be info on that in the big Sur-ron thread somewhere). If they're using a simple voltage-percentage table, then while it isn't as "accurate" (for instance, it will show lower than actual capacity while under load), it won't need to change over time to match what the battery is actually capable of; the coulomb counter has to be periodically recalibrated (possibly automatically) for the true battery capacity.

Assuming they're just going by voltage, then based on the above chart for that one particular cell, 76v / 20s = 3.8v / cell. Assuming a static no load test to get that, then that's a bit above half full, near the 2.0Ah / 4.8Ah used mark on that chart on the 0.2A load (lowest available load data). 66.5v / 20s = 3.325v/cell, which is on the "cliff" at about 4.3Ah/4.8Ah used.

So, one of the assumptions has to be wrong (is it a 20s pack? is it Li(XX) chemistry? is it a voltage-based meter? ) etc. or the system or battery has a problem of a kind I can't think of ATM.


Is it safe to assume you're not actually having a problem other than the display percentage is wrong?

 

[E-HP]

Since it's original, then yes, presumably the system ought to be calibrated for it.

I use voltage readout and simple math to determine the battery charge, and use the display to determine the remaining bars and remaining bar percentage. I think the percentage on displays is usually fairly well calibrated to the bars.

 

[Nuckingfuts]

Since it's original, then yes, presumably the system ought to be calibrated for it.

So if it is using the cells you found data for, then they should charge to at least 4.1v, if not 4.2v, but this depends on how they setup the BMS--if they set it up for long pack life then they might've used a lower full voltage (and possibly a higher empty one too), so that you only actually get some large fraction of the cell capacity--this is how large-EV packs are designed, to maximize lifespan, and minimize expensive warranty work.

But we'll assume for this discussion that it's like nearly all the other small-EV packs and maximizes capacity and capability over pack life, and charges to full. So if it charges to 4.1v, then a 20s (72v Li(XX) pack would be 82v. 4.2v would be 84v. If it discharges to say, 3.0v, then that would be 60v empty. If it's 2.8v, that's 56v.

Charge curves aren't linear, being somewhat linear from say, 25% to 75%, with some usually very curved portions at the ends, rapidly falling at empty and rising at full.

Not knowing what actual cell is used in there, can't point you to a tested curve for them, but this site has several Samsung 21700 cells; this is the last one on their list; it shows the discharge curve at several different currents for two samples of the cell.

Battery test-review 26650 comparator

Test of Samsung INR21700-50G 5000mAh (Green)

View attachment 350166

We don't know what the bike manufacturer considers, voltage wise, for any particular capacity percentage, though, or if they are actually using a "coulomb counter" that measures true capacity in Wh or Ah. (there may be info on that in the big Sur-ron thread somewhere). If they're using a simple voltage-percentage table, then while it isn't as "accurate" (for instance, it will show lower than actual capacity while under load), it won't need to change over time to match what the battery is actually capable of; the coulomb counter has to be periodically recalibrated (possibly automatically) for the true battery capacity.

Assuming they're just going by voltage, then based on the above chart for that one particular cell, 76v / 20s = 3.8v / cell. Assuming a static no load test to get that, then that's a bit above half full, near the 2.0Ah / 4.8Ah used mark on that chart on the 0.2A load (lowest available load data). 66.5v / 20s = 3.325v/cell, which is on the "cliff" at about 4.3Ah/4.8Ah used.

So, one of the assumptions has to be wrong (is it a 20s pack? is it Li(XX) chemistry? is it a voltage-based meter? ) etc. or the system or battery has a problem of a kind I can't think of ATM.


Is it safe to assume you're not actually having a problem other than the display percentage is wrong?

I apologize as I miss quoted, the Surron ultra bee battery spec is 74V/55Ah. None the less I follow your logic and thank you for the detailed write-up. Ive searched quite a bit and haven't been able to find any confirmed specs on the cells used or anyone who has torn down this battery case to see what's inside. I'm not currently experiencing any issues with the battery and the range I'm getting is consistent with other users but I'd really like to know if my charger is set to place it excessively safe and there's more range I can safely get from this pack.

The charger/battery combos I have from EM3EV for my ebikes charge to 90% to get the best range:battery life, Id like to know if I'm getting a charge somewhere in this range. My voltage reading at 100% sure doesn't look like it and according to those curves it's only charging to ~70% at best.

 

[amberwolf]

Well, a "74v" battery isn't one of the "standard" Li-xx nominal voltages, terminology-wise, so I don't know how many series cells it uses, therefore I don't know what it's full charge or empty is.

So...it probably is not a 20s battery, and they might actually be using it's real full-charge voltage as the battery voltage listing. If so, then 74v is probably the full voltage of an 18s battery that charges each cell to 4.1v full.

In that case, you'd have to see if there's a cell (LiMn, NMC, etc) in that database with a 4.1v full charge to better guesstimate capacity vs voltage for that curve, and then mulitply that by the 18s instead of 20. I'd guess that this will line up better with the percentage you get vs voltage.