Safe RC-LiPo undercharge when bulk charging without BMS?

My next 77.7V-nominal (88.2V-hot) RC-LiPo pack will normally be balance charged through the balance leads, but carry an onboard low-current bulk charger for occasional range-extension/emergency use.

Assume cycle life at 4.2V/cell is acceptable.
Assume cells are matched, and balanced before a single (partial) discharge, then bulk charged.

How much would you undercharge to? 4.15V (87.15V-hot)? 4.10V (86.1V-hot)?

Of course, the definitive answer is to test and measure, but I'm curious for current practices or opinions.

fatty said:

Of course, the definitive answer is to test and measure, but I'm curious for current practices or opinions.

Click to expand...

I go all the way with balanced cells. Right to 4.2v / cell with well behaved cells.


Not very well behaved cells watch like a hawk. I wouldnt bring them under 3.8-9v/ and wouldn't go over 4.1v...

The worse the cell(s) bottomed out, the worse the imbalance is... you know that and are staying away from this.

For example my big cells, the Chevy Volt ones, have been in my bike for 6 months plus without a balance. All are still very well balanced. The curve drops off around 3.3v-3.2v. No problem shoving a 1C charge rate into them for months now, still all wihthin 0.001v of each other.

My Lipo reserve pack, ( 5Ah, 72v, 20s) is another story. As soon as cells go under 3.8v/cell... look out. You gona have imbalance. But I have been doing ok bulk charging them too.. you just have to watch. the lesser capacity cells WILL rear their heads by showing the voltage incongruousness. It took me like three or four bulk charges to throw them out and require the normal balance to charge a cycle. Where the EV cells have been using have taken 80+ charge (full) cycles, (over roughtly 250 charging periods, ) and never went empty. I do empty my hobby lipo 5Ah pouchs ( Pulse 65C 5Ah x 20s) though... it is easy in 4-5 miles of reserve.

If they never go empty, you are fine Id go as far to say. In the RC hobby where I beat these things regularly, and balance charge on good chargers, they still die in a year or two under abuse.. IF you go under 1/2 capacity. That means,:

Example: 5 Ah pack, never use more than half the mAh .. Ie 2500mAh. Then you are safe.

Never use more than 1/2 your capcity... never a problem.

good cells 4.2 is okay

don’t let sit there for long

use almost immediately

reduces degradation


not sure when next use

stop around 4.0-4.1

then quickly top off

for full capacity if needed


50% dod is smart money

Thanks, much appreciated.

Since the on-board charger may be used for low-SoC "emergencies" (though not over-discharge), but I only expect to use it in that emergency case, and only within one discharge of a balance, I think I may split the difference and set for 4.15V.

Of course you are assuming cells are balanced at the top

which means it is at the bottom of SoC% where the "unevenness" is revealed and lack of a BMS needs to be compensated for by setting LVC higher.

Those who bottom balance their cells and go bareback no BMS, can relax a bit more if sometimes they need to "limp home" overriding their usual pack-level LVC.

And can use cell level monitoring attached only while charging to stop when the weakest cell hits 4.2V first.

I think I'm following you, but I want to confirm:

john61ct said:

which means it is at the bottom of SoC% where the "unevenness" is revealed and lack of a BMS needs to be compensated for by setting LVC higher.

Click to expand...

Isn't this always where unevenness is revealed? Or do you mean you'll notice it when bottom-balancing vs top-balancing?

john61ct said:

Those who bottom balance their cells and go bareback no BMS, can relax a bit more if sometimes they need to "limp home" overriding their usual pack-level LVC.

Click to expand...

Only in so far as you are familiar with how uneven the pack becomes, if you bottom-balance.

So you're referring to the user's knowledge of the pack, correct?

The spot in the SoC% curve where you choose to balance the cells, is the only spot where they will match precisely.

The manufacturer usually IME delivers them balanced at the midpoint, close to the (rounded) nameplate nominal voltage.

In that case (IMO ideally left that way) there will be imbalances revealed both at the top and bottom of the cycles, and either

1 both HVC and LVC are based on a per-cell based trigger, usually that being the core function of a BMS.

Or, assuming the owner has the gear and knowledge to calibrate just how much to stay away from the pack-level voltage "shoulders" at each end, and the cells are of good quality, well matched capacities and still in good condition

2. Setting pack-level LVC high enough and pack-level charging setpoint / HVC low enough, provides a safety margin for the weakest cells to be protected.

Even if using #1, using #2 **as well** will **greatly** extend cycle-count lifespan, with the top margin at say 4.05V, and for the bottom higher the better, a small difference in LVC can double lifespan if C-rates aren't too high.

fatty said:

john61ct said:

which means it is at the bottom of SoC% where the "unevenness" is revealed and lack of a BMS needs to be compensated for by setting LVC higher.

Click to expand...

Isn't this always where unevenness is revealed? Or do you mean you'll notice it when bottom-balancing vs top-balancing?

Click to expand...

A bottom-balanced pack will have zero unevenness at that point, and maximum unevenness at the top, fully charged - the top-balancing function of any BMS and chargers must be disabled to preserve the balanced state at the bottom.

A top-balanced pack (the norm these days) will correspondingly have maximum unevenness at the bottom.

fatty said:

john61ct said:

Those who bottom balance their cells and go bareback no BMS, can relax a bit more if sometimes they need to "limp home" overriding their usual pack-level LVC.

Click to expand...

Only in so far as you are familiar with how uneven the pack becomes, if you bottom-balance.

So you're referring to the user's knowledge of the pack, correct?

Click to expand...

Well everything relies on that with no BMS, being conservative about preserving pack health means sacrificing more range.

The BMS is what lets you get maximum range and still protect the weakest cell/groups.

An ignorant user going bareback (no automated cutoff) **and** flying blind (no cell level monitoring) needs to sacrifice the most capacity, or risk murdering* the pack.

* the definition of damage IMO includes dropping lots of potential lifespan, not necessarily immediately detectable

john61ct said:

[info]

Click to expand...

I think I did misunderstand -- I thought you were referring to the user's observation of cell (im)balance before recharging, such that he/she would only observe that imbalance when connecting the charger (and thus at low SoC).
Thanks for explaining -- all sounds correct/in agreement.
I edited my OP for clarity.

Proper balancing only pertains to the chosen balancing point.

That's why I used "unevenness" for away from the chosen balancing point.

Imbalance at any other point is not actionable or even meaningful data, the cells are **supposed** to be uneven there.

Only midpoint balancing minimises the unevenness at both top and bottom.

john61ct said:

That's why I used "unevenness" for away from the chosen balancing point.

Click to expand...

Interesting -- I hadn't heard that terminology distinction before, but it makes sense.

Yes many think that once a pack is "balanced" that means the differences in voltage go away at all SoC% points

Or even worse that it is possible to balance at more than one SoC% point.

I mean it "is possible", but counterproductive; balancing say at 3.65V "undoes" your top or bottom balance.

Just pick a strategy and stick to it consistently.

There are balancers that they claim can be left going, "transferring" from the higher to lower all the time, in theory increasing overall capacity

but IMO that's snake oil, a false goal in the first place.

What's "undercharge"?
Been using LiPoly on my ebikes and now, a boat, for 9 years and here is how I do it.
Always stay at 1P, if you need more capacity, use larger capacity bricks(pouches). If I need more than my pair of bricks will provide, I carry another pair of bricks and swap them out. I never parallel.
I always store at 3.90V/cell, maybe 3.95V if I'm going to use it within a couple of hours or so.
I keep the cells within .03V of each other, this is probably somewhat obsessive.
I top charge to 4.10V - 4.15V as the cells start to stray above that and it takes too much time to charge/balance to 4.20V for the little that is added.
I never let the cells get below 3.65V, again, they start to stray(dive off the deep end) below that. Actually, I almost always am done around 3.80V. If you find yourself regularly flirting with the deep-end, carry more capacity.
I store them in the fridge, they will store forever without degrading.
I don't count cycles, but they last years.
It's that simple.

motomech said:

What's "undercharge"?

Click to expand...

Just undercharge from the 4.2V max, like your 4.10-4.15V.

motomech said:

Always stay at 1P, if you need more capacity, use larger capacity bricks(pouches).

Click to expand...

I read this in another post of yours, but I'm not clear on why?
With cells in parallel, capacity dispersion is averaged out.
Live example attached. Defaults to 8000mAh +/-1%.

Note that 4.15V or even 4.05V can still get to the same definition of 100% Full

as using a 4.2V or even 4.35V charge setpoint.

Not of course if you are using a Bulk / CC-only "charge to X and stop" profile, which works just fine and in fact is what I recommend for normal daily-usage cycling

And not if you are holding CV / Absorb stage until amps trails down to zero, which I do not recommend ever.

But, if holding CV / Absorb stage until a certain endAmps spec

the combination of that spec with the CV voltage spec, can get you to whatever %capacity utilization you want

at any voltage setpoint above say 4.02V.

fatty said:

motomech said:

Always stay at 1P, if you need more capacity, use larger capacity bricks(pouches).

Click to expand...

I read this in another post of yours, but I'm not clear on why?

Click to expand...

So long as the paralleling is done at the lowest level 1S voltage first, you are correct in theory.*

However, assuming any combination was easily available at the same price per kWh and the same energy density

getting to say 180Ah, which would you choose from a practical POV?

A. a single string of 180Ah cells

B. a single string of groups of 100 1800mAh 18650 cells comprising 180Ah in parallel

C. a single string of 90Ah pairs in parallel

D. two strings, serial'd at the pack level voltage, each a set of 90Ah cells?

E. six strings, serial'd at the pack level voltage, each a set of 30Ah cells?

______
Only A and D are "optimal", D is a compromise.

Left as an exercise for the reader, try to describe why, there are several factors at play.

* D & E both violate this guideline

fatty said:

motomech said:

What's "undercharge"?

Click to expand...

Just undercharge from the 4.2V max, like your 4.10-4.15V.

motomech said:

Always stay at 1P, if you need more capacity, use larger capacity bricks(pouches).

Click to expand...

I read this in another post of yours, but I'm not clear on why?
With cells in parallel, capacity dispersion is averaged out.
Live example attached. Defaults to 8000mAh +/-1%.

Click to expand...

I should have been more specific.
I don't like to parallel so I can bulk charge w/ my indestructible Mean Well and ck and balance w/ Battery Medics. And Battery Medics will not balance the cells correctly if they are paralleled.
The RC chargers never seemed to last much more than a year in an ebike application and I used some name brands. Yes, I used the 2 into 1 Y-style balance lead adapters(And those can be tricky, better have the paired cells dead-on even, or there can be an in-rush w/ melted wires), but what I think was hard on the RC chargers was this scenario; Say if you had a pair of 5,000mAh cells paralleled, in other words a 10,000mAh cell and it had gotten off the others, the charger worked too hard to bring it back in-line. I found it to be better to break the pack down and charge in 1P
And I feel the less breaks, either S or P, one has to make w/ LiPoly, in general, the better.
To me, the whole thing about using LiPoly on ebikes is not about fires and safety, that's been blown way out of proportion The biggest con is time, with the multiple steps, it can be time consuming and by not breaking down a pack to charge is one less step.
It also makes for less wires, paralleling can make for a rat's nest real quick.
And while I won't leave charging LiPoly completely unattended, I have no problem leaving the Battery Medics on while I'm not watching them.
So what I'm saying is, to keep w/ the KISS plan, use the biggest bricks that will fit in the space.
But, and this a big but, when the Multistars were discontinued, it hurt the ability to implement this approach. They were available in bricks of 10,000mAh, 15,000mAh and even 20,000mAh and Hobbyking had them on sale all the time. They were such a good deal, that I used to recommended to folks here to consider LiPoly for an ebike application, something I no longer do because of all the flak I used to receive about fires and all that. In fact, several years ago, I bought 8 MS 6S/10,000mAh bricks for $50 (shipped) ea. 4 made a 12S/ 20,000mAh group for $200! At the same time, I bought a bunch of Turnigy 6S/5,000 25C bricks that were on sale for $40 ea. For energy density/ Dollar value, It's hard to beat the Turnigy 25C when it's on sale.
But now days, I don't think that HK puts much of it's LiPoly on sale, but since they really screwed-up their web site, I no longer browse their selection.

Packs in parallel can connect the balance leads "across" so one balancer handles them all at once.

If the only connection is at the pack-level voltage main power leads

then you get unbalanced current flows causing uneven wear shorter lifespan.

But **cells** paralleled at the 1S level, then those groups connected all in a single string, no problem so long as you don't get catastrophic failures per cell internal shorting.

Lots of people series-connect high C-rate hobby LiPo packs to get up to high voltages

and break those connections to enable use of cheap 6S or 8S balancing chargers

10S and 12S units getting rare and pricey these days[emoji19]

Of course combining parallel and serial connections of already-built LiPo packs would indeed be a challenge to do so that breaking/reassembly is bot convenient and safe.

john61ct said:

Note that 4.15V or even 4.05V can still get to the same definition of 100% Full
as using a 4.2V or even 4.35V charge setpoint.
Not of course if you are using a Bulk / CC-only "charge to X and stop" profile, which works just fine and in fact is what I recommend for normal daily-usage cycling
And not if you are holding CV / Absorb stage until amps trails down to zero, which I do not recommend ever.
But, if holding CV / Absorb stage until a certain endAmps spec
the combination of that spec with the CV voltage spec, can get you to whatever %capacity utilization you want
at any voltage setpoint above say 4.02V.

Click to expand...

Hm, not with 4.35V since you'd exceed the 4.2V limit before the charger even hits CV. I guess if you mean manually terminating the charge during CC at spec Wh?

john61ct said:

So long as the paralleling is done at the lowest level 1S voltage first, you are correct in theory.*
However, assuming any combination was easily available at the same price per kWh and the same energy density
getting to say 180Ah, which would you choose from a practical POV?
A. a single string of 180Ah cells
B. a single string of groups of 100 1800mAh 18650 cells comprising 180Ah in parallel
C. a single string of 90Ah pairs in parallel
D. two strings, serial'd at the pack level voltage, each a set of 90Ah cells?
E. six strings, serial'd at the pack level voltage, each a set of 30Ah cells?
Only A and D are "optimal", D is a compromise.

Click to expand...

I think you mean A and B are "optimal"?
I require single-string balance charging without reconfiguring the packs, so D. and E. are excluded.
A. requires the fewest connections and has fewest points of failure, but greatest capacity dispersion and most difficult packing.
B. requires the most connections and has most points of failure, but smallest capacity dispersion and most flexible packing.
C. is simply a compromise between the two.

motomech said:

I don't like to parallel so I can bulk charge w/ my indestructible Mean Well and ck and balance w/ Battery Medics. And Battery Medics will not balance the cells correctly if they are paralleled.

Click to expand...

I believe john61 points this out above, but you can parallel the balance leads first and the paralleled cells electrically become one larger cell that can be balanced the same as a physically larger single cell. There is no electrical difference. With LiPo bricks where the actual cells can't be seen, it helps to draw out an electrical diagram.

I do think a lot of the confusion comes from the bad habit of writing the series connection before the parallel connection when describing packs, such as 21s2p even when the cells are paralleled first, which would more properly be written 2p21s.