Proper charging voltage for TrustFire 14500 protected Li-Ion cells

[Gadget_Guru]

Howdy all,

I've been using unprotected lithium-ion cells since early 2004 or so. I usually charge them with my Triton charger, which has a maximum charging voltage of 4.18V, according to my most-trusted DMM (Mastech MAS-345 with RS232 interface). This is not quite the proper full charge voltage of 4.20V, but close enough.

I just received my first protected cells. They are TrustFire "900mAh" (hah!) from DealExtreme. I'm wondering if the cells' protection circuit drops a little voltage during the charge cycle, preventing the cell from reaching the full voltage that the charger provides. This would be a shame, since my Triton is already slightly undercharging unprotected cells.

Has anyone here actually measured the voltage drop across a protection circuit while under various charging currents?

If there is a substantial voltage drop, perhaps charging to a higher voltage to accomodate the drop would gain useful cell capacity, while still never allowing the cell itself to go above 4.20V. If the voltage drop is large enough, though, the protection circuit might not let charging voltage go high enough to get the cell to a full 4.20V before the cell's over-voltage protection kicks in and shuts down the charge.

My otherwise rather useless and perhaps dangerous UltraFire WF-138 10440/14500 charger might be useful for charging protected cells. It peaks at 4.32V before shutting off, with current still at a fairly high level. It's not a proper constant current / constant voltage (CC/CV) charger, as far as I know. I haven't yet plotted the current curve during a charge cycle. But if it's anything like my similar Ultrafire WF-138 RCR123A (16340) 3V/3.6V charger, the current will start out at about the rated 250mA with a fully discharged cell, then slowly taper off to about half that level when the charger finally switches off at full voltage.

 

[mudman cj]

Re: Proper charging voltage for protected Li-Ion cells

You should not exceed 4.20V during charging because if there is a small voltage drop across the circuit, it is likely in the form of resistance, and towards the end of the charging cycle when current is low, the voltage drop will be minimal and therefore the cell will be subjected to the potential at the battery charging terminals.

As for your WF138, I believe that is one of the chargers that overcharges unprotected cells left on the charger. You would be better off not using it for unprotected cells, but they tend to work fine for protected cells, terminating the charge at or below 4.20V.

 

[Gadget_Guru]

Re: Proper charging voltage for protected Li-Ion cells

You should not exceed 4.20V during charging because if there is a small voltage drop across the circuit, it is likely in the form of resistance, and towards the end of the charging cycle when current is low, the voltage drop will be minimal and therefore the cell will be subjected to the potential at the battery charging terminals.

I thought it might be the case that any voltage drop across the protection circuit might be resistance, which is why I asked about measured voltage at various applied currents. If the measured voltage drop at minimum charging current (at the end of a proper CC/CV charge cycle) was known, that small drop could be safely added to the charging voltage to fully charge the cell to 4.20V.

Looks like I'll end up sacrificing one of my protected new cells to "science". I'm just too curious about this issue to let the cell rest unmolested. It will probably soon have a date with a sharp knife and a soldering iron...

As for your WF138, I believe that is one of the chargers that overcharges unprotected cells left on the charger. You would be better off not using it for unprotected cells, but they tend to work fine for protected cells, terminating the charge at or below 4.20V.

I'll have to measure that. I recall that the charging current on my RCR123A version of this charger went to 0, or at least below the minimum current that my DMM could display in the 10A scale, which is 10mA. I'll run the test on my 10440/14500 version with the DMM in the 400mA scale, to see if there is any residual charging current after the charger status LED goes green. The extra resistance of the DMM's current shunt in 400mA mode won't hurt, as it will just keep the cell from going to quite as high a charge voltage.

I just had another thought. My UltraFire charger doesn't do a proper CC/CV charge, and instead just gradually tapers the current right from the start of charging a fully depleted cell. When the charger hits it's shut-off voltage of 4.32V or so, the current is still quite high, much higher than the 1/10 of full charge current that a proper CC/CV charge profile would usually entail. So if charge current is still at say 150mA when the charge terminates, all I have to do is use Ohm's Law to calculate that a .8 Ohm resistor added in series between the charger and cell would limit charge voltage to 4.20V. This would not allow a full charge, of course, since the current never had a chance to slowly taper off to 1/10 of full charge current. But at least the cell would never see any over-voltage, making this charger safe to use.

 

[mdocod]

Re: Proper charging voltage for protected Li-Ion cells

I would not concern yourself with trying to squeeze that last 0.02V into a cell, or whatever...

Many people purposely limit charge voltage to ~4.10V to increase cell cycle life.

Being a few hundredths of a volt UNDER 4.20V is not going to have any meaningful impact on the runtime of a device anyways, over time, the cell will degrade slower with this slightly lower charge voltage, so after a few hundred cycles, the cells would actually probably have better capacity than a cell would had it been charged to 4.20V each time.

 

[Gadget_Guru]

Re: Proper charging voltage for protected Li-Ion cells

I would not concern yourself with trying to squeeze that last 0.02V into a cell, or whatever...

Many people purposely limit charge voltage to ~4.10V to increase cell cycle life.

Being a few hundredths of a volt UNDER 4.20V is not going to have any meaningful impact on the runtime of a device anyways, over time, the cell will degrade slower with this slightly lower charge voltage, so after a few hundred cycles, the cells would actually probably have better capacity than a cell would had it been charged to 4.20V each time.

All very true.

But I'm seeking maximum cell capacity purely for the sake of optimizing my cheap cells. I have an assortment of high-quality used 18650 cells pulled from notebook PC packs, and some cheap new Li-Ion cells of various smaller sizes from Deal Extreme.

I'm just too cheap to spend say $16.00 on a new Japanese 18650 cell that really will meet it's rated 2600mAh capacity when new. So my very cheap (a few dollars each) used Japanese 18650 cells that were rated at 2400mAh when new, and still do 2150mAh when charged to 4.18V and discharged at 1A to 3V will have to do. I might as well get as much mAh as I can out of them, without taking them to crazy high voltage like some of the cheap chargers do.

In my typical use, my Li-Ion cells will die of old age before they see 100 cycles, so eeking out a few extra mAh is more important to me than an extra 100 cycles.

 

[mudman cj]

Re: Proper charging voltage for protected Li-Ion cells

Here is a thread with some data that should put this in perspective. A charger designed for 18650 cells should terminate at around C/20 give or take, so final current would be in the neighborhood of 2200/20=110mA. The data is quite linear, and therefore mostly Ohmic in nature. It can be fit by the equation: Vdrop=0.0252*current-0.0014 with an R^2 of 0.9997. Extrapolating to 110mA gives a voltage drop of 1.3 mV.

 

[Meterman]

Re: Proper charging voltage for protected Li-Ion cells

Looks like I'll end up sacrificing one of my protected new cells to "science". I'm just too curious about this issue to let the cell rest unmolested. It will probably soon have a date with a sharp knife and a soldering iron...

Why killing a new cell? In this thread I've shown the voltage drop across the PCB when discharging the cell at various currents.

You can easily do the appropriate measurement when chargig a cell. Measure the voltage drop across the PCB by contacting the bottom contact and (by pricking through the shrink tube by means of a sharp tip of a test lead) and the body of the cell. But don't hit the flat wire running from top to bottom!

So you can see the voltage loss at any desired current.


Wulf

EDIT: I was too slow!

 

[Gadget_Guru]

Re: Proper charging voltage for protected Li-Ion cells

Why killing a new cell? You can easily do the appropriate measurement when chargig a cell. Measure the voltage drop across the PCB by contacting the bottom contact and (by pricking through the shrink tube by means of a sharp tip of a test lead) and the body of the cell. But don't hit the flat wire running from top to bottom!

Wow, I had almost exactly that idea last night. I intend to cut away a small spot of the shrink wrap, and stick a small NIB magnet to the cell body (safely away from the positive lead under the shrink wrap). I'll let the steel alligator clip on one of my DMM's test leads stick to the magnet, and clip the other DMM lead to the negative terminal on my cell charging holder. I'll stick some tape over the "wound" when done.

I intend to measure voltage drop at various charging currents from 400mA on down.

On another note, the protection circuits on these TrustFire 14500 cells are pretty lousy, anyway. Tested LVC was 2.5V, and HVC was 4.33V, both outside the range I'd really like to see. But the TrustFire cells have tested at higher capacity than the unprotected UltraFire 14500 cells I bought at the same time from Deal Extreme, so I'll live with the PCB's lousy performance, in exchange for higher capacity.

Also, I think I may have damaged the PCB in one of the cells during my HVC test. I had my Triton charger set to 500mA charge rate in the 6V SLA mode, so that voltage would keep going after 4.20V. When the cell's PCB HVC activated at 4.33V, the Triton saw an "open circuit", and the Triton's output voltage shot up to a much higher level. I know very little current was flowing into the PCB at that point, but it may have been enough to do some damage, as the cell would no longer work under load, nor would it charge. I removed the PCB and tested the cell directly, and it works normally. Quite amazingly well, in fact, as I got 730mAh measured discharge capacity after charging manually with a variable power supply to 4.22V (40mA cutoff).

 

[Meterman]

Re: Proper charging voltage for protected Li-Ion cells

On another note, the protection circuits on these TrustFire 14500 cells are pretty lousy, anyway. Tested LVC was 2.5V, and HVC was 4.33V, both outside the range I'd really like to see.

These values must be a bit outside the normal working ranges. That's an emergency brake to avoid fatal results if all normal regulations have failed.

Also, I think I may have damaged the PCB in one of the cells during my HVC test. I had my Triton charger set to 500mA charge rate in the 6V SLA mode, so that voltage would keep going after 4.20V. When the cell's PCB HVC activated at 4.33V, the Triton saw an "open circuit", and the Triton's output voltage shot up to a much higher level. I know very little current was flowing into the PCB at that point, but it may have been enough to do some damage, as the cell would no longer work under load, nor would it charge. I removed the PCB and tested the cell directly, and it works normally. Quite amazingly well, in fact, as I got 730mAh measured discharge capacity after charging manually with a variable power supply to 4.22V (40mA cutoff).

When the PCB reacts by shutting off the cell, there is normally no defect. Application of a sufficient voltage will usually "revive" the cell. But there are a lot of posts in CPF regarding this matter.

Wulf

 

[mdocod]

Re: Proper charging voltage for protected Li-Ion cells

Actually, the 2.5-4.33V range you have found on your PCBs is not unusual. Most of the PCBs on the market, regardless of brand, operate with similar limits....

Personally, I'd like to see the low-voltage cutoff on these PCBs higher, like 3.0V, that way, under a load, we could expect to see the cell rebound to ~3.5V or higher after the load has been removed, this would be healthier for the cell, with the 2.5V low-volt cutoff, under lower loads, (like lower powered LED lights, or multi-mode lights being used on lower settings), the voltage of the cell will not even rebound to 3.0V when drained under the "light" load to 2.5V.

However, the ~4.33V high volt cutoff actually has some logic behind it, and it has to do with compatibility I think. Specifically, compatibility with chargers that do not use constant-voltage at the end of the charge. There are a number of chargers out there, that take cell voltage above 4.20V DURING the charge temporarily by continuing a CC charge method even after the cell has exceeded 4.20V, then the charge is terminated such that the cell is expected to "rest" back down to ~4.20V after it's complete. Usually the cell only spends maybe 15-30 minutes dealing with voltages above 4.20V, as such, the charge method is generally considered acceptable and safe. Building a PCB with limits set say- at like, 4.23V, for instance, would not allow the cell to reach full charge on some of these chargers.

 

[Gadget_Guru]

Here's the results of my discharge tests on a single TrustFire 14500 protected cell. Note that the second and third cycles are sans protection circuit. Although I have not yet tested voltage drop across the protection circuit, it seems that removing the circuit adds very little to cell capacity.

Charging manually to 4.22V does add a useful extra little bit of capacity, though, compared to my Triton's charging voltage of 4.18V: