Review of / Measurement on UltraFire WF-188 charger

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
[SIZE=+3]Charger UltraFire WF-188[/SIZE]

DSC_6436.jpg


This charger is a multi LiIon chemistry charger that supports a wide variety of LiIon sizes.

DSC_6439.jpg
DSC_6442.jpg


The charger is powered from the mains and is universal voltage (100-240 VAC 50/60 Hz), the version I bought was supplied with an American mains plug.

DSC_6438.jpg
DSC_6440.jpg
DSC_6437.jpg


There are switches to select chemistry (3.7 or 3.2 volt) and charger current (300 or 650 mA) and a led for each channel. The switches have a click and enough movement to distinctly mark position. The charger has a led for each channel, it will be red when charging, green otherwise.

supportedBatteryTypes.png


supportedBatterySizes.png

DSC_6443.jpg
DSC_6444.jpg
DSC_6445.jpg
DSC_6446.jpg
DSC_6447.jpg
DSC_6448.jpg


The charger can easily handle 70 mm long batteries and up to D cells in diameter (32 mm), with both flat and button tops cells. But the slider that is connection to the battery minus pole does not slide easy enough, it will easily lock up and give a bad connection. Using a tool to secure it is pushing against the battery is a very good idea.




[SIZE=+2]Measurements[/SIZE]

Below 2.4 volt the charger is charging with 15-22 uA.
Between 2.4 volt and 4.2 volt the charger is applying regular charge current (See curve below).
When the battery goes above 4.2 volt the current is turned off (below 3uA current).
The charger will not restart charging before the battery has been removed.
When charger is disconnected from power, but with a battery in, it will draw below 300uA from the battery.
More batteries will not change the charge current.
Because each channel is independent, the current varies between the channels.

UltraFire%20WF188%20650mA%20%28AW18650-22%29.png


The 650 mA setting is a bit below 600 mA and drops when the battery is near full charge. This is not a CC/CV charge profile.

UltraFire%20WF188%20650mA%20%28AW18650-22%29%20B.png


The second channel has 600 mA in charge current and starts reducing the current earlier. The total charger time is nearly the same (3 hour, 24 minutes)

UltraFire%20WF188%20300mA%20%28AW18650-22%29.png


Reducing the charge current to 300 mA increases the charge time to 5 hour 50 minutes. The charger keeps the current constant during the charge, but the final cell voltage is lower than above.

UltraFire%20WF188%20650mA%20%28AW16340-IMR%29.png


My old IMR cell has a large voltage drop when current is turned off (This is as expected).

UltraFire%20WF188%20300mA%20%28AW16340-IMR%29.png


Again the 300 mA setting has a lower final voltage than the 650 mA setting (This is strange).

UltraFire%20WF188%20650mA%20%2818650%20LiFePO4%29.png


Charging a LiFePO4 on the 650mA setting works fine. The fast rise in voltage when the cell is nearly full is due to the LiFePO4 chemistry.

MeasurePuls.png


All the noise on the curves above is because the charger turns off the current each 1.5 second to check the voltage. This makes it possible to measure the actual cell voltage, ignoring any resistance both in the battery and in the connections.

MeasurePulsNoise.png


But there might be another reason for this pulsing, the charging circuit does make a lot of electric noise. This curve is exactly the same as above, but without applying any filtering.

Noise.png


Zooming in on the noise (Expanding the time scale 100000 times), show the actual noise pulses that makes the noise. These curves are for 300 mA charge current, 600 mA does look the same way.



[SIZE=+2]Conclusion[/SIZE]

The charger does not follow the recommended charge procedure for LiIon batteries, but it does charge to 4.2 volt and then turns off. It is dual channel and can do many battery sizes, the charge time is reasonable for 18650 batteries.
I believe that it is a useable charger, but watch out for the slider connection.



[SIZE=+3]Notes[/SIZE]

Here is an explanation on how I did the above charge curves: How do I test a charger
 

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
I have one of these. It works as a CC , inside there is a current source PWM based on a SD4840 chip. A microcontroller interrupts the CC to measure the battery voltage and open the charge circuit when close to 4.2V.

The voltage across the battery when the CC is applied may trigger the protection on protected battery. It happens to me.
Not detrimental. True, one measure about 5.2V but the battery element doesn't see it.

Unprotected terminates properly too, not exceeding 4.2V. What about old one with a higher Ri ? I don't have any so far. Anyway, one may assume that these dying batteries will be euthanized ?

Interesting enough but this circuit is more complex than a normal chip based CC/CV ?

May be the intent was to shorten the charge time by trying to push as much current as it can ?

The bulk of the charge is done during the CC phase in a regular CC/CV and the CV stage lasts about the same time as the CC.
 
Last edited:

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
It looks like for 650mA charge current, the float voltage equals or exceeds 4.25V. That doesn't look good.
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
It looks like for 650mA charge current, the float voltage equals or exceeds 4.25V. That doesn't look good.

The actual charge voltage is above 4.2 volt, but the chemistry is not charged to more than 4.2 volt. This can be seen at the point where the charge current is turned off, the voltage immediately drops below 4.2 volt. This is the same charging algorithm as WF-139.
 

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
The actual charge voltage is above 4.2 volt, but the chemistry is not charged to more than 4.2 volt. This can be seen at the point where the charge current is turned off, the voltage immediately drops below 4.2 volt. This is the same charging algorithm as WF-139.

Keep in mind the voltage drop against contacts and the protection circuitry. It is why this charger interrupts the current to measure the tension.

To get accurate measurements, one should use separate contacts across the battery element terminals (Kelvin method).
 

VidPro

Flashlight Enthusiast
Joined
Apr 7, 2004
Messages
5,441
Location
Lost In Space
This is not a CC/CV charge profile
like Really , beautfull pictures,
Is it really finishing a li battery charge without having slowed down but to only about 1/2 its charge rate?
the pics are wonderfull, but i cannot believe my eyes , cause that sucks so badly :)

did ultrafire ever get the MEMO :devil:, slow down neer the end of charge, , , at LEAST, even if your not going to actually do CV :shrug:
 
Last edited:

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
The actual charge voltage is above 4.2 volt, but the chemistry is not charged to more than 4.2 volt. This can be seen at the point where the charge current is turned off, the voltage immediately drops below 4.2 volt. This is the same charging algorithm as WF-139.

Of course the "chemistry" is being charged at over 4.2V. What else could be getting charged? The wrapper? The saving grace for this charger is that it terminates just about immediately upon reaching the float voltage. Thus, the "chemistry" basically spends zero time at 4.2+V.
 

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
Of course the "chemistry" is being charged at over 4.2V. What else could be getting charged? The wrapper? The saving grace for this charger is that it terminates just about immediately upon reaching the float voltage. Thus, the "chemistry" basically spends zero time at 4.2+V.

How can you be so sure ? Did you measure the voltage across the "chemistry" ? Once again, one has to account for the voltage drop across the contact R and protection circuit !

Anyway, no need to worry about. For most users, their batteries will die of old age (3 years average) instead of this charger. This happens with portable PC, Mobile phone, camera ... . The vast majority ends up in the recycle bin because of dead batteries too expensive to replace when available.
Here at least 18650 are cheap and plentifull.

And if you want a cheap and real CC/CV charger, see my thread about the use of Samsung TCH137 charger.
 

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
How can you be so sure ? Did you measure the voltage across the "chemistry" ? Once again, one has to account for the voltage drop across the contact R and protection circuit !

If there is voltage drop in the measurement, then the true float voltage will actually be even higher. Sure, there can be all sorts of sources for measurement error. If you choose to not believe the data presented, then what's point of making any conclusions one way or another? You may as well just assume whatever the manufacturer tells you.
 
Last edited:

Rej

Newly Enlightened
Joined
Mar 10, 2011
Messages
58
Location
Stoney Creek, Ontario Canada
I've had one of these for about a month now... and find as a 'backup' charger to my Hobbico Elite & 'gutted' UF-139 (charging cradle now) it is not bad. I gutted the UF-139 as in the past it had gone as high as 4.68volts during charging!!

I did hardwire the internal sliding mechanism with 16G wire...so as not to rely on the sliding contact (another CPF member suggested this), which seemed 'ify' at best.

My unit never goes above 4.24-4.25 volts during charging, but I have noticed at the 600ma setting; I rarely get higher than 4.15 resting voltage on my AW 18650-2600's. If I use the 300ma mode I can usually hit around 4.17-4.18 volts, but usually only after several attempts.

Definitely stops the charge when done; (forgive the crappy cell phone pictures!)

1st shot resting after a full charge; (note, My small voltmeter reads 0.05V TOO high, actual open voltage is 5.43-5.44V);
03142011241.jpg


(Open voltage = 5.44 | Mah = -0.005)

2nd shot is near the end of charging at 300ma mode;

03142011239.jpg


(actual voltage = 4.21 / mah = 293)

It does not do a full charge, as my Hobby charger setup... but in a pinch it is definitely better than the UF-139.
 

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
If there is voltage drop in the measurement, then the true float voltage will actually be even higher. Sure, there can be all sorts of sources for measurement error. If you choose to not believe the data presented, then what's point of making any conclusions one way or another? You may as well just assume whatever the manufacturer tells you.

I believe the data presented and I pointed out that the real voltage across the battery element is the voltage shown in them minus the drop in contact and protection circuit !

And indeed, when the current is turned off, we can see that the voltage measured drops and it corresponds to the battery open voltage (no current, no drop, the MOSFET can be considered pure resistor).

The measurment set up should be such that the voltage be measured across the battery element, this requires removing the wraping and using contact point independent of the charger contact (we can neglect the contact internal to the battery).
 

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
No, the cell simply is not charged to 100% and thus the OCV is less than 4.2V which is viewed as 100% SOC. It's not voltage sag from contact and protection circuit resistance. You'd see the same behavior for true CC/CV charge if the charge were terminated as soon as the float voltage of 4.2V was reached. A Li-ion, protected or non-protected, isn't fully charged at that point and both types of cells would show a drop in OCV.
 
Last edited:

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
No, the cell simply is not charged to 100% and thus the OCV is less than 4.2V which is viewed as 100% SOC. It's not voltage sag from contact and protection circuit resistance. You'd see the same behavior for true CC/CV charge if the charge were terminated as soon as the float voltage of 4.2V was reached. A Li-ion, protected or non-protected, isn't fully charged at that point and both types of cells would show a drop in OCV.

Are you answering to my post ? If yes, sorry but we are not talking about the same things.

In my post I am refering to the voltages displayed in red in HKJ post. These are the voltage applied to the element under charge (when I speak about element, I speak about the real element inside the package !) plus the voltage drops.

A battery can be seen as a perfect voltage source (which varies with the charge level) with various resitances in series.

And a protected battery will be less charged than an unprotected one all thing considered equals. Although we are speaking a few if not fractions of % here.
 

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
The IMR16340 that was tested above has no protection circuit. Hard to detect any SOC difference vs the protected cells also tested. Granted, the protected cells were not 16340 size and they were also different chemistry.

The voltage drops when the charger stops charging because the cells are not at 100% SOC, which is commonly assumed to be at 4.20V OCV. Has nothing to do with voltage drop from any measurement arising from contact resistance or protection circuit.

If the charger implemented a true CC/CV algorithm and used a float voltage of 4.20V, then if the cell voltage after a full charge measures say 4.15V, then I agree that this is probably because of some parasitic resistance that fooled the charger into thinking that the cell was at 100% SOC. But this charger immediately terminates charging as soon as its designed-in float voltage is reached. So it is hard to say what contribution parasitic resistance plays in the voltage drop. Based on the IMR16340 results, I would say minimal, especially if all of the tested cells are healthy.

Quite clearly, the cells are not fully charged and that definitely plays a key role in the lower voltage after charge termination. Assuming that the cells were fully discharged prior to the charging tests, the blue lines in the graphs above clearly show that the mAh capacity is not close to 100%.

And batteries have more than just resistance, which actually should be impedance. There is capacitance and inductance as well in most any model of a battery equivalent circuit.
 
Last edited:

Rej

Newly Enlightened
Joined
Mar 10, 2011
Messages
58
Location
Stoney Creek, Ontario Canada
I'd have to agree Justin.... as my hobby charger will hit 4.200V at say 600mah, then continue to charge (@ 4.20V) for another hour or more till it hits around 40-50mah. My protected AW 18650-2600's ALWAYS come off around 4.19-4.205V, and are still there resting after 30 minutes.

I've never hit higher than 4.17 V resting (in either bay) with the SAME cells on my UF-188; obviously it's terminating the charge early. And to reach that voltage I usually have to switch over to the 300mah mode and charge several more times! And this is with a direct-wired negative slider....bypassing the 'poorly designed' negative contact IMO.

Might be a good thing for cell Longevity....but the method of charge is questionable, as HKJ has shown :shrug:
 

VidPro

Flashlight Enthusiast
Joined
Apr 7, 2004
Messages
5,441
Location
Lost In Space
I'd have to agree Justin.... as my hobby charger will hit 4.200V at say 600mah, then continue to charge (@ 4.20V) for another hour or more till it hits around 40-50mah. My protected AW 18650-2600's ALWAYS come off around 4.19-4.205V, and are still there resting after 30 minutes.

aka , uses a CV style , and terminates when the current flow is low enough that it is basically done.


I've never hit higher than 4.17 V resting (in either bay) with the SAME cells on my UF-188; obviously it's terminating the charge early. And to reach that voltage I usually have to switch over to the 300mah mode and charge several more times! And this is with a direct-wired negative slider....bypassing the 'poorly designed' negative contact IMO.

Might be a good thing for cell Longevity....but the method of charge is questionable, as HKJ has shown :shrug:

the problem i see, is it is not doing the CV , or even acting to the battery like CV, where the input to the battery is slowed down as the number of chemicals that CAN charge are limited.

If it terminates early, then that is going to help some, IF you keep stuffing a cell back on this thing, or changing the rate and putting it back on, your STILL not slowing down at the end of charge like your hobby charger is doing, and the "proper" methods.

To me it is all about the chemicals that are "mostly all charged", and few chemicals capable of being charged Left to charge (needing time), and this thing trying to pound that in when it just wont happen with the average high capacity low current type of li-ion, without time. ( probably ok for robust high speed batts)

i am saying IMO you should not worry about that last ~.05V or put it back on to try and squeese it up to a higher voltage, that would make what it is already doing incorrectally worse.

when the battery is mostly charged and your trying to get that last bit charged, any power that does not actually charge the limited stuff that CAN charge, will just be WASTEd, power going the "wrong places", or even causing a bit of gas in places.

someone help me say it correctally :)
 
Last edited:

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
The IMR16340 that was tested above has no protection circuit. Hard to detect any SOC difference vs the protected cells also tested. Granted, the protected cells were not 16340 size and they were also different chemistry.

The voltage drops when the charger stops charging because the cells are not at 100% SOC, which is commonly assumed to be at 4.20V OCV. Has nothing to do with voltage drop from any measurement arising from contact resistance or protection circuit.

If the charger implemented a true CC/CV algorithm and used a float voltage of 4.20V, then if the cell voltage after a full charge measures say 4.15V, then I agree that this is probably because of some parasitic resistance that fooled the charger into thinking that the cell was at 100% SOC. But this charger immediately terminates charging as soon as its designed-in float voltage is reached. So it is hard to say what contribution parasitic resistance plays in the voltage drop. Based on the IMR16340 results, I would say minimal, especially if all of the tested cells are healthy.

Quite clearly, the cells are not fully charged and that definitely plays a key role in the lower voltage after charge termination. Assuming that the cells were fully discharged prior to the charging tests, the blue lines in the graphs above clearly show that the mAh capacity is not close to 100%.

And batteries have more than just resistance, which actually should be impedance. There is capacitance and inductance as well in most any model of a battery equivalent circuit.

Capacitance and inductance have no sensible effect here.

On this charger, the measured voltage drops (small dips on the red curve) while charging correspond to the Ri x Ic. The charging current is periodically interrupted in order for the microcontroller to measure the battery voltage and decide to stop the charging.

We all know that this is not the prescribed CC/CV method and as I said, I assume the designer intended to reduce the charging time by pusshing more current then would be possible with a CV of 4.2V.


With a CV of 4.20V the charging current goes down in an asymptotic way. It will take forever for the battery to reach 4.2V. It is why in a CC/CV, they stop when the current is about 1/10 the initial charging current corresponding to the CC phase.

The voltage sag seen once the battery is removed is mainly due to the chemical process. It takes indeed seconds, so nothing to do with internal resistance, L or C.

If I put the batteries charged with my WF188 in the charger I made with a Samsung which is a true CC/CV, this one starts to charge then stops after about 5 min.

Anyway, we are dealing here with a few percent of undercharge. Nothing important and acceptable considering the various elements and measurments tolerances.

As long as the charger doesn't subject the battery element to more than 4.2 V for a long period of time, doesn't trickle charge and ends up with a battery slightly undercharged no problemo.

One has to adapt the charging current to the battery size to avoid this.

Older batteries with higher Ri may be subjected to slightly more than 4.2V, OK, so what ? They may see their life shortened but they are close to their end anyway.

Once again, this WF188 is more complex than a regular CC/CV ! So the guy who has designed it, has attempted do do something he thought was better.

Did he succeeded ? In fact I don't care, it is a serviceable charger for regular use at a rock bottom price.

True, the sliding contact system is lame. I put direct wires with proper strain relief.

Did anyone of you reversed engineered this charger WF 188 and a true CC/CV ?.
 
Last edited:

Justin Case

Flashlight Enthusiast
Joined
Mar 19, 2008
Messages
3,797
Here is a CC-CV charge curve for a Panasonic 18650 from the datasheet. Charge rate is 0.67C. Float voltage is hit after about 70 min, for which the cell is at about 78% capacity. If the charge had been terminated at that point instead of continuing with the full CV phase of the charge, quite clearly the OCV would not stay at the float charge level of 4.2V.

Panasonic18650CC-CVchargecharacteri.gif
 

BoarHunter

Newly Enlightened
Joined
Feb 7, 2009
Messages
150
Here is a CC-CV charge curve for a Panasonic 18650 from the datasheet. Charge rate is 0.67C. Float voltage is hit after about 70 min, for which the cell is at about 78% capacity. If the charge had been terminated at that point instead of continuing with the full CV phase of the charge, quite clearly the OCV would not stay at the float charge level of 4.2V.

Panasonic18650CC-CVchargecharacteri.gif

It is obvious that if the charging is interrupted just at that point (CC to CV transition), the full charge is not reached and the open circuit voltage of the battery will NOT be 4.2 V. The voltage you see on the curve is the battery element voltage PLUS Ic x Ri !!!

And it is why the WF188 cuts the charging current in order to measure this Voc and decides to proceed till it is say about 4.15 V.
Without that the WF188 would wrongly assume the battery is charged.

Again, if you look at the internals of this Wf 188, you will see a PWM CC acting at the same time as main AC DC converter, MOS switches and a microcontroller with AD converter to orchestrate all this.
A more complex design than the regular CC/CV linear circuits.
The WF188 drops its charging current at a lower pace, thus reducing the charging time.
I doubt the engineer who designed this charger was unaware of CC/CV circuits and I am pretty sure he wanted to do something he felt was better.
Is the gain worthwhile ? Is it detrimental to aging batteries ? This requires a more thoroughfull analysis.
 
Last edited:

Rej

Newly Enlightened
Joined
Mar 10, 2011
Messages
58
Location
Stoney Creek, Ontario Canada
i am saying IMO you should not worry about that last ~.05V or put it back on to try and squeese it up to a higher voltage, that would make what it is already doing incorrectally worse.

when the battery is mostly charged and your trying to get that last bit charged, any power that does not actually charge the limited stuff that CAN charge, will just be WASTEd, power going the "wrong places", or even causing a bit of gas in places.

Thanks...I'll take that advice, as the charger cost less than one AW battery!
 
Top