Running tests on the WF-139. Interesting results. A must read for WF-139 users.

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
I have taken my "conclusions" from later in this thread and decided to edited them right into the top of the thread for all to see more clearly:

From Post #60 in this thread:
  • All WF-139 chargers, regardless of what "version" it is. Use a Constant Current only charging algorithm that can reach *charging* voltages of ~4.25-4.35+V during charging depending on the capacity of the cell being charged.
  • Charging a small unprotected cell in with this type of charger could cause charging voltages of 4.4V or higher to be reached in theory.
  • Most would agree that this charging method does not follow most li-ion manufactures recommendations for charging, and will probably reduce the life of the cell.
  • Charging a small unprotected cell (like a RCR123) in this charger, could come with increased risk of fire/explosion.
  • The Older WF-139 chargers, will measure ~10-12V open circuit, and will continue to trickle charge cells after the charge has been terminated. This could lead to fire/explosion if left unattended for long periods of time.
  • If the charge is terminated by the cells protection, which is very common when protected RCR123 or similar size cells are charged in this charger, the cell will not be trickle charged because the cells protection will hold the circuit closed after it has tripped.
  • Newer WF-139 chargers, will measure ~5V open circuit, and will not trickle charge after termination.

-----------------------------------------------------------------

Video is of a pair of 3.7V protected AW brand RCR123s on a WF-139 charger.

When voltage got to around 4.30V I had to grab the camera.

http://www.youtube.com/watch?v=w05TzgYHayY <watch this video.

This amazed me.

Ever wonder what's going on behind the scene with your WF-139?

I decided to run some tests to explain some of the weird behavior of this charger.

MANY people have reported that when charging AW RCR123s in this charger, they "terminate" at around 4.16V. I have speculated that the charge was being terminated by the cell at a voltage much higher than this, and then settling down to 4.16V without a supply voltage, I was correct. This video not only proves that (notice the voltage jumps to the open circuit voltage of the WF-139 when the charge "terminates"), but also that during the charge, the WF-139 is indeed holding the cell in a state of over-charge.

Here's a picture of the cell moments after charge was terminated:
100_3216.jpg


THE WF-139 DOES NOT APPEAR TO HAVE A CV STAGE AT ALL.

The WF-139 is holding a CC stage until the cell is somewhere in the 4.3X volt range, and then, terminating with the expected result of the cell settling to around 4.20V. This charge method is not considered terribly bad, but it's not terribly good either.

If you notice in the video, the voltage reading fluctuates, it "dips" down during the charge every few seconds. I believe the charger takes a "reading" of the cell voltage with the charge stopped every few seconds to see how much farther it needs to go to get to around 4.20V open circuit. My multi-meter is probably not showing how far down the cell voltage is compared with the charge voltage on these little "blips" because it's not responsive enough to catch it, so to speak. It's probably "RMSing" the value a bit so we're not seeing the whole picture.

On a smaller cell, the voltage is ramping up a lot faster than on larger cells, all the way to the point of tripping the PCB before it even makes it to 4.20V open circuit. This is why the smaller protected cells are not coming off the charger at 4.20+V like larger cells often do with this charger, the PCB is not only terminating the charge, it is also holding the circuit open until the input voltage is removed, this prevents the WF-139 from being able to do it's famous over-charge via trickle method.

------------------

More testing Observations:

On a pair of AW brand protected 17500s:
Charge voltage was at 4.31V when the charger terminated the charge and voltage dropped to 4.22V still on the charger. ~20 minutes later this had settled to 4.21V. I'm going to leave this on the charger with the volt-meter in place for another hour or more and try to capture the voltage rising as I expect it to from trickle charge.

-----------------

More test results and pictures to follow.

Eric
 
Last edited:

lctorana

Flashlight Enthusiast
Joined
Jun 28, 2007
Messages
2,123
Location
Melbourne, Australia
Wow.

So the UF-139 will charge the cells quite nicely, but looking at SilverFox's ready reckoner, it will shorten the cell life quite drastically whether we leave the cells on the charger after the light turns green or not.

Lesson #2
Small unprotected cells really, really, would not like the UF-139.
 

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
The ~11V reading is fine, it's the open circuit voltage of the charger and the cell is NOT seeing that voltage since the PCB has severed the circuit, it's nothing to be concerned about.
 

PhantomPhoton

Flashlight Enthusiast
Joined
Jan 15, 2007
Messages
3,116
Location
NV
Thanks, I haven't even thought to poke around like that yet to monitor what was going on there. Doesn't surprise me that it doesn't use a CV being so cheap.

I've been using the 139 charger for a year and a half now, but only light usage on my LiIon cells. The only problem I've had with a cell is a relatively new AW RCR123 not wanting to charge anymore. I suspect it was a cell problem and not a charger problem though as it's brother still works fine.
 

Black Rose

Flashlight Enthusiast
Joined
Mar 8, 2008
Messages
4,626
Location
Ottawa, ON, Canada
The ~11V reading is fine, it's the open circuit voltage of the charger and the cell is NOT seeing that voltage since the PCB has severed the circuit, it's nothing to be concerned about.
If the charger is rated to have an output of 4.2V (450 mA), why does it jump up to 10 - 11 volts right after termination?

I watched the video again, and noticed that the open circuit voltage stays high from around 30 seconds until the end of the video.
I would have expected it to drop once the battery indicated "no more".
 
Last edited:

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
The 11V reading is what you would get with no cell in place but with the charger plugged in. When the circuit has no load, it happens to have an output voltage around 11V. Once the PCB of the cell tripped, the voltage of the charger under no load becomes present.

Like when a CR123 is fresh from a package, it reads 3.2V, but when it is under the load of a P91, it's ~2.2V. Same type of thing.

Often times li-ion chargers for consumer style "loose" protected cells are designed to have this high open circuit voltage to re-set PCBs that have tripped off due to over-discharge. The 11V of the WF-139 may be a bit excessive, but most PCBs I'm aware of in good quality cells can handle up to around 12V without a problem.
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
I might perhaps take a less favorable view and suggest that the high voltage is typical of a constant current circuit with a high resistance load. In such cases, the voltage will naturally rise to the maximum.

It seems to me it would be much better if the voltage didn't rise above 4.2 V under any circumstances.
 

zipplet

Flashlight Enthusiast
Joined
Dec 11, 2006
Messages
1,139
Location
Ireland
This is awful - I thought the nano charger was bad (4.26V open circuit, so it has the potential to charge that high with trickle charging) but this WF thing looks like a piece of junk after seeing these tests.
 

McLux

Newly Enlightened
Joined
Apr 8, 2007
Messages
36
Location
Belgium
I am using a DIY li-ion charger for a few years now, with a voltage regulator, adjusted at 4.20 volts, and a current limiter @ approx. 1 A. The open circuit voltage is always 4.20 volt, I am wondering why the manufacturers don't make their devices that way.
My charger did't stop automaticly (I use a timer for safety), but after 4 hours, the current is only a few mA anymore, so the battery stops charging by itself.
 
Last edited:

zipplet

Flashlight Enthusiast
Joined
Dec 11, 2006
Messages
1,139
Location
Ireland
I wonder the same thing as I'm currently designing my own charger. Infact, in my post here I say the exact same thing as you. It costs nothing to do, but they don't do it.
 

WildChild

Flashlight Enthusiast
Joined
Apr 26, 2005
Messages
1,424
Location
Québec, Canada
This is why I trashed mine when I saw that! Mine was going to around 4.29v at the end of the charge and the two charge slot gave different readings. They were not charging the batteries at the same level at all. No problem since I switched to the Pila IBC. And at the end of the charge process, I never read over 4.20V on the terminals.
 

harddrive

Enlightened
Joined
Dec 21, 2006
Messages
392
Location
Australia
Makes you feel guilty about using the WF139. I also use the WF-137 for 18650 and suspect it is no better.

I think I might have to get a Pila IBC but I can't find one for less than about $70 USD delivered to Australia. :(
 

zipplet

Flashlight Enthusiast
Joined
Dec 11, 2006
Messages
1,139
Location
Ireland
This entire business is why im building my own charger. It's actually quite simple to do and easier than building a -dV nicd/nimh charger but I wouldn't recommend it to most people because the potential for something to go wrong is high.

Another member here is doing the same but I can't remember what his thread was called so I can't find it...
 

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
Good thread, and an excellent idea to do a video of it.

This is the main reason I stopped using the WF-139 some while ago - I got suspicious of the high voltage of cells right off the charger. They soon settled back down again, but still >4.20V, too high.

To rely on the cell's over-charge protection for termination is asking for trouble. It is not safe practice, and the fact that there is not a proper CV phase is not good for the cell.
 
Last edited:

Tohuwabohu

Newly Enlightened
Joined
May 8, 2008
Messages
184
Location
Germany
There is quite a lot of information on several cheap chargers including the Ultrafire WF-139 on this japanese page.
I can't read japanese and the Google translation is not very helpful but there are a lot of interesting graphs and circuit details on that page.

I don't have the WF-139 but I have the RCR123 version of the Ulrafire WF-138 which uses a very simalar curcuit.
With most of my protected RCR123's the charger will terminate the charge, but with some the protection circuit of the cell will kick in. If anybody is interested I could supply some scope graphs of the charging process.
I have ordered the Pila IBC and thought of comparing it to the WF-138 when it arrives.

The absenece of the CV stage in the charging process may not be good for the cell but it also has an advantage: the total charging time is shorter.

The HXY-042V2000A mentioned on the japanes page and also recommended in the german Messerforum.net could be a cheap and good alternative to the WF-139.
But, as often with this cheap stuff, you can't be sure what you get. I ordered the single cell version from KD an got an identical looking charger from another manufacturer with a lot worse build quality and lower charging current. That was the final reason for me to order the Pila IBC.
 
Last edited:

geek4christ

Enlightened
Joined
Feb 16, 2008
Messages
339
Location
Oklahoma, USA
Excellent detective work, MD. Thanks for doing these tests. I hope AW will consider offering a better charger in the future. A cheap charger at the expense of greatly reduced cycle life is just not a good trade-off IMHO.

One alternative (other than the Pila IBC) that I don't see mentioned very often is an affordable hobby charger like the Great Planes PolyCharge or the Apache 2020 which retail for around $23 and $30 respectively. All you need is one of those chargers, a cheap 12v power source (an old PC power supply can be used for this with a little modification) and some way to get the charging leads connected to the cells. I don't have either of those chargers, but they seem like very workable solutions.

Drawbacks that I see to going with a low cost pack charger:
  • Added cost of power supply or effort in building one from a PC power supply
  • Can't charge two cells independently
  • Must be careful to not set the charger to 2s or 3s if charging a single li-ion cell (although the PolyCharge has a safety mechanism that won't allow you to do that)
  • Cost is getting close to the Pila IBC anyway when you include shipping
  • Have to make a clamp or wire up a bay or use magnets to contact the loose li-ion cells
Benefits:
  • Selectable charge rate, so you can safely charge a wider variety of cells (10440 all the way to 18650)
  • Proven reliability of the products
  • Can run off 12v car battery when away from home without an extra adapter of any sort
  • Reputable dealers offer good warranties
  • Proper CC/CV algorithm and termination voltage unlike the UF-139
So that's my perception at the current time. Am I missing anything?
 

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
For comparisons sake, i just did some testing on another popular budget charger known as the DSD. I used a 5.7V 800mA Nokia cell phone adapter (12V car style) as the power source to get a feel for variations in behavior compared to the original 450mA adapter. Using the different adapter has actually helped reveal some things.

I see no evidence of a CV stage in this charger either, and I think this charger ordinarily just takes advantage of it's normally very slow charging speed to top off cell without the need for a separate CV stage.

I measured charging current right at ~810mA into a single 17670. In several tests the Charge always climbed steadily to around ~4.21-4.22V and then the charge terminated. The cells always settled to ~4.08-4.12V after termination. My gut tells me that with the charge current split between 2 cells, and using the original wall-wart adapter at 450mA... The termination at around 4.22V would semi-consistently result in larger size cells settling to ~4.18-4.20V, which is what I recall from years ago using this charger on a regular basis. (I have misplaced the original 450mA adapter, so I'm basing comparisons from memory of using this charger years ago).

So the DSD, is another example of a cheap charger with no apparent CV stage, that simply relies on characteristics of the cell to result in a charge that is semi-"acceptable."

With slow charge rates, a CV stage is less important though, consider a situation where you have 2 18650 cells loaded into a charger that's splitting ~450mA between the 2. That's about a 0.1C charge rate, which is what a CV stage should terminate at anyways, this charger avoids the need for a CV stage by simply charging so slow that it's semi-irrelevant. On smaller cells, I would expect the charge termination to result in voltage settling even lower.

Eric
 
Top