MAHA C204W charge cycle

[wptski]

I was asked to if possible post some scope screens from a C204W charger cycle. These were done using two AA cells, so the maximum current is 2A but again this is another PWM cycled charger! It appeared to be the best way to show the cycle was to plot the duty cycle, so all five screens are showing duty cycle. Whatever the percentage it is that part of 2A. The first four are zoomed in traveling from left to right each one overlaps a bit. The last screen is after the charge LED turned green and it was in trickle mode which lasted around two hours(8% or 160ma) and although I don't have a screen of it, the current dropped to around 130ma, still PWM. It slowly alternated between 60ma and 130ma.

For a charger that claims high initial currents to shock or jolt a old cell it starts off rather wimpy at around an effective .5A and stays at 1A for some time!!

I remember testing this charger for cell temperature and it got over 140F but most of that heat must be generated by the built in power supply, etc. The cells ran much cooler because the cells were extended outside the charger for this test.

[bcwang]

Hi Bill,

Woah, this charger is all over the place. Based on the time scale, it looks like it slowly moves overall charge current up and down gradually, but I wonder why. One thing that doesn't make sense is if the charger is supposed to average 2amp charge current, it doesn't seem to spend much time at 2 amps.

Do you know what pulse rate is used for the pwm?

Did you happen to get the current during the final trickle charge stage?

What temperatures did the cells reach outside the charger? Maybe a little fan blowing on the charger will keep things cool since it's not the batteries overheating from the inside, but probably heat transfer from the power supply to the batteries.

Thanks,
Ben

[wptski]

bcwang:

The pulse rate varies so much, it's hard to tell! The trickle is 160ma for about two hours and it changes between 130ma/60ma.

The temperature was around 120F max. A fan would help but 120F isn't that high, although I had a fan on the cells at first.

[SilverFox]

Hello Bill,

Very interesting...

Tom

[bcwang]

bcwang:

The pulse rate varies so much, it's hard to tell! The trickle is 160ma for about two hours and it changes between 130ma/60ma.

The temperature was around 120F max. A fan would help but 120F isn't that high, although I had a fan on the cells at first.

At least from what you could tell, was the pulse rate in the range of seconds per pulse, or milleseconds? I know when charging you can hear a pulse sound every interval which feels to be over a second. However, I never noticed it changing intervals, maybe it was too gradual to notice over a few seconds of listening. Maybe if you spend enough time listening to the pulse, you can hear the rate change like it does in your graph.

Strange final trickle charge, do you know how much time it spends at 130ma vs 60ma? 130ma seems too high for a trickle, even 60ma seems to high for some of the lower capacity cells.

I actually meant using a fan when using the charger as intended with the batteries inside the charger and not hotwired to the outside. Maybe the flow of air will cool the electronics so the batteries don't absorb all the heat.

You know, it's still odd that your graph shows very little time at 2 amps. I know this charger does average a 2amp charge rate with two cells, if we use your graph for comparison, there is no way it averages 2 amps over the duration of the charge. Something must be off here. You didn't happen to play with the other bank while the log was collecting data causing it to switch in and out of the 2amp mode did ya?

You should try this mode of data collecting on the C808M as well. I wonder if it'll exhibit this same varying charge current as well. Somehow it is able to keep cells very cool, might be related to not using constant charge current.

Ben

[wptski]

bcwang:

Actually, it's very slow, look at the timebase, 100s. I ran several tests with different cells and it was pretty much the same. Could it be fooling me and the scope? Youbetcha! I'll try some different ways to see if I get different results. It could very well be beyond my equipment too. I did run into a quirk where after the green LED comes on and I was trending the duty cycle in trickle. The screen goes from left to right which takes an hour, it then compresses to about a half a screen and continues to draw. At two hours, it just stops, I save the screen, it's blank and resets back to the start again. I might be too close to the bottom end of the duty cycle range, not sure though. The reason I used duty cycle because of the constant change, trying to get a picture in one shot instead of a bunch of screens. I mentioned this in our PM's that it might be difficult to capture.

I post screens from the C808M too but if I remember correctly, it's pulse rate was in seconds also but nowhere near as slow as the C204W.

Yes, it doesn't show much time at high duty cycle or 2A unless for some reason or somehow the net effective current is higher than what these graphs show but I kind-of doubt that! Like I said, I'll play more with it to see if I can find fault in what I've posted.

[bcwang]

Hi Bill,

Let me see if I can help you figure out the problem. Can you tell me the settings you used to do the capture? Things like sample rate, sample value (time averaged or point sample). I can think of a possible scenario of setting up the capture that would produce a graph like that when the actual charging rate is nearly 2amp the whole time.

Thanks,
Ben

[wptski]

bcwang:

The manual leads you to believe that the scope automatically set up everything but I'm "not" sure about this but it was set to 1A div and 200ms timebase and that's it!

I just ran another test but reading current instead of duty cycle and "somehow" lost my saved screen plus it reset again! This is really pissing me off! I was showing a bit different trickle too. Of course, the fricking battery died too. I may have to send this pack to SilverFox and have hime cycle it fifty times or so, he likes to do stuff like that. Not sure if it's defective or just having a slow break in time?

This charger may be a real ongoing project.

[wptski]

bcwang:

I may have found the problem! The div and timebase prior to entering TrendPlot does affect its operation. Timebase affects the starting timebase of the TrendPlot. I "think" that the problem was that the div should have been at 2A instead of 1A. I could tell the difference while watching it but I won't really know till I look at it closely tomorrow!

[bcwang]

[wptski]

bcwang:

Okay, here's four zoomed in screen at 2A div! There appears to be a few spots that show a skipped part of the trace for some reason, nothing is perfect. They are better but is that it? I'm not done playing and always willing to learn more.

[bcwang]

Hi Bill,

I wonder if the scope is not sampling often enough so your not getting consistent numbers. It might be giving you an instant value at each sample and miss a whole bunch of values in between. Is there a way to setup the scope to time average the current as the sample value, like what you did when testing the energizer 15 min charger?

Thanks,
Ben

[wptski]

bcwang:

The scope only has a 20mHz sampling rate. I was looking at the 40mHz version too but it listed about $200 higher and there weren't any deals around on that version. It also wasn't supported by another software package that I already had which in some ways is better than the Fluke software.


I have tried before and I'm trying that again and I've tried so many different ways, I'm getting a bit confused! There's a problem in what the scope captures, puts into memory and what/how the software reads/shows that data. I've recalled a screen from memory, download it into the Fluke software and I'll have a blank screen on my PC. That same data "will" show in the other software mentioned above but it's weird looking. This charger's cycle is unlike any other charger that I've looked at so far. The secret of how it works might remain a secret too!

I may have a handle on my screen lockups and data loss, so that's a big help.

[bcwang]

bcwang:

The scope only has a 20mHz sampling rate. I was looking at the 40mHz version too but it listed about $200 higher and there weren't any deals around on that version. It also wasn't supported by another software package that I already had which in some ways is better than the Fluke software.


I have tried before and I'm trying that again and I've tried so many different ways, I'm getting a bit confused! There's a problem in what the scope captures, puts into memory and what/how the software reads/shows that data. I've recalled a screen from memory, download it into the Fluke software and I'll have a blank screen on my PC. That same data "will" show in the other software mentioned above but it's weird looking. This charger's cycle is unlike any other charger that I've looked at so far. The secret of how it works might remain a secret too!

I may have a handle on my screen lockups and data loss, so that's a big help.

Maybe that's why I can't get any answers from Maha about this charger. It must be some super secret mysterious charge system that they don't want anyone to know about. It's so secret that even trying to use measuring equipment with it results in defeat!

By the way, have you noticed cells coming off the C204W 2 hours after the light turns green last longer than off your other chargers? I'm going to wait til I get a CBA II to make a more conclusive test. But it seems to be the best of all my chargers in this respect.

[wptski]

bcwang:

I've never compared AA cells off the C204W to other chargers.

Now here's a couple of screens. The first is the actual screen dispaly of the C204W in trickle mode which I left running. Notice how the screen gets compressed over time and the blurb is where the battery in the probe died but a have a second probe which I connected on the fly. The constant trickle was around a 100ma but as you can see it did go down to 50ma. It also pulsed to higher peaks which I caught at 645ma. The second capture is what the Fluke software makes of the same screen. Geez, I forgot to grab that screen with my other software!

In normal scope mode the average current will read at 100ma but the slowly pulsed peaks are much higher. The faster the timebase the higher the peaked current will be. At 1s the current may be 200ma but 500ms it would be 400ma and at 250ms it would be 800ma.

A while back I was working on my gas dryer. If it did start it drew 5A(normal) running but the problem was at times it didn't start and drew 35A! Using a Fluke DMM min/max function which read down to 100ms glitches, it read 45A. In Fast min/max at 250 micro seconds it read 60A. I ended up replacing the drive motor, not a fun job. Actually, one of the reasons I got this scope was to look at the old motor measuring phase shift between the start and run winding but never got to it yet!

[bcwang]

Maha specifies a 50ma trickle current. I'm surprised again to find you are averaging 100ma, with 50ma as the minimum, and seeing 645ma sometimes.

I thought you would see 2 amp current pulses at a very low duty cycle for the trickle averaging 50ma. This is very weird.

[wptski]

bcwang:

It is 2A duty cycled down. Here's one that shows 2A but not at 50ma, it's trial/error trying to catch that split-second moment! It is wierd and confusing.

[bcwang]

Does the scope have triggering so you can catch two pulses in a row and stop? Once you know the pulse width and the pulse frequency you can figure out the overall charge current. I notice your picture shows 310 ma but it might have calculated that current from whatever was in the graph, there might be more dead time before the next pulse that would have brought the average current farther down had the scope continued.

[bcwang]

Hi Bill,

So I did a little experiment. Using your last graph as an accurate charging pulse of what looks like 2 amps at 310 ms. I made an assumption that the charger always uses a pulse of that width and magnitude. So I took a stopwatch and tried to hand time the duration between pulses for each of the 4 charging stages. I would have liked to use a more sophisticated way to get an exact timing between pulses but all I had was a stopwatch, my ear, and my trigger finger. I did it quite a few times to see if I could get consistent results and when the same number came up many times I used it as my value.

My results
Stage 1 - 1.25 sec
Stage 2 - 0.31 sec
Stage 3 - 3.85 sec
Stage 4 - 12.85 sec

So using your data that each pulse is 310ms and 2 amps, it equals 620 ma current delivered per pulse. Using my measured pulse duration and the equation to give average delivered current over time of current_pulse/time_between pulses, the results are:

Stage 1 - 496 ma
Stage 2 - 2000 ma
Stage 3 - 161 ma
Stage 4 - 48 ma

This mostly matches the results from your first post, but I get much more consistency. I listened to the charger many times over the duration of the charge, the rates are the same the whole time during each stage and do not vary. This data makes much more sense now, as I know this charger does average a 2amp charge rate overall. I think the graphs from your first post may be from sampling with not enough resolution causing aliasing in your sample data which gives the quite regular slope. I'm sure there is a way for the scope to be setup to give the correct results though, just gotta figure out how to do it.

So the next question is, how does this charger behave when both banks are in action? My guess is for Stage 2 at least, that the charger alternates each pulse between the two banks so they each average 1amp. However, it's the other 3 stages that I don't know about and would like to find out. I'll try to listen for the charge rate again when both banks are in action to see what it does.

Could you please try your scope on the AAA section of the charger and try to get the same type of picture as that last one? I'd like to know the characteristics of the charge pulse on a AAA battery and using my ears I can try to figure out the charge rates in each stage.

Ben

[wptski]

bcwang:

Okay, here's a few more for you to look at! The first three are charging before the green LED comes on and the other three are during trickle. Some but not all have multiple pulses as you asked for.

The first stage as you refer to, I guess the charger is analyzing the cell and nothing is really happening on the scope. The first screen is where the charging starts at about .5A and then progresses to the second and third screen. I just changed the parameters there to show different values like frequency, etc. I can post the waveforms for any of the below. A quick check of a AAA cycle just show less current, around 800ma.

I've only tested one bank so far but I'd guess that they are seperate circuits, detect the other bank and reduce the current.

[bcwang]

Your graphs confirm some of my findings. The first graph shows the pulses 1.25 seconds apart, averaging 500ma charge rate in stage 1. The second graph is not as discrete but you do see mostly the drop in current every .3 seconds or so which is the charge rate in stage 2. The third graph makes it easier to see the same thing, a .3 sec charge every .31 seconds.

Your bottom 3 graphs are actually of the stage that maha calls "top-off", which also lines up with my findings. A pulse every 3.85 seconds or so. If you wait two hours after the light turns green, you should start seeing the pulses around 12.85 seconds apart.

So Bill, you said for AAA the current you saw is 800ma. Do you mean the graph now shows a amplitude of 800ma and the pulse width is still the same 300ms long? When I get a chance I'm going to try listening to the AAA charge cycle to see what pulse rate it uses in each stage of charging. Hopefully this info along with your pulse info will help figure out what currents it uses for AAA cells for all 4 stages.

[wptski]

bcwang:

Yes, the same 300ms or so but the amplitude of pulses aren't 2A as the AA cycle but a bit less than 1A on the graph, value reads around 800ma.

I'll try to work on graphing both bays but I may have to make up another cord, not sure.

EDIT: Wouldn't you like to tour the R&D department at Duracell, Sanyo, MAHA, Energizer, etc.? I wonder how many tests go up in smoke?

[bcwang]

Ok, so I finally got around to doing the AAA charge testing. I got the same cycle times as the AA results, so the only difference is the charge current. Based on that, and all the previous data, this is my summary on how I think this charger works. This is based on wptski's measurement of 800ma current.

AA 2 cell AA 4 cell AAA 2 cell AAA 4 cell
1 496ma 496ma 198ma 198ma
2 2000ma 1000ma 800ma 400ma
3 161ma 161ma 64ma 64ma
4 48ma 48ma 19ma 19ma

However, maha claims 700ma charge rate in AAA mode, so if you use that for the calculation you get:

AA 2 cell AA 4 cell AAA 2 cell AAA 4 cell
1 496ma 496ma 174ma 174ma
2 2000ma 1000ma 700ma 350ma
3 161ma 161ma 56ma 56ma
4 48ma 48ma 17ma 17ma

I guess it all makes sense if you consider AAA cells are about 1/3 the capacity, capable of 1/3 the current draw, and require charging at 1/3 the rate of AA cells.

I can't figure out how to add tabs to the text so everything in the table is offset. It seems to swallow added spaces too. It's still readable though.

[wptski]

bcwang:

Here's a couple more graphs for you. The first is charging four cells and just as you said, it's using alternating pulses.

The second is a full charging cycle of two AA cells. The time compression compressed the initial start of .5A right out of the graph but we've looked at that above somewhere. It does show the main charge of 2A till the green LED lites, after that it's 160ma some two hours before it dropped to around 20ma although it shows 40ma where I stopped it. So now you have all four portions of the charge cycle!

The first graph was done on my other C204W which cooked a cell one time. It cooked more than the cell, other graphs I did showed unstable lower current compared to my other unit!

[bcwang]

Hi wptski,

Do you happen to have that last graph with everything time compressed, except this time for a AAA battery?

[wptski]

bcwang:

No but I can do that in the next few days! The compression is something the scope does every time the trace reaches the screens right edge.

[wptski]

bcwang:

Here's the C204W charging AAA cells. I'm posting this first one so you can see the start of the cycle, after it compresses a few screens, you can't see the 200ma or so starting charge.

The second one is the whole cycle, stage three is around 60ma and trickle was as low as 6ma at times.

[bcwang]

Cool, looks like it about lines up with everything having the same timing as AA charging except using a 800ma current instead of 2000ma. I finally got my CBA II so I'll be testing completeness of charge for a couple of different chargers. I have a feeling this one will do well.

Thanks for the graphs wptski!

[bcwang]

Something we never investigated was the discharge rate of the c204w. I think for AA it is supposed to be 400ma, but this is not verified, and I have no idea what it is for AAA batteries. If anyone tested the discharge rate of the c204w that would be great to know.

[moldyoldy]

Regarding the 204W from website:

= = = = = from the Maha website = = = =
Improvements on an already intelligent charger

When the light turns green, the batteries will be nearly 98% charged so you can get the most of the high capacity batteries. The MH-C204W has a 5th generation PowerEx microprocessor that features an unique four stage charging process:

-Ramp-up
-Rapid Charging
-Top-off charging
-Maintenance Charging

The Ramp-up jolts the batteries if needed to revive old batteries. This jolting feature is a series of carefully timed high current pulses that can activate both new and inactive batteries.

The MH-C204W is designed with an intermediate, medium current, top-off charge so that the battery will retain most of its charge.

The Maintenance Charging feature is an extremely low current pulse charging that gives the batteries a brief pulse charge periodically to ensure the batteries are charged but not overcharged.

= = = = = = = end snip = = = = = = =

So my question is - how does it determine "if needed"? an "old" battery does not seem to be sufficient. It must be measuring impedance some how. Other charger designs claim do that as well (AccuPower 20 & 10). IOW, if you have a regularly cycled cell, then no "jolt" is needed at the beginning. Also, the end top-off charge is probably rather high based on the website entry. certainly not the "trickle-charge" that we think is common. no info given as to the invocation time difference between a "top-off" charge and a "maintenance" charge.

Tim