Pulse Charging NiMh Batteries?

I have been using a Titanium Innovations Charger (this one) for some time.

All seems OK, it has independent channels, refreshing function, and does not heat up batts since it takes quite long to charge them (3hours or so)

I also has a display that shows charging progress. For each channel there is a 4 segment bar and it blinks with a rate of 1 sec or so.

Sorry for being kinda long but I had to show how it works...

So just of the heck of it, I tried measuring charging current with my DMM.
What I saw is that the DMM readings pulsed from 0 to 1.69 following the rate the "charging bars" blinked.

Seems that the charger pulses charging current then rests to measure battery voltage and avoid heat buildup(?).

Now...if it is so, how good a charging algorithm is this?

TNX, kostas

Kostas:

Pulse charging is less an algorithm than a technique for charging Nickel-based cells. As you mentioned, it allows the charger to sample the battery voltage and allows the charger to terminate the charge based upon -dV/dt. In practice there is a slight (~.10v) voltage drop as a nickel-based battery reaches maximum charge. If the charger can detect this it can drop down to a trickle charger or terminate altogether. A termination algorithm is necessary for NiMH chargers that charge at rates higher than 0.1C. Some chargers also have a safety feature where they will terminate at some maxV if the deltaV is missed. This also requires pulse charging as far as I can tell. The spec. sheet for your charger indicates that it also uses 0V/dt which is merely the detection of no increase in voltage over some fixed amount of time.

Fancier chargers will add a thermocouple to measure dT/dt (temperature rise over time) as an alternate acceptable termination technique. Some really good chargers (like the MAHA C-9000) use a (secret) combination of all of these.

wow! thanks for the correction and the additional info on teh termination determining methods....

thread saved/archived!

:thinking: I have the Accumanager 20, and it's also a pulse charger. I currently am only charging the (D size Accuevolution) LSD cells. It takes about 15 hrs to complete the charging cycle when charging four of these 10,000maH cells. The D cells never get hot only warm, and the finishing end voltage is about 1.46V. It charges at 700ma at each position.

Pulse charging is used by most of the popular chargers for example:


The MH-C9000 charging at the default rate of 1A (a pulsed 2A charge current)



This is the default 2A charge rate of the MH-C808M charger, for the most part constant, with pauses to measure battery voltage. (some nice voltage spikes on this waveform)


Here is the MH-C808M "soft" charging at 1A (2A charge pulse at 50% duty cycle)



The Sanyo NC-MQN06U charger. In this case, it is pretty much constant with a pause every so often to measure battery voltage, I suspect.


This is after the Sanyo charger switches from bulk charging to top off.

As pertains to Nickel based rechargeable cells, the negative pulse, or "burp" charging method (which I'm pretty sure the Accumanager chargers use, as it supposedly negates the need for cycling cells etc. I don't know for sure about the Titanium Innovations Chargers, or others) has been debated for 40 years, or so. Some originally thought that it was the greatest thing since sliced bread. Others, who would seem to be more in the know, determined that while it doesn't really hurt anything, it doesn't actually do anything either.

If you search around, you can find articles supporting both sides of the debate. I think this article sums up pretty well, just how effective pulse, or burp charging actually is though.

As far as charging cells at, for example, a 10A pulsed rate 10% of the time vs. charging the same cells at a constant 1A rate, the accumulated charge will be equal and the cells will maintain the same temperature, provided all other factors are similar. Think about it. An effective 1A/h rate, is a 1A/h rate, no matter how you dole it out.

Note that not all chargers use negative pulse charging. Some simply use a pulsed charge. In either event, personally, I don't think there is any advantage (or disadvantage) to using either of these pulse charging methods for charging nickle based cells, as applies to single rate chargers. For example, there is really no need to check cell voltage between pulses with nickel based cells other than if voltage is the only means to determine the point of charge termination. And, as I said, there would be no significant difference in temperature of the cells, provided that the effective charge rate was the same as a constant current charger.

Dave

then you have to go on about current control, and how many of the ni-?? chargers didnt have any intention of actually using a pulse charge alogrythm, they just use it to set the RATE, which is an "averaged" rate, not the actual rate of charge. and that difference in rate can make a huge difference in what is really occuring.

Ex: a charger that is claiming a 200ma rate of charge, that is PWM a 1000ma power input, to "simulate" a 200ma charge , which it is not a 200ma charge at all. it is a PWM control of a 1000ma rate to average out to that rate.
Vrses a Charger which has a 250ma current control or capability and charges at an actual 250ma.

EX2: a charger that has only one actual "rate" but chops up that one power to provide a different rate (averaged) to smaller cells.
Vrses a Charger which might have included a different actual rate, for dealing with smaller cells.

EX3: topping off the battery using hard but short pulses.
VS a charger capable of topping using an actual reduced rate.

so you got :
Stop and check for voltage , which most of them use that method for checking the V-drop for termination.
you got Pulse it to change the "averaged" charge rate
Then less used is
charging the battery by purposfully pulsing it as the charge alogrythm intent, not as the "rate" adjustement. which does the same thing, just different intent.
and reverse pulse, which is usually not used, that was supposed to cycle it better, where it is acutually discharged some when the pulse is in the off state.

In Old School chargers, before everything was "switched" and so "binary" the "rate of charge" was actual current controlled rate, so 200 was 200. even if it also might have paused for a look at the voltage. now way more things are using pulse as rate control.

When a "topping" rate was applied it would actually be 20ma, not 1000ma for 20miliseconds. kinda analog.
Again, which can make a huge difference, as the action and activity on the battery are completly different, with the different methods.
when the "method" changes, some of the "how to" changes along with it.

:kewlpics:russel

VidPro said:

......so you got :
Stop and check for voltage , which most of them use that method for checking the V-drop for termination.

Click to expand...

Your probably way ahead of me here Vid, But I thought that the -dV during charging of nickel based cells could be detected by monitoring the charging circuit voltage just as easily as measuring the OC voltage of the cell. Therefore, there is really no real need to pulse the charge current and measure the OC voltage of the cell in order to detect the -dV drop, concerning single rate chargers. No?

Dave

Yes i think back in the old school, (depending on the charger) when they charged slower, some did not bother to STOP and check, they just had a solid current control going to the battery, then viewed BOTH the battery voltage and the charge voltage, for a voltage drop, or for the voltage to just quit moving up , the 0 thing.
before the 500 lines of code microcontrolled processes, were the normal.
Unless the power in stops, the battery voltage reading will be effected by the charge power.

could be each method has its own ramifications ? because stopping the charge, the voltage drop might lighten up some, each time the power is removed, not be as visable. and it can overcharge more on robust good resistance batts.
Then when NOT stopping, it could just blindly miss altogether on a battery that had really bad resistance?

If the Rate is going to be really high, like it is on Many of the new things, that might have changed the Need to stop to check, because a strong holding charge current with a high voltage limit would "obscure" the voltage changes occuring on the battery itself??? So mabey that another reason most of them stop to check?

Plus in the mix, some really "cool" chargers will stop on a voltage max , which has its plus and minuses.
and some old school chargers and series and cheap simple , that had Limited charge voltages (like 1.5v per cell), that would slow down as the voltage differentials (between charge and battery) were smaller, simple , more CV kind of control, and does not depend on v-drop for termination.

Hello Dave,

Another possibility is that a charger can be looking for a drop in voltage during the charge cycle, but may also look for a change in internal resistance during a pause in the charge.

I haven't been able to find a paper on this, but I believe at least one high end charger manufacturer incorporates something like this into their charging algorithm.

Tom

45/70 said:

As pertains to Nickel based rechargeable cells, the negative pulse, or "burp" charging method (which I'm pretty sure the Accumanager chargers use, as it supposedly negates the need for cycling cells etc. I don't know for sure about the Titanium Innovations Chargers, or others) has been debated for 40 years, or so. Some originally thought that it was the greatest thing since sliced bread. Others, who would seem to be more in the know, determined that while it doesn't really hurt anything, it doesn't actually do anything either.

Click to expand...

I am definitely in the later camp, having spent a considerable amount of time looking into this issue a year ago. You summed it up nicely as "while it doesn't really hurt anything, it doesn't actually do anything either". Perfect.

Not being an electrical engineer, it was my understanding that using a constant current power supply, combined with PWM, was the cheapest way to make a variable current charging system like the C9000.

edit: to address VidPro's response, there may be very good reasons to pulse the current in order to get good termination detection. This is not really a topic I know much about and therefore will simply leave it there.

Cheers,
BG

Very interesting. Would this pulse charging you all are referring to be the same as "negative Flex pulse" charging? This is what the MH 401fs touts as eliminating the need for a break in, or forming charge. thanks


Keith

i also assume the my accumanager 10 also pulse chargers too if the accumanager 20 does.

budynabuick said:

Would this pulse charging you all are referring to be the same as "negative Flex pulse" charging?

Click to expand...

Hi budy. This article covers the mythical advantages of "Burp Charging", pretty well. It's more of a marketing fantasy than anything else, as it really doesn't do anything. This is a bit different than a charging circuit that simply applies current in pulses, as there is polarity reversal involved, in burp charging.

Dave

45/70 said:

Hi budy. This article covers the mythical advantages of "Burp Charging", pretty well. It's more of a marketing fantasy than anything else, as it really doesn't do anything. This is a bit different than a charging circuit that simply applies current in pulses, as there is polarity reversal involved, in burp charging.

Dave

Click to expand...

hey Dave. Read your attached piece and while there is a dearth of hard evedence on this issue i decided (well it was really by accident) to try a little test for myself. Here is how it played out. I purchased 8 XX eneloop for my PA40 cause i heard they performed well in the cold. I might add that they do indeed! So i discharged all 8 of the XX's and all fell in the 1400-1500 range. I then put 4 in the c-9000 break in. While waiting impatiently for the forming charge i said "what the heck" and decided to see for myself if there was any merit to the claim of not needing a forming process if using the neg flex pulse claims made by the 401fs. The results of the break-in the final numbers were in the 2400 range for all four of the batts that were formed which i thought was pretty good. Now i took the other 4 XX's that i charged (no breakin) in the 401fs and discharged/recharged them and the numbers read lowest 2595 with the highest being 2650! Now i realize this is just a 8 battery test and not really conclusive but very compelling non the less. So i am now forced to buy more batts (darn it LOL) to see if i can get similar results. I will say without reservation that this 401fs does indeed give a deep/full charge without any heat problems (i have the updated model without the lid) on fast charge which is really not all that fast as it is only 1000 ma for AA. The jury is still out but I will do more tests when I get more batteries. But i think (even with just this one test) the numbers reveal there just might be something to this neg flex pulse process.

Keith

budynabuick said:

...I then put 4 in the c-9000 break in. While waiting impatiently for the forming charge i said "what the heck" and decided to see for myself if there was any merit to the claim of not needing a forming process if using the neg flex pulse claims made by the 401fs. The results of the break-in the final numbers were in the 2400 range for all four of the batts that were formed which i thought was pretty good. Now i took the other 4 XX's that i charged (no breakin) in the 401fs and discharged/recharged them and the numbers read lowest 2595 with the highest being 2650! Now i realize this is just a 8 battery test and not really conclusive but very compelling non the less...

Click to expand...

Before you get too excited, if you own a voltmeter, take a voltage reading of the cells INSIDE the 401fs right when the charger signals 'Done'. Then, remove the cells and let them 'rest' for 2 hours. THEN, discharge them in the C9000.

You might be seeing 'Surface Charge'. The C9000 is a very gentle charger - it terminates @ 1.47VDC and then charges @ 100mA for 2 more hours. It also has 'Rest Periods' of 1-2 hours between the Charge and Discharge functions in a programmed cycle. This allows the 'Surface Charge' to bleed off and the cells to cool if they've become hot.

Folks with the BC-900 (and the rest of the La Crosse family) consistently claim higher Capacities than folks with C9000s. That's because the BC-900 terminates @ ~1.53VDC and then IMMEDIATELY begins the Discharge function, resulting in misleading Capacity claims (~100mAh higher).

45/70 said:

...the -dV during charging of nickel based cells could be detected by monitoring the charging circuit voltage just as easily as measuring the OC voltage of the cell. Therefore, there is really no real need to pulse the charge current and measure the OC voltage of the cell in order to detect the -dV drop, concerning single rate chargers. No?

Dave

Click to expand...

The voltage drop along both positive and negative conductors from the charging circuit to the battery, and especially the voltage drop across the variably more-or-less crappy contacts between battery holder and battery make accurate voltage detection difficult when conducting current.
Vdrop​ = Icharge​*Rconductors ​
(6 inches of 20 AWG wire inside a typical shan zhai POS is about 0.005 ohms, and if its carrying 1 A of charging current, then you have 0.005 V being dropped.... %100 of the dV 'full-charge' threshold.)
That said, if the charging current were precisely the same, and nothing changed in the resistance of the conductors (the battery didn't get wiggled in the holder, temperature, moisture, etc), then the only thing changing would be the internal resistance and charge of the battery and you could reliably monitor the voltage at the current carrying conductors at the charger outlet even while charging for the NDV event.... but those factors are not reliably constant.
Negative dV 'NDV' chargers look at the diminishing voltage gains occurring when the battery is full and voltage ceases to climb, and if that's all they have to go on, then they are vulnerable to innacurate readings. When ya remove or diminish the charging current from time 1 to time 2, that voltage drop would diminish or disappear so V1 ​would be 'artificially' higher than V2​,so it would appear like a battery asking for more charge.... even if the battery is actually full! An increasing charging current would cause more voltage drop, so V1 ​would be lower than V2​, which would make it look like the battery were full (a false peak detection).

'Cause of all that, it's way better to take the current off in order to sense Voltage. A good NDV charger therefore pauses now and then to compare those readings at as close as it can get to a quiescent state, and a better charger also looks at temperature.... it's either do that, or run a totally separate wire alla way to the battery (all the way) which is not used for conducting current.
(Incidentally, that's the same reason that many alternators use a dedicated "sense" wire.)

Anyway, I hope that helps explain why it's a good idea to pause I and check V.

I know it's an old thread, but i ran into this one with a NiMH problem I had recently with voltage sensing charge termination.
In addition to the above regarding voltage drop of charging current interfering with accurate readings, the other problem the charger had was the ripple voltage of the switching power supply. Especially under charging load, the charger's output had quite a lot of ripple in it (the waveform was pretty ugly), and even with the battery absorbing the noise and whatever filtering the charger employed in the voltage measurements it wouldn't take much to screw up the voltage measurement by a mere 5 mV. Better to pull that crap off the battery before taking a reading.

anyway, thanks for the helpful thread

Odyss said:

Anyway, I hope that helps explain why it's a good idea to pause I and check V.

Click to expand...

It could also be because they uses a crappy switcher circuit for the charging that makes a lot of noise, making it difficult to measure the voltage with current on. The disadvantage by turning the current off is that the -dv is lower.