Simple NiMH Charging Circuit and Resistance

[LukeDupont]

Hello everyone!

I'm quite new to electronics. For one of my first projects, I decided to make a USB / Solar NiMH trickle charger. I decided to do this with only a diode and a resistor, as I do not need nor want quick charging, and while I know there are more complicated circuits to provide a more regulated constant current, I want to build something simple, and power efficient which can tolerate dips in voltage.

So, here's the simple circuit I came up with and built.

My Diode drops the voltage by about 0.7V, and my input voltage is about 4-5V. As I'm charging a pair of 2500mAH NiMHs, I selected a 25Ω 1W Resistor, which should get the current down to about 250mA at 5V.

...Or so I thought! The actual current I got was far, far less -- barely enough to power an LED or two!
I found that a 10Ω Resistor got me up to 140-160mA, but I dont understand why my calculations are so off. Did I place the resistor in the wrong part of my circuit, or am I not accounting for the resistance of the batteries? I tried measuring one of my batteries' resistance by selecting Ohms on my multimeter and probing the positive and negative terminals, but got no reading. They can't have that much resistance, because when I short the resistor, I get a very large current equal to roughly what my power source can supply.

What am I missing? Should I have calculated the resistor's value across the voltage of the two NiMH (2.4V) as opposed to the power source (5V), or the power source minus the diode (4.2V)? I'm a little confused as to what the resistor is acting on, and whether its function were to change if, say, I placed it before the diode, or at some other location.

 

[Chemelec]

Your battery is a Nominal 1.2 Volts.
But when charging, it Charges up to about 1.6 volts.

 

[BobK]

I selected a 25Ω 1W Resistor, which should get the current down to about 250mA at 5V.

How do you figure that? Not even considering the 0.7V drop of the diode or the 2.4V drop of the batteries:

I = V / R = 5 / 25 = 0.2 = 200mA.

The correct calculation is:

V = 5 - 2.4 - 0.7 = 1.9

R = V / I = 1.9 / 0.250 = 7.6Ω

But a trickle charge for NiMH should actually be 0.05C or 125mA for a 2500mAH battery.

Bob

 

[Audioguru]

Energizer battery company and a Japanese one say that a Ni-MH battery must have a trickle charge no higher than 1/40th its mAh rating to avoid damaging the battery. 1/40th of 2500mAh is 62.5mA.

 

[LukeDupont]

How do you figure that? Not even considering the 0.7V drop of the diode or the 2.4V drop of the batteries:

I = V / R = 5 / 25 = 0.2 = 200mA.

The correct calculation is:

V = 5 - 2.4 - 0.7 = 1.9

R = V / I = 1.9 / 0.250 = 7.6Ω

But a trickle charge for NiMH should actually be 0.05C or 125mA for a 2500mAH battery.

Bob

Ohh, I see! I have to account for the voltage drop of each component, and then calculate the resistance based on the remaining voltage. I think I'm a little confused as to what a resistor "acts on."

This means that as the "left over" voltage lessens, so does the current allowed to pass through with a given resistor, correct?

It seems I could use this to my advantage, as the voltage of each cell rises as it charges, ending at around 1.5V. This means I can have more current when the cell is at a lower voltage, and less when it nears full charge.

I did some new calculations based on 10Ω @ 5V, and they seem to line up with my observations of the circuit I built:

10Ω @ 5 - (1.5*2) - 0.7 V = 130mA #at end of charging cycle
10Ω @ 5 - (1.2*2) - 0.7 V = 190mA #at start/middle of charge cycle

I get between 110mA and 170mA depending on the charge/voltage of the batteries, and the particular USB cable I use, which would probably make sense once you account for the added resistance in the batteries, wires, and cables.

This results in about 0.05C, which should be fairly safe for when the device is plugged into the wall for prolonged periods. However, I primarily want to use this with a small 5V solar panel, and that requires some more considerations.

I'd like to charge at a higher rate from solar, given limited / varying sunlight, and the fact that solar panels are generally over-rated. From my experience, I usually get about 4.5V from a 5V panel, and about 0.66% the rated current. Higher than that is the exception rather than the rule.

I've read up quite a bit on this, and find that people generally pick solar panels used with nimhs that can provide 0.15-0.2C at their rated current (*without the use of a current limiting resistor). This results in closer to 0.1C in real world conditions, and doesn't adversely affect the battery.

I've heard varying claims that trickle charging for NiMH should be limited to 0.1C, 0.05C, 0.025C, or even that you shouldn't trickle charge at all. Without more precise data as to how long at what current what extent of damage occurs, I can only assume that the lower end of these estimates are *very* conservative and assume that the battery is being left to charge indefinitely from the wall. I do not ever intend to do that, so I think it's fair to select 0.1C, especially considering the efficiency of my solar panel.

Alternatively I can have a selectable resistance for wall/solar charging, with a more conservative current limit for charging from the wall (maybe 0.05C?)

 

[Audioguru]

Modern Ni-MH batteries made by Eneloop (Sanyo/Panasonic), Energizer and Duracell hold a charge for one year and are sold precharged. They do not recommend trickle charging if you want them to last for 5 years.
0.1C is a recommended overnight (10-14 hours) charge rate for Ni-MH batteries. If you accidently charge a fully charged battery at 0.1C overnight then it gets pretty hot and its life is shortened, which is why a smart charger circuit is recommended to sense a full charge then switch off the charging.

My solar garden lights use AAA or AA Ni-MH batteries that are cooked in sunlight all day at a fairly high current. They get hot and last only a couple of years. Old Chinese ones rusted away in a month or two.

Solar panels? If they have a glass cover they last for a long time. But if they are cheap with a plastic cover then the plastic gets sunburned by UV and becomes cloudy blocking some of the light.

 

[LukeDupont]

Modern Ni-MH batteries made by Eneloop (Sanyo/Panasonic), Energizer and Duracell hold a charge for one year and are sold precharged. They do not recommend trickle charging if you want them to last for 5 years.
0.1C is a recommended overnight (10-14 hours) charge rate for Ni-MH batteries. If you accidently charge a fully charged battery at 0.1C overnight then it gets pretty hot and its life is shortened, which is why a smart charger circuit is recommended to sense a full charge then switch off the charging.

My solar garden lights use AAA or AA Ni-MH batteries that are cooked in sunlight all day at a fairly high current. They get hot and last only a couple of years. Old Chinese ones rusted away in a month or two.

Solar panels? If they have a glass cover they last for a long time. But if they are cheap with a plastic cover then the plastic gets sunburned by UV and becomes cloudy blocking some of the light.

Thanks! That gives me a good rough idea of how much damage might occur from over charging at 0.1C.

It occurred to me that if I could have reverse current protection and regulate from 5ish to 3.4 volts accurately, I could select a resistance that initially charges at something like 0.2C and ends at 0.05C as the cell's voltage rises. The problem is that a linear regulator requires an input of at least 2V above the output, and I still need a diode for reverse current protection, putting me at roughly 2.7V drop. As I'd like this to work down to at least 4.5V, that means I have only 1.5V to spare. Two diodes would drop me down 1.4 or 1.6V, but that's relative to the input voltage and not a fixed reference, so no good.

Maybe I'll look into limiting regulating with Zeners, or maybe I'll just stick with a simple low current charger.

Here's what my current, simple design with a simple diode and resistor looks like:

#Low solar input at 4.3V, 0.8V dropped across diode, 7.5Ω resistor:
4.3 - 0.8 = 3.5
3.5 - 2.4 = 1.1V / 7.5Ω = 147mA
3.5 - 2.7 = 0.8V / 7.5Ω = 107mA
3.5 - 3.0 = 0.5V / 7.5Ω = 66mA

#High solar or wall input at 5V, 0.8V dropped across diode, 7.5Ω resistor:
5.0 - 0.8 = 4.2
4.2 - 2.4 = 1.8V / 7.5Ω = 240mA
4.2 - 2.7 = 1.5V / 7.5Ω = 200mA
4.2 - 3.0 = 1.2V / 7.5Ω = 160mA

So that's ending C/15 at 1.5V for 2500mAh cells, and C/12 for 1900mAh cells. If the batteries attempt to charge above 1.5V, the current will continue to decrease slightly. I think I will give this setup a try; it should be fairly safe with 2500mAh and 1900mAh NiMH, especially when charging from solar, and so long as I don't leave full batteries plugged into the wall overnight.

 

[Audioguru]

Your voltage idea does not work because each battery cell and each manufacturer and maybe temperature changes the voltage. Didn't you learn about how a full charge on a Ni-MH is supposed to be done and how all charger ICs do it? They sense a full charge when the charging voltage drops! Here is a graph of a typical (some cells have a higher voltage and others have less), also look at the temperature rising. At a current of 0.1C or 0.3C the voltage for many cells never reaches 1.5V but the cell is being damaged by over charging::

 

[LukeDupont]

Your voltage idea does not work because each battery cell and each manufacturer and maybe temperature changes the voltage. Didn't you learn about how a full charge on a Ni-MH is supposed to be done and how all charger ICs do it? They sense a full charge when the charging voltage drops! Here is a graph of a typical (some cells have a higher voltage and others have less), also look at the temperature rising. At a current of 0.1C or 0.3C the voltage for many cells never reaches 1.5V but the cell is being damaged by over charging::

Ah, thanks! Those charts are useful.

I have read about charger ICs and was under the impression that voltage drop and temperature change sensed by these ICs does not appear when charging under 0.1C. This is collaborated by your second diagram.

What I didn't know, but your second diagram makes clear, is that charging at such a low current, the cells will not exceed 1.4V. This is what I've found in my own charger using a 10Ω resistor, which finishes at about 110-130mA @ 1.4V. I was thinking it wasn't charging them fully, but apparently it is!

All of my NiMH seem to finish at this voltage for a current between 0.1C and 0.05C, so I suppose I should plan on ending at 1.4V and not 1.5.

That said, I'll reconsider my "Semi-fast" charger idea, because it's unclear what the curve would look like. In theory, I still think it would work, but it would take some careful calculation and testing with various battery types in order to avoid "over stimulating" the battery and raising the voltage too much.

 

[Audioguru]

The graphs show a "typical" battery. Some have a higher voltage and others have a lower voltage. A battery charger IC charges at 0.3C to 1C and detects the voltage drop just after a full charge.

 

[LukeDupont]

Ok, well, I think I'll stick a simple low current, fairly linear trickle charge ending at around 0.05C for wall charging, and maybe a little higher limit around 0.1C for solar. I'll stay away from charging any higher than 0.1C without a temperature / voltage sensing circuit.

 

[Audioguru]

Why don't you use a charger IC to do it properly?

 

[LukeDupont]

Why don't you use a charger IC to do it properly?

Because that would be a sure way to over charge the battery catastrophically when charging from solar, as the charging cycle would be interrupted frequently.

 

[Audioguru]

I trust that the manufacturer of a charger IC does everything correctly.

 

[LukeDupont]

I trust that the manufacturer of a charger IC does everything correctly.

Fast charging ICs are designed to remain plugged into the wall, providing a consistant, uninterrupted power source. They're not designed to be interrupted. With a solar power source, power output varies drastically and is interrupted often. This will cause the cells to cool off, the voltage to drop prematurely, and you'll either receive a "stop" signal too early, or, more catastrophically, miss it altogether.

It's not that such ICs are bad, they just isn't made for such applications. At least, this is what I have gathered from my research.

 

[Audioguru]

My cheap solar garden lights use a Ni-MH battery cell and it has no problem with solar charging with clouds mixed in.

 

[BobK]

My cheap solar garden lights use a Ni-MH battery cell and it has no problem with solar charging with clouds mixed in.

Trickle charged with no termination condition, relying on it eventually getting dark and discharging the battery.

Bob

 

[Audioguru]

My cheap solar garden lights come with a Ni-MH cell now instead of the old Ni-Cad. But the battery cell is probably 2/3rds full of rice since the AAA cells are only 300mAh compared to a 800mAh Energizer or 850mAh Duracell that are both made in Japan and might be Eneloops. The cell is cooked in the sun all day plus the charging creates heat. The lights are still lit in the morning after being on all night long. The original Chinese cells rust away in one or two months. The Energizer and Duracell cells last for years.