2
\$\begingroup\$

I have built a solar charger that charges 6 AA NiMH rechargeable batteries in series using a constant current source at a little under C/10. That part is fine.

I want to use the batteries to power a 3 W LED that is driven by a different constant current driver. Again, that part is fine. To power the LED I want to use 2 lots of 3 series-connected batteries, connected in parallel.

To change over between charge mode and power-the-LED mode, I intend to use a double-pole, double-throw switch using this arrangement:

charge / power-the-led arrangement

Is this safe, or are their risks involved when switching from charge mode to power-the-LED mode?

Updated circuit (Schottky diodes added) to address feedback received:

enter image description here

\$\endgroup\$
5
  • 1
    \$\begingroup\$ How are you handling cell balancing during charging? Each battery has different output impedance, so they will not all discharge at the same rate. If the capacity or discharge level of each cell is initially different when charging begins then can your charger still charge all the cells as much as you need, without over-charging any of them? \$\endgroup\$
    – user4574
    May 31, 2022 at 23:06
  • \$\begingroup\$ It depends how you handle charge cutoff and UVP with LED Vf min \$\endgroup\$ May 31, 2022 at 23:33
  • 2
    \$\begingroup\$ This approach seems fraught, instead find a constant current driver that will efficiently operate from 7.2V \$\endgroup\$ Jun 1, 2022 at 0:27
  • 1
    \$\begingroup\$ @user4574 I've added that the charger uses just under C/10 to the question. I don't believe that overcharging is an issue at that current. @ TonyStewartEE75, UVP is me seeing that the LED is getting dim and switching it off. (I'm using it as a replacement to a desk lamp.) @ Jasen I'm not sure of your reasoning?? \$\endgroup\$ Jun 1, 2022 at 12:00
  • \$\begingroup\$ @user4574 If the charge current is low enough, NiMH will self-balance during end of charging. \$\endgroup\$
    – winny
    Jun 1, 2022 at 14:05

4 Answers 4

2
\$\begingroup\$

You need to ensure that the switch is 'break before make' -- i.e. when the switch is changing over; in the middle, you have all switches open, not all closed. Most simple SPST slide switches are like this, but some rotary switches may not.

If you are charging in series, be aware that if the cells are not at equal states of (dis)charge (e..g some have aged differently from others), then you are at risk of over charging those cells as you wait for the whoel stack to reach a termination votlage. If your charging current is less than the continuous trickle charge limit of the cells, then there would be less of an issue.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ That's a helpful point to make about the switch, @jp314. I've edited the question to state that charging is at just under C/10, so overcharging isn't an issue. \$\endgroup\$ Jun 1, 2022 at 12:09
2
\$\begingroup\$

It might be safe but the cells will likely suffer.

It is probably safe in the sense that it won't burn your house down and if you follow the suggestions in the answer by jp314, it may work fairly well.

However, if you just charge the series string of six cells without any special balancing, the two halves of that stack will likely end up at slightly different final voltages. Once you flip the switch over, the higher-voltage 3-cell battery will immediately start discharging into the other 3-cell battery and will keep doing so until their voltages equalize. Even if the voltage difference is just 0.1 V, the small resistances involved will lead to non-negligible currents, potentially overcharging and damaging the cells in the receiving half.

Worst case, once one of the cells fails (loses capacity, develops a short), the resulting large equalization current may easily kill the other two cells in the same stack.

Finally, there's also the risk of you forgetting to flip the switch into the right position and then frying the LED driver by feeding >8 V to it. Similarly, if your charger is dumb enough, connecting the parallel arrangement to it may damage the cells. (And if it doesn't, this means you could have just used the parallel arrangement all the time and skip the switch altogether.)

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for your detailed response TooTea. I've updated the circuit to include Schottky diodes that, hopefully will address the issue you highlighted when the two stacks are at slightly different voltages. As for potentially frying the LED driver, I'll just have to be careful. One thing in my favour is that I will have to unplug the LED and plug in the charger when the time comes and so I should remember to flip the switch when neither is plugged in. \$\endgroup\$ Jun 2, 2022 at 17:12
  • \$\begingroup\$ I see now that it needs to be a 3 pole double throw switch. \$\endgroup\$ Jun 3, 2022 at 17:21
  • \$\begingroup\$ @MarkHightonRidley -- You could use a microcontroller to sense any voltage coming in before you actually connect to it, and this will foolproof it so that you don't have to "be careful". You can use latching relays so that they only draw power when changing states. \$\endgroup\$ Jun 4, 2022 at 12:14
  • \$\begingroup\$ @MicroservicesOnDDD Right,but then OP could also just use a LED driver capable of handling 9V and live happily ever after without switches,relays,microcontrollers and all that complexity. \$\endgroup\$
    – TooTea
    Jun 4, 2022 at 17:50
  • \$\begingroup\$ Thanks for the suggestion, @MicroservicesOnDDD but it's overkill for what I'm trying to achieve. \$\endgroup\$ Jun 4, 2022 at 21:02
1
\$\begingroup\$

In it's original form, it wasn't entirely safe (from the perspective of the batteries) for reasons given in the comments on it. I asked the same question in the Electronics Hobbyists Facebook group and also got some helpful suggestions. As a result of both sets of feedback, I've updated the schematic and built and tested it. It works well.

Updated Schematic

The schottky diodes (1N5817) drop 0.45V on each battery stack when powering the 3W LED, and that's just fine. They also prevent any issues that would otherwise arise if the two battery stacks end up at different voltages after charging or when powering the LED.

I do have to remember to set the switch to the "Charge" position before plugging in the charger, and to the "Power-the-LED" position when the LED is plugged in and before I switch it on.

\$\endgroup\$
0
\$\begingroup\$

There are issues with charging in series. Unless you are using a certain NiMH formula for charging in series, the weaker cells will tend to get less charge and overdischarge. If you use a higher charging current, you may cause those weaker ones to overcharge, bulge and vent.

The problem is that they don't charge up exactly the same, even at a 0.1 C rate of charge.

There were some brands where you could charge in series on slow and the capacities would stay the same with a smart slow charger. The company used a certain NiMH formula I have not seen with other brands (Maha) C size. They didn't put much effort into matching their cells during manufacture, though.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.