I need to charge a 10 cell nimh pack, but my input voltage is only
12VDC. I dont want to use a switching power supply.
I'm thinking that if I charge the 10 cells in two banks, i.e., 5 at a time, and I switch between the banks very quickly (100 times per second say), that I will be able to charge the entire 10 cells okay. And since I am switching between the packs so quickly I could still use dt/dt termination (change in temperature rise) since its a relatively slow phenomenon and switching between packs is essentially like using a pulsed-current charge method which is pretty standard.
Whats the point of you even responding? I'm NOT going to use a switching power supply. Thats the whole point of the post.
Haven't you people ever heard of experimenting for experimentings sake? How did any of you learn electronics with your boring can't-do, shouldn't-try, attitudes?
Oh, wait you don't want the best solution, only one that you've already determined is what you want.
Sure you can do it that way.
But WHY? Design your charger to charge the first five cells then the other five, why 'high speed' switch between the two banks. Do one bank, then the other.
I understand there might be a better way to do things, but I _want_ to do it this way to see if it can be done! Plus it could be a lot cheaper than a switching version.
Because this pack will be used in a device that needs to be available at any time, and if someone decides to stop the charge at some random point during the charge cycle, you end up with a battery pack where half the cells are dead and the other half are partially charged. So switching between them at a relatively high speed ensures the cells all get a roughly equal amount of charge per unit time. (human time anyway)
All you need is a 555 timer switching the current source between the banks 60 times a second or so.
Okay then two seperate LM317's charging the banks seperately, one charge controller that uses dt/dt (BQ2002T, ~$1 @ 1k qty), a relay for connecting it all up, and there you have a way to charge 10 nimh cells with only 12V and no switching power supply, and you do it by charging
It seems if you are not careful, you can get the two different groups of batteries at different potentials. As you discharge the batteries, you only have one pack. I think this can lead to a scenario where you get cell reversal. Take a look at this:
formatting link
The nice thing about charging the pack as one group is if the batteries are well matched, they see the same charge and discharge current, so there is less of a likelihood of cell reversal.
Remember though, the charge termination comes from one thermistor, which is in thermal contact with both "banks", so the charge termination will occur at the same point as if you were charging all the cells as one pack. I'm not sure there is any way, from the cells perspective, that a difference between charging all the cells at once and charging two banks at once could be perceived.
dT/dt isn't all that precise. In any event, you would have to implement a discharge before charge if you really want to trust that both sides got charged to the same level.
dT/dt is really flaky. Remember, you use multiple end of charge detection schemes, so how do you know voltage depression didn't terminate the charge?
I had a customer problem one time that ended up being due to an extra layer of kapton tape between the batteries and the thermistor.
When one of the banks reaches the DT/dt set point, it shuts down charging. The other bank does not get a full charge, except by chance. You need individual control and sensing of each bank.
If you have the capability to separate the thing into
2 banks as you mentioned in your earlier post: "Okay then two seperate LM317's charging the banks seperately," then there is no issue. Run one supply into two separate controllers, with equal termination set points. The controllers can share some circuitry, but each bank needs to be terminated depending on the signal from its own sensor.
There's drawbacks/compromises/etc to whichever way you do it. I like your idea separate LM317's, provided you can separate the pack into two individual banks of 5 cells each, and you terminate the charge individually.
Actually, since anything you do is a compromise when you split the pack, I'd set each 317 for constant current with a taper charge when terminal voltage is reached. The easy way is to set a TL431 to the terminal voltage, and have it shunt current from the source when that voltage is reached. That way, you can continue to pump constant current from the 317, but divert some portion of it through the 431, and provide the NiMh with whatever C/X you decide.
A nice side benefit is that you can light a red led and extinguish a green led with that approach, so that you see that the pack is fully charged, and use the current through the led as part of what you divert. You can also put a resistor in parallel with the led to divert more current.
The TL431 has a max of 100 mA AIRC - if you need to shunt more than say 75 mA, you can control a transistor with the
431.
Is that the best possible approach for charging an NiMh. No. But it is simple, it works, and you are forced to compromise in your situation, so it seems a reasonable way.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.