NiCad Charger Question

You risk smoking the chip if either 24V or 48V goes to ground. Why not just use a resistor. A light bulb also works.

"Chris"

.... Phil

The 7812 will dissipate 0.72W into a short circuit and 0.32W into a flat battery. No problem there provided there is enough heatsinking.

You might need capacitors on the inpuit and output of the regulator.

I'm not sure the 7812 will tolerate 24V on the output when the 48V is absent. I'd fit a diode in series with the output so the battery can't discharge through the 7812 when the 48V is absent.

Well, I was considering the changing voltage as the battery charge peaks. Since there are so many cells as the voltage across the resistor will change, the current will change significantly. Here is a revision:

I don't this will smoke. I don't expect the bulb to light much at

Chris

Not sure why (puzzled look).

Done. I also removed the load incandescent lamp per your post. The .png has been updated to reflect this.

Thanks, Chris

"Chris"

Well, I was considering the changing voltage as the battery charge peaks. Since there are so many cells as the voltage across the resistor will change, the current will change significantly.

A 20 cell NiCd pack will start charging at 24 volts and peak at 28.

That makes the current variation using a resistor only 20% = peanuts.

Use an 1100 ohm resistor and the variation is only +/- 10% around 20mA.

..... Phil

Allright, easy enough.

Chris

Well, I was considering the changing voltage as the battery charge peaks. Since there are so many cells as the voltage across the resistor will change, the current will change significantly. Here is a revision:

I don't this will smoke. I don't expect the bulb to light much at

Chris

Consider using an LDO type... I use LM117T but VinMAx is too low for this.I use them for 180 mA NiCad charger at 12V i/p with no heatsink, just the tab soldered to a 'ground plane'.

An LM317 would work fine, just as long as the output is not shorted while it is powered up to 48 volts. For the OP an LM317 is perfect if you can get the supply down to 31 to 40 volts.

Huh? 12V IC for 24V battery pack? You do realize that once you get the battery pack past 12V you'll not be able to charge it up any more?

If you want a simple constant current charger then rig one up using a bjt. It is quite simple(although wastes power). In any case you'll need 24V+. The problem with NiCad is that some cells will "short" due to dendrite buildup which will only result in heating up the batteries without ever fully charging them. To solve that you have to pass a large current through the batteries for a short period of time. I tend to "zap" them using a 12V car battery for about 1/2 to 1 second(which is relatively dangerous). This works in recovering about 80% of the dead cells. In many cases though the dendrites will rather quickly recover and need to be zapped several times in rapid succession.

The voltage regulator is being used as a constant current source.

Iout ~ (12-0.6)/600 = 19mA

Although without the decoupling capacitors on input and output it is quite likely to squeg at about 10MHz and be generally unhappy.

With 24v over voltage available he would be better off with a resistor of about 1.2k and power rating 2W. Simple and indestructible. Initial charge rate 20mA and when at full charge ~17mA. Even if the cells were badly discharged they quickly come up to 1.2v/cell and the maximum current into a dead short would be limited to 40mA.

You might want to add something to switch it off or to trickle charging when the terminal voltage reaches 1.4v/cell (ie 28v). That way when you forget about the batteries on charge they do not suffer unduly.

Regards, Martin Brown

It doesn't matter what it is. A "constant current source" is just a variable voltage supply that varies in a way to keep the current constant. The voltage source used has a limit and it is called the voltage compliance.

What is going to happen if you use a constant current source with a 12V supply, when the battery pack gets above 12V the constant current source will then go out of compliance and never be able to send any current. It is very simple.

G--V---R---B---G

Where V is the power supply voltage, R is a regulator for constant current, and B is the batter pack. Simple analysis how's that

V - I*R - B = 0

if V < B then I < 0 and current will flow out of the battery.

V must always be larger than B which says that the compliance for a constant current power supply must always be larger than the maximum voltage you will be charging too.

Simple result: If you use a constant current source with a compliance of 12V you'll never be able to charge anything above 12V.

You seem to believe that a constant current source allows you to charge up any voltage. Can you use a constant current source to charge up a 1M voltage battery? Even if the constant current source is derived from a 1V battery? One can, of course, insert a step up before the constant current source to increase it's compliance but since the OP specifically mentioned using a 12V source and not stepping it up implies that we are dealing with a 12V compliance(max).

If the OP were to use a 12V to 30V DC-DC converter then he is increasing the compliance. He will need to do something like this else he will never charge up the batteries past 12V.

A very simple analogy is to think of the constant current source as supplying a small amount of gas into a closed container at a constant rate. Once the pressure of of the whatever is supply the gas reaches the pressure of the container we can no longer pump any more gas in. What happens as we near that point is that our "constant current source" no longer becomes constant as it does not have enough pressure to overcome what is pushing back. If we increase the supply pressure we are then able to supply more gas to the container.

So even though our "supply pressure" is constant the pressure in the container is not. It is varying and once it gets filled to the maximum pressure we can supply we then cannot supply any more unless we get a bigger pump.

I suggest you look carefully at his circuit diagram again. It is *NOT* on a 12v supply. The 12v regulator is being used to generate a constant current from a *48v* DC supply voltage. 48v >> 24v so the thing will work but isn't well protected or properly decoupled against RF oscillations as drawn. I expect 48v would kill a 7812 if the output was shorted to ground since max input voltage is something like 35v.

It is clear you do not understand the circuit. There is regulated 12v across R1, D1 and all that current that flows through the battery. There is 48-28 v dropped across the 7812 when the battery is fully charged. This is way more than is needed for regulation.

The 7812 is bogus a 1.2k 2W resistor would be a lot more robust, and perfectly adequate under the circumstances working from a 48v supply.

Regards, Martin Brown