Control circuit for NiCd battery charger

[soldar]

I have a drill with 18V (nominal) battery packs. The charger is very simple and basically what is in the left part of the image. The values shown are not important, just a general idea.

The source is just a wall wart transformer followed by rectifier bridge and smoothing capacitor. Rs is just the source output resistance.

I have put in series a 220 ohm resistor for trickle charging and I short it with a switch for full "real" charging. The 5 ohm and the diode are original and I do not want to touch them.

So I manually close a switch for full charging. What I want to do is design a control circuit with a relay that will act as follows:

Upon connection relay is open so in trickle charge.

Momentarily pushing the "push to start" activates the relay which latches.

Circuit measures voltage at S ... which is the same as Vc when relay is closed .

Two conditions unlatch the relay:
1- Voltage over Vh which indicates battery is charged
2- Battery under VL which indicates a shorted cell

once the relay lets go it can only be activated again by pushing the button. Then the charger is in trickle mode and the battery a power supply are safe. (I once had a burnt transformer from a shorted cell.)

Vc is very variable, let's say 20 to 34 in order to give us some margin. A 24 V relay should work because it woul be activated at more than that and then the voltage would fall to about 23 V with the battery charging.

So, I am looking for ideas for the simplest way to implement this.

Maybe I would start with dropping the voltage to, say, 15 volt and then have to comparators, one to detect VH and one to detect VL and deactivate the relay.

There are probably already circuits I could use or modify. Maybe I can use a single comparator that detects a range? Also i am not sure how precise I need to voltages to be. Maybe detecting a shorted cell by low voltage would be easier but maybe detecting "fully charged" voltage needs to be more precise? In which case I would rather err on the side of stopping short and not overcharging.

I suppose another easy to implement addition would be a timer. One of those plug-in timers should work well.

Anyway. This is my first look at the problem and I am looking for ideas, schematics and anything useful.

[MarkF]

I know this is not for NiCd batteries.

  https://www.eleccircuit.com/gel-cell-battery-charger-circuit/

But, this gel-cell battery charger may do what you want.
It allows you to set the charging voltage (via pot) and
charging current (via series resistor in return path).

As the battery voltage approaches the charging voltage, the charge current will
decrease to a trickle.

The LED is 'On' while the battery is charging and goes off when charged.

[soldar]

I know this is not for NiCd batteries.

  https://www.eleccircuit.com/gel-cell-battery-charger-circuit/

But, this gel-cell battery charger may do what you want.
It allows you to set the charging voltage (via pot) and
charging current (via series resistor in return path).

As the battery voltage approaches the charging voltage, the charge current will
decrease to a trickle.

The LED is 'On' while the battery is charging and goes off when charged.

Thanks but it does not really do what I want. That is a classic float charger of which I have designed and built many for lead-acid batteries.  They have their uses but it is not what I need. It is good for other applications but not for what I need.

I might start designing something myself.

[Benta]

The best NiCd/NiMH charger ICs I've worked with were the Temic (now Vishay) U2402B and U2405B.
Obsolete since years, but the data sheet might give you some inspiration for a roll-your-own charger, especially the algorithm.
Would be an easy job for a small MCU with A/D.

[Kim Christensen]

The most brain-dead simple is a timer that controls the relay. Some old rapid chargers worked that way. The rapid charge current timer was set so that the amount of mAh fed was about 80%-90% of capacity and then relied upon the 0.1C trickle charge to do the top up. But these required the operator to be smart enough to only charge exhausted batteries. Hitting the button on a fully charged battery would damage it with that circuit.

A modern NICAD/NiMH charger monitors the battery voltage and cuts the current after the voltage has peaked and starts to decline again. Better ones will also measure battery temperature as a backup and will terminate charging after a set amount of temperature rise and won't charge a too hot battery at all.

[soldar]

So far this is the idea I am working on. Q1 and Q2 form a bistable and they both conduct or not. When power is first applied they do not and are "reset". Pushing the start button sets the bistable and they both conduct so the 220 ohm resistor is shorted and the charger is in "high". If Q3 conducts the bistable is reset.

Two comparators detect when the voltage is too high or too low and signal Q3 to conduct. This puts the charger in trickle current charge. If it is because the battery is full (or no battery connected) then good. If it is because the battery voltage is too low then it prevents over-current which can damage the transformer.

I just realized the 5R resistor raises the voltage by 0.5 volts per 100 mA charging So I would need to take that into account. Or sense after the resistor but that means one more wire from the existing charger. One advantage of what I have is that I only need to cut the positive wire to insert the 220V resistor.

I need to think about this a bit more. Especially about the comparators.

[Kim Christensen]

Configure Q1 so that it operates as a constant current source. Then the voltage across the 5R will not change much when in rapid charge mode.

If you want to get even fancier, you could eliminate the 220R and configure Q1 to be a two step current source. (Low current when Q2 is off and high current when Q2 is on)
EDIT: Just add a 2nd current source (trickle current) in parallel (with Q1's) and switch the Q1 current source on/off.

[soldar]

Configure Q1 so that it operates as a constant current source. Then the voltage across the 5R will not change much when in rapid charge mode.

If you want to get even fancier, you could eliminate the 220R and configure Q1 to be a two step current source. (Low current when Q2 is off and high current when Q2 is on)
EDIT: Just add a 2nd current source (trickle current) in parallel (with Q1's) and switch the Q1 current source on/off.

yeah, this is the usual evolution....

Right now I have just cut the cable and inserted a 220 ohm resistor for trickle charge and I short it with alligator clips for normal charge. So I think, you know, I could automatically remove the short when it is charged. And then I could also do it if the voltage is low. and then... the project begins to escalate and it makes more sense to design a new charger from scratch... and then it seems like not worth the effort and nothing is done. It is the story of my life, not only with electronic projects but in general. Home renovations etc. A simple little project ends up being so complicated and expensive it never gets done. I really suffer from this.

If I want a better charger then I should probably design and build from scratch because the transformer has high output resistance, designed to limit the current, but it soon gets too low for any control circuit to have any headroom.

I will have to keep thinking about this and decide which way to go.