# Schmitt trigger design to control lead acid battery charger

I've found a very good explanation on schmitt trigger circuit design here -What is a Schmitt Trigger and its application exactly?

But the design is done considering same reference voltage and the comparator biasing voltage v1. It means threshold voltages (high and low) will be varied along with the supply voltage. I want to achieve a constant threshold values to design a battery charge controller. For example Vth(H)=13.8v and Vth(L)=13.0v. How can I achieve this?

• Use a voltage regulator Jun 26, 2015 at 8:40
• Yes, I did it. I used 5.1v zener diode. Jun 26, 2015 at 11:59
• No, use a voltage regulator. Jun 26, 2015 at 12:26

This might not be efficient, but works.

You can go for a voltage reference IC like this one. This can be used as supply voltage for your Schmitt trigger IC alone. You can scale down the threshold of 13.8 V and 13.0 V to 4.6 V and 4.3 V respectively (or any other values), using high value resistor divider at the input side of the Scmitt trigger(So that, it doesn't waste current there). The UTP and LTP are now referenced to an independent constant voltage reference from the IC. The inputs however will vary depending on battery voltage.

• Pretty much the approach any circuit that needs to generate a constant output voltage takes, too - eg, power regulators. Jun 26, 2015 at 9:29
• @Nick Yes, i agree. A Buck regulator can be used(LDO wastes more power). i wanted to simplify. Jun 26, 2015 at 9:39
• Using zener diode I made Vs and Vref both 5.1v. Using R1=10k, R2=1M and Rfb=220k I got threshold as 5v and 4.6v. While testing, higher threshold seems work but lower doesn't. While increasing Vin, output goes low at around 5v Vin, if I decrease the Vin a little, output goes high again. While decreasing, output shouldn't go high again if Vin doesn't go lower than 4.6v Jun 26, 2015 at 12:09

Since this is a battery-powered system I do not advise running the Schmitt trigger itself from an LDO. Run the trigger's comparator on Vbat itself and choose a comparator that

• has sufficient PSRR, since you expect your supply to vary;
• is rail-to-rail; and
• supports a single-ended supply at the voltages you expect.

Use a low-power reference such as a fixed, 2.5V LM4040 for your negative terminal. This draws as little as 45uA typ, though you'll want to exceed its max of 60uA in the worst case. Your threshold analysis will need to take into account the two different supply voltages expected at each threshold. In other words your hysteresis curve will not be rectangular but will look, crudely, like with transition analysis points in red. A possible solution for this system is simulate this circuit – Schematic created using CircuitLab

where

$$\frac {R_2}{R_4} = \frac {69} {65} \left( \frac {13} {V_{ref}} - 1 \right)$$

$$\frac {R_3}{R_4} = \frac {69} {4} \left( \frac {13} {V_{ref}} - 1 \right)$$

This yields the correct results: 