# Latching Solenoid Drive Voltage Problem

I have a circuit which drives a latching selonoid valve(valve changes state for every 50ms pulse)

Below there is an easy go schematic of the circuit. Basically a 3.7V LiPo Battery supplies for 3.3V mcu regulator and a boost converter of 9V's (LM2623).

To switch the valve mcu enables the boost converter and waits for 10ms for stabile 9V. And then set drive pulses to valve driver.

When I connect the board to MSP430 Launchpad(debugger, vcc=3.5V) and in debug state; if I run the code, valve perfectly switches between state. In the switching time, voltage on coil only decreases to ~7V.

When I disconnect the debugger and run with only battery, code tries to swith the valve but, it not switches. And voltage on coil decreases till ~5v or so.

How can this extra voltage decrease can be compansated. There is already 100uF connected to valve driver supply.

simulate this circuit – Schematic created using CircuitLab

You can calculate the required capacitor from:

C= $\frac{I \cdot T}{\Delta V}$

Where I is the coil current, T is the pulse time, and $\Delta V$ is the allowable voltage droop.

For example, if the current is 1A, the allowable droop is 1V and time is 50ms, then a 50mF capacitor (50,000uF) is required. Naturally, a larger capacitor will take more than 10ms to charge if the battery cannot supply enough current to hold the supply up for 50ms.

Basically as described I was wait for 10ms to stabilize 9V output. But after @Spehro Pefhany's comment. I re-inspected with oscilloscope and I saw It was not reaching to 9V's.

To be sure I made a delay of 750ms and inspected on scope. And I saw It was taking ~100ms to reach stabile booster output voltage and charging of capacitors. Therefore it started to latch the relay.

Here is an scope inspection(booster configured to 11.8V) :