Solenoid driver circuit

[earckens]

A 12V DC heavy electrovalve needs to be driven by a circuit (with 555 timer) that gives a start pulse of 12V and after a set time delay (here R3,C2 gives about 2.5secs: C2 is replaced by a 2.2uF cap) the valve supply voltage drops to 4.5V dc, enough to keep it activated but low enough to draw a minimal (750mA) current.

When using a resistive load the circuit works fine. When using the electrovalve I notice that pin 3 of the 555 gives a downward voltage drop to 0V dc (as it should do) but this drop however merely is a spike and it goes back to 5V dc, hence keeping the TIP125 on, and the transistion from 12V to 4.5V does not occur. This spike occurs every time RC has passed.

What could be the reason for this, given this circuit?

 

[crutschow]

You may be getting some feedback from the solenoid inductive spike through the MOSFET Q3's collector-gate capacitance to the base of Q1 and retriggering the 555.
Try placing a diode in series with the MOSFET gate (cathode to gate) with a 10kΩ resistor from the gate to ground.

 

[earckens]

You may be getting some feedback from the solenoid inductive spike through the MOSFET Q3's collector-gate capacitance to the base of Q1 and retriggering the 555.
Try placing a diode in series with the MOSFET gate (cathode to gate) with a 10kΩ resistor from the gate to ground.

Thanks! Will try that and let you know.

 

[earckens]

You may be getting some feedback from the solenoid inductive spike through the MOSFET Q3's collector-gate capacitance to the base of Q1 and retriggering the 555.
Try placing a diode in series with the MOSFET gate (cathode to gate) with a 10kΩ resistor from the gate to ground.

Tried but not yet solved; the downspike also occurs on the 5V rail: straight (0.5ms) down, then 15ms recharge to 5V (due to C5 470uF).

Replaced C1 with 100nF, same problem.

Do you have any other ideas? I feel stuck at the moment..

 

[earckens]

I finaly replaced C1 with 22uF and it works. It seems as if the current draw at the moment the solenoid is being fed by the 7805 momentarily is so high that it spikes down. Since the pulse on the optocoupler input will be long enough to overcome the large C1R2 time a 22uF value for C1 seems ok. I now have to optimize C1 so that its value is reduced to the point where no triggering will be cause by the 7805 downspike.

The solenoid draws more almost 3A at 12V, when fed with 5V this drops to 500mA, so this circuit is worth it.

 

[earckens]

C1 at 22uF was so high that R3C2 did not matter anymore and pin 3 would drop when C1R2 became high enough for Q1 to drop again, after 4 seconds. So I reduced C1 to 2.2uF enough for the spike on the 5V rail not to impact, and C2 at 10nF. Pin 3 definitely drops again after 400ms. Good enough I think.

 

[AnalogKid]

There is no decoupling on the 555. Add a 0.1 uF ceramic and something in the 1 uF - 22 uF range electrolytic in parallel across the power and GND pins, as close in as possible with short leads.

ak

 

[earckens]

C1 at 22uF was so high that R3C2 did not matter anymore and pin 3 would drop when C1R2 bacame high enough for Q1 to drop again. So I reduced C1 to 2.2uF enough for the spike on the 5V rail not to impact, and C2 at 10nF. Pin 3 definitely drops again after 400ms. Good enough I think.

There is no decoupling on the 555. Add a 0.1 uF ceramic and something in the 1 uF - 22 uF range electrolytic in parallel across the power and GND pins, as close in as possible with short leads.

ak

Hi, there already is a 470uF and a 100nF decoupling cap right near the 555.
I am afraid the transition from 12V to the 5V 7805 is so immediate that this IC instantaneously cannot cope and for a fraction of time goes into short circuit protection.