Driving 3V Solenoid with 6V

[Wallace Gasiewicz]

Put the Zener in SERIES with the coil. Not to Ground. Neg on zener to 6 V supply and other end to coil.

Another suggestion:
https://www.mouser.com/c/power/power-management-ics/ldo-voltage-regulators/?output%20voltage=3.3%20V

Or big one:

https://www.sparkfun.com/products/526

More one watt Zener this time 3 Volt

https://www.ebay.com/itm/254955329925?hash=item3b5c859985:g:NqMAAOSw2R9ghRw0&amdata=enc%3AAQAHAAAAoHPHQMNDQ6093PeLHLPd0EJBpJhNLxHGbWZDfaSoM%2BnvOZgyzFbzix1eIOo2lpCRtsZ%2F8eIHccijPWWqAvovjqlYcNosxseedPwH0o4KD8OIR4v%2F2UiQs%2BuXQIjkRwnDt2eaIl6LmAZtJdFVAwkFsVLPQoSuhVzLGqpLa3VwgzrAuwACVwCB8gtlMVjMUSKBs8YEWz5zSMpd6ypf%2BP4LoSk%3D%7Ctkp%3ABk9SR6anuoi2YQ

[Zero999]

Here is a one watt 3.3 V  SMD Zener:

https://www.ebay.com/itm/284996435482?hash=item425b1c7e1a:g:4bAAAOSwTcdiBt01&amdata=enc%3AAQAHAAAAoNHFyGuZc6wusIvQpSTlTa1MIa8lB6iRMlN0hJtXW8UlURJdyKYdbnY4hvIWRtbTeqK38iKPCBsuDePKr0kuETV7T3ej6oG8vme33ldcSwQM1N%2BKnX62ZumvZMcWsn%2FkQldDYtzkXfpheK7XIavb1U7fbnGbU6%2FYbFJsKUmvlMlX913QeTja%2Br9Vfrtxob2IbPsqJ4osrkjUJSDunMtrvVI%3D%7Ctkp%3ABk9SR_yY_oW2YQ

But I do not know where your solenoid is., maybe it would be easy to put a regular thu hole axial zener in the wire feeding it and cover the zener with shrink tube.

Except it won't be able to dissipate 1W, without putting it on a huge copper plane to act as a heat sink and even then it's optimistic.
https://www.mouser.co.uk/datasheet/2/427/zm4728a-1767906.pdf

I think I can also put it directly onto the SV with a shrink tube. Then I will just connect the Zener diode's anode to GND and the cathode to the other lead of the SV.
But I think this Zener diode increases my power consumption compared without using it.

That won't work, because the heat won't be able to escape, causing it to get too hot.

It doesn't matter whether a resistor, or zener diode, it will still have to dissipate the same amount of heat.

[elefurtronik]

I am now thinking to put a small linear voltage regulator to the source of my p-channel MOSFET to attenuate the Vdd voltage. This might work. My system will work maximum 1500ms and I don't think it is going to get too hot.

[Zero999]

I am now thinking to put a small linear voltage regulator to the source of my p-channel MOSFET to attenuate the Vdd voltage. This might work. My system will work maximum 1500ms and I don't think it is going to get too hot.

It doesn't matter. It will still have to dissipate the same amount of power as a resistor or zener diode. The fact you're only running this for 1.5 seconds (you should have mentioned that as the start of the thread) which will limit the heating. How often is it going to be operated? Just use a chunky 1W ceramic power resistor, which will be much more reliable than an active circuit such as a linear regulator.

[james_s]

Oh yeah, if it's only 1.5 seconds then a resistor is the way to go IMO, at least if the duty cycle is low. I was assuming much longer periods.

[elefurtronik]

I am now thinking to put a small linear voltage regulator to the source of my p-channel MOSFET to attenuate the Vdd voltage. This might work. My system will work maximum 1500ms and I don't think it is going to get too hot.

It doesn't matter. It will still have to dissipate the same amount of power as a resistor or zener diode. The fact you're only running this for 1.5 seconds (you should have mentioned that as the start of the thread) which will limit the heating. How often is it going to be operated? Just use a chunky 1W ceramic power resistor, which will be much more reliable than an active circuit such as a linear regulator.

I don't have space in my box for the chunky resistor. :/ It will operate 20 times a day.

Oh yeah, if it's only 1.5 seconds then a resistor is the way to go IMO, at least if the duty cycle is low. I was assuming much longer periods.

If I can control SV with PWM then it can be as much as lower. However, it is digital on/off for moment.

[james_s]

"Chunky" is relative. You can get a small surface mount resistor that is perfectly capable of dissipating a watt for 1.5 seconds at such a low duty cycle. You can typically overload a resistor quite severely if the duty cycle is low.

The key thing to remember is that absolutely any kind of linear dropper is going to dissipate the same amount of heat. Doesn't matter if you use a resistor, a zener, a transistor based regulator, a linear regulator IC, anything, the amount of heat is the same.

[Zero999]

I am now thinking to put a small linear voltage regulator to the source of my p-channel MOSFET to attenuate the Vdd voltage. This might work. My system will work maximum 1500ms and I don't think it is going to get too hot.

It doesn't matter. It will still have to dissipate the same amount of power as a resistor or zener diode. The fact you're only running this for 1.5 seconds (you should have mentioned that as the start of the thread) which will limit the heating. How often is it going to be operated? Just use a chunky 1W ceramic power resistor, which will be much more reliable than an active circuit such as a linear regulator.

I don't have space in my box for the chunky resistor. :/ It will operate 20 times a day.

Then use a smaller resistor and it will get hotter. It doesn't matter what device is used to dissipate the power, it will still produce the same amount of heat. You have a choice between a higher temperature and smaller component, or lower temperature and a larger component. It's just physics. P = V×I and the more power, the more heat. If you put 1W of heat into a tiny space, it will get much hotter, than 1W into a much larger volume.

1.5 seconds 20 times per day is a low duty cycle, which will limit the heating to some extent.

[Siwastaja]

PWM is the correct answer. You need the freewheeling diode and proper layout anyway, even if you switch once a year. So why not switch 10000 times per second?

With PWM, current varies in triangular waveform. Use high enough frequency, and force ripple becomes low enough not to matter (make dI = 10% of DC current or so. dI = difference between maximum and minimum current. Mechanical force is relative to current). With high enough frequency, the thing operates in CCM, meaning current is never zero (which is what you want to avoid force ripple, of course). Then the formula for duty cycle is easy: D = Vout/Vin. So in this case, simply 50%.

Efficiency is good, too. For even better, you might want to consider synchronous rectification because even with a schottky diode with say Vf=0.4V, efficiency will be diode-limited to (3-0.4)/3 = 87%. But that's not bad compared to the resistor idea.

In general, high-efficiency strategies help with miniaturization because you don't need the space for heat sinking / dissipation. Also eases the design. In this case, it's a total no-brainer as you need all the components there anyway. I'm assuming you are using microcontroller or similar which can easily produce the square wave.