Button press (input pulse) output alternating voltage pulse

[Patwong]

Appreciate some help here to design a circuit
I would like to control a pulse solenoid valve. The pulse solenoid works by positive input voltage pulse, the valve open, reverse the voltage pulse, the valve will close. then no voltage is applied, the valve state remain as it was.

I would like the circuitry to be simple and consume only power when the button is press ( so not microcontroller or arduino here). The preferred power source is a 18650 battery (3.7v). 1 battery or at the most 2 battery to power the circuitry. When the button is release, the whole circuitry is open and therefore no power is consumed.

When I press a button, the positive output voltage will be generated (can be a pulse or so long as the button is press, then output voltage is maintain). when the button is release, there is no voltage applied. When the button is pressed again, a negative output voltage will be generated.

Below is what I am trying to achieve

 

[KlausST]

Hi,

Two extremely important informations are missing:
* timing of the pulses
* expectable current (waveform)
... urgent to decide how to control the valve.

Some remark about not using a microcontroller:
* if you want a positive , then a negative , then a positive ... pulse --> then surely you need something to "memorize" the last pulse direction. If this is the job of "the electronics", then you can´t power OFF it completely. It needs a supply .... in most cases.
* current consumption of a modern microcontroller may be reduced by some power saving modes. Usually there always is some kind of power saving mode, which means a current consumption below some microamperes.

If I had to design it, I´d surely go with a microcontroller. It could be a simple 8 pin one using the internal oscillator, thus you have low part count, while being very flexible with software.


Klaus

 

[betwixt]

Not sure whats available on the market at the moment with enough relay contacts but it can be done with a latching relay alone if the contacts are wired to reverse the coil polarity. The pulse has to be short enough that it doesn't make a mechanical oscillator!

Brian.

 

[BradtheRad]

Link to discussion how to make toggle circuits (with schematics) at an electronics website:

electronics.stackexchange.com/questions/4060/make-a-momentary-switch-control-a-toggle

Flip-flops are constructed so the Q and bar-Q can send bipolar waveforms. Then a series CR shortens these pulses. However the pulses may be too weak to energize a latching relay. You may need to amplify the signals through an H-bridge.

 

[Patwong]

Hi,

Two extremely important informations are missing:
* timing of the pulses
* expectable current (waveform)
... urgent to decide how to control the valve.

Some remark about not using a microcontroller:
* if you want a positive , then a negative , then a positive ... pulse --> then surely you need something to "memorize" the last pulse direction. If this is the job of "the electronics", then you can´t power OFF it completely. It needs a supply .... in most cases.
* current consumption of a modern microcontroller may be reduced by some power saving modes. Usually there always is some kind of power saving mode, which means a current consumption below some microamperes.

If I had to design it, I´d surely go with a microcontroller. It could be a simple 8 pin one using the internal oscillator, thus you have low part count, while being very flexible with software.


Klaus

I dont really have a specific on the timing of pulse since the data sheet did not specify, my guess would be 500ms. A close to square wave form will do.
would definitely dont want a microcontroller.

 

[KlausST]

Hi,

I´m quite sure the datasheet specifies switching time as well as current.

I´ll be back when there is useful information to go on.

Klaus

 

[crutschow]

Without knowing the coil current, it's not possible to design an optimum circuit to do what you want.

 

[danadakk]

I dont really have a specific on the timing of pulse since the data sheet did not specify, my guess would be 500ms. A close to square wave form will do.
would definitely dont want a microcontroller.

Whats the issue in your mind using a micro controller ? Is it programming you do not know how
to do ?

Regards, Dana.

 

[D.A.(Tony)Stewart]

My immediate answer will use both leads of the solenoid with a full bridge or two half-bridge circuits to drive the bipolar output.

Choices depend on your V, I, DCR, t or power and $ budget specs.

Most likely P and N type transistors for momentary switching, rather than dual N type which need a low-side charge pump to drive the high side gate. Bridge ICs might cost much more than discrete logic-level FETs.

The 1-shot timer may be chosen in many, simple ways with a defined time constant and include the D FF toggle circuit which has a very low static current.

--- Updated Nov 20, 2023 ---

proof of concept.

The depressing the switch changes direction starts the RC timer but the 1-shot is ~ 0.5 s (56nF * 10Meg) starts when the button is released.
Solenoid resistance is estimated at 300 mohms and 100 mH inductance.

 

[D.A.(Tony)Stewart]

Disregard diodes which are redundant.