PWM for controlling liquid flow through a Solenoid Valve

[kdv1988]

Hello guys, I am looking to control the flowrate of water through a Solenoid Valve in a continuous closed-loop process.

I inquired about Proportional Flow Solenoid Valves and these are a little too expensive and only a handful of manufacturers are supplying those.

But I looked online for some PWM Controllers and found a few, but I don't really know much about this topic as this field is completely new to me (I am a machine designer by profession). I've come here to ask you guys for some guidance

So the application is to control the flow-rate using PWM. I was told that you need to purchase Solenoid Valves with PWM-specific coils and a Controller to convert the 0-20mA signal from the PLC into PWM signals.

How do I go about this? And could you'll advise which PWM Controller would work for an application like this?

Thanks,
KV

 

[kdv1988]

More info :-

I have to introduce a coolant, flow-rates ranging from 500ml to 2000ml into my system where it has to mix downstream with a flow of water coming under gravity. The water shall be introduced into the system under gravity or through the plant's water supply line, due to which the flow-rate might fluctuate.

I need to control the flow of coolant to effectively match that of water in real-time so that my final mixture doesn't get affected with an accuracy of approx +- 0.5%.

Hope this info shall suffice. Feel free to let me know if you need any more details & parameters.

 

[TugboatEng]

Any DC coil should work with PWM. The inrusb on AC could will cause it to oveeheat. This control scheme is very commonly used for displacement control of hydraulic pumps. I don't know that it provides precise control. It's normally used as a soft start/stop feature.

 

[kdv1988]

But what about the PWM Controller? I have been suggested to use this -

https://www.asco.com/en-gb/Pages/accessories-series-603.aspx

It's a tad on the expensive side. I was unable to find a similar alternative, but did see a lot of PWM controllers with manual Potentiometers.

 

[TugboatEng]

Are you using a PLC? You can use a digital out to drive a solid state relay that operates the solenoid. Otherwise, any light duty converter through a solid state relay will work.

 

[Compositepro]

If it takes 10 minutes to fill a tank then the tank will be 10% full after the valve has been open for 1 minute. Simple. This does require constant flow rate, which requires constant supply pressure. This can be done with a pressure regulator or by supply from a reservoir that is maintained at a constant level. Constant level can be maintained by a float valve or by an overflow weir.

 

[kdv1988]

@TugboatEng, Yes I am using a PLC.

What is a light duty converter?

 

[TugboatEng]

One that doesn't cost as much. You don't need an I to PWM converter capable of driving a solenoid if you put a solid state relay between the converter and your solenoid. If you really want cheap, a simple transistor can replace the SSR. You really don't need optical isolation if both circuits are running the same 24VDC.

However, your PLC's digital outputs are capable of generating a PWM signal. It may or may not have the power to drive a solenoid but the SSR can easily be used to amplify the signal from the PLC. My tugboats use this system to control our hydraulic steering pumps. It's an axial piston Rexroth A4VG pump. The PWM controls a pilot valve that positions the swash plate.

 

[kdv1988]

@TugboatEng
You don't need an I to PWM converter capable of driving a solenoid if you put a solid state relay between the converter and your solenoid
So I would need a converter or not? I am a little confused sorry.

Also if I use an SSR & a converter (or not), would a simple Direct Acting Solenoid valve running on 24VDC work?

 

[kdv1988]

Secondly, I have seen people make Ventilators using stepper motors in this current Pandemic. Please see the attached image.

There is an inlet port with a flow-sensor that controls the outlet port with a spring controlled mechanism.

Any comments on this? The image only shows the in & out ports and the spring mechanism.

 

[TugboatEng]

Now that you mention it, all PWM controlled solenoid valves I have seen are solenoid actuated spring return spool valves. The spring and the average force from the solenoid result in the valve position.

 

[kdv1988]

@TugboatEng Please respond to my previous 2 queries.

1. Would I need a converter if I use an SSR? I mean I would right? How else will an analog/digital signal be converted into PWM?
2. The Ventilator product using a Stepper motor. Will a similar tech work for my application?

 

[TugboatEng]

1. The digital outputs on a PLC can be pulsed to mimic PWM. If the digital outputs has the correct voltages sufficient current rating for your solenoid you can pulse the solenoid directly. In this case make sure to use a flyback diode on the solenoid to protect the PLC. If the PLC voltage/current is not sufficient the you simply use a SSR as an amplifier.

2. Sure it could work but if you're going the distance of using a variable speed motor then you might as well attach that motor to any off the shelf positive displacement pump.

 

[kdv1988]

@TugboatEng,

In the first case, would a normal Solenoid Valve with a 24VDC Coil work? Or does it require a special coil for PWM?

 

[itsmoked]

It's safe to say that if the valve is opening and closing at some PWM rate it will last a few hours - maybe. If your control keeps it always open but at some intermediate position -not engaging the seat- you should be fine. Running it with a PLC should allow you to find the acceptable operating region at the flow and pressure and material and to stay out of the unacceptable region.

A normal DC coil should work okay. There are cases where they will not be okay if they have duty cycle limits or are valves for set-back where a big current operates it and a small holding current keeps it operated. One of those types would not work.

You are at the point where you need to get some hardware maybe from ebay and try this out yourself. We can't tell you much more. Remember this is Eng-Tips not Eng-Do-All The-Engineering-For-You.

 

[kdv1988]

Of course. I don't mean for you'll to do the engineering & trials for me. But mainly guide me to proper selection of the equipment so as to prevent time & money loss if I select the wrong instrumentation.

Trials I shall take soon once I understand completely how to approach this application.

 

[moltenmetal]

There are two kinds of pulse-width modulation control for solenoid valves.

The obvious kind is actually in industry called "time proportional control", where the duty cycle (% of the time that the solenoid valve is OPEN) is used to control flowrate in average terms, even though the flow is intermittent in time. We've used that with great success for steam control to heating coils in tanks- works like a charm.

The other kind is pulse width modulation of the current to the solenoid coil. This CAN produce continuous modulation of valve POSITION and hence of flowrate in CERTAIN designs of solenoid valve. It is done in thermal massflow controllers such as those offered by Brooks, Bronkhorst etc. The resulting Cv vs % of coil current curve will be very complex and VERY hysteratic, meaning that it will not behave like a pneumatic control valve which does not have a stem positioner fitted- only WORSE. Is it useful- yes, it can be, but only for things that can tolerate sloppy, hysteratic control behavior on the final control element.

Will you manage to build your own arrangement that gives you adequate control of flow? Depends. How many man-hours to you have to screw around doing this?

Got $400? You can buy a PWM driver board and a solenoid valve which are matched and which give you decent control within fairly broad limits. Not great control- not PRECISE control, but decent control when used within their limits. You can get that from Burkert or Asco. That's by the way cheaper than an I/P converter to DRIVE a pneumatic control valve...

 

[moltenmetal]

Forgot to add:

- solenoid must be DIRECT ACTING type- this PWM drive of the coil is very unlikely to work well on a pilot-operated solenoid valve

- by "certain designs", I mean designs which don't use "snap" actions, but rather solenoid valves which look like conventional globe vales with a plug that rises off a seat and is forced back onto the seat by spring pressure

Good luck- it'll be fun, and won't save you any money unless you're making hundreds of them or this is a hobby project and your time is worth nothing