Running AC solenoid vibration pump with DC voltage

Thread Starter

Amir Mohammad Soltanpour

Joined Sep 10, 2017
23
Ulka e series solenoid vibration pump works with Ac voltage directly connected to main power.
They are rated at 220, 110, 24 voltage 50-60 Hz.
These pumps use an internal diode for removing half cycle of AC voltage.
The manufacturer specification does not determine the voltage as pure sine ac voltage.

Full specification of Ulka e series vibration pumps : http://ulka-ceme.co.uk/Ulka_E_Models.html

The question is that can these kind of pumps be running with square wave DC voltage controlled with pwm at 50-60 Hz without problem?
Will it damage the pump ? or reduce its lifetime?
 

MisterBill2

Joined Jan 23, 2018
8,016
Quite often directly driven vibration systems run in a resonant mode, which gives a lot more motion than otherwise, because of the resonance. So if your DC application has a different frequency the pump may not deliver nearly as much flow as it would running on chopped mains AC power. And mechanical resonances are often fairly sharply peaked, so you will probably need to set the drive frequency exactly.
 

MisterBill2

Joined Jan 23, 2018
8,016
Probably if you use a a series power transistor you can make it work with a pulsed waveform. You will need to adjust the frequency and the duty cycle to get the best performance. But if you have the AC voltage available why would you use DC? It will be more complex and less efficient.
 

Alec_t

Joined Sep 17, 2013
11,900
The question is that can these kind of pumps be running with square wave DC voltage controlled with pwm at 50-60 Hz without problem?
What would you control using pwm? As stated above, the pump may be need to operate at a specific resonant frequency. You might also find that below a certain pulse amplitude the pump's internal valves wouldn't open/close properly
 

MisterBill2

Joined Jan 23, 2018
8,016
There will need to be a quite stable frequency control, and the on time will need to be less than 50% because of the drive being a square wave instead of a sine wave. The big deal will be to hold it at the resonant frequency.
 

kaindub

Joined Oct 28, 2019
53
These pumps do not run in resonant mode. You can think of them as being a DC coil that has a 50hz 50% duty cycle signal applied to the solenoid inside.
if you apply an AC voltage of a different frequency the pump will pump more or less depending on the frequency. But in order to keep the magnetic circuit from saturating the voltage would have to vary in proportion to the frequency. this is somewhat ifferent to PWM.
For the reason stated above, the pump probably would not work on a PWM signal as the square waveform would saturate the magnetics and the coil would overheat.
You could vary the output by just applying pulses of AC to the pump. It a basic form of PWM but the frame rate would be in the seconds if not tens of seconds
 

Thread Starter

Amir Mohammad Soltanpour

Joined Sep 10, 2017
23
These pumps do not run in resonant mode. You can think of them as being a DC coil that has a 50hz 50% duty cycle signal applied to the solenoid inside.
if you apply an AC voltage of a different frequency the pump will pump more or less depending on the frequency. But in order to keep the magnetic circuit from saturating the voltage would have to vary in proportion to the frequency. this is somewhat ifferent to PWM.
For the reason stated above, the pump probably would not work on a PWM signal as the square waveform would saturate the magnetics and the coil would overheat.
You could vary the output by just applying pulses of AC to the pump. It a basic form of PWM but the frame rate would be in the seconds if not tens of seconds
The point of this thread is to understand that are these bumps sensitive to pure sine wave or will square wave voltage make any effect on their output or life time.
They are rated at AC voltage and as we know main power of buildings is pure sine AC voltage.
I'm trying to use these pumps with DC voltage coming from backup batteries and definitely i don't want to design an expensive and sophisticated pure sine wave inverter.
So the idea is to take DC voltage from batteries and feed it through a single MOSFET or IGBT and control the gate with a microcontroller and make the output voltage exactly at 50Hz or to what ever these pump are rated at.
These pump use built in diode to cut half of ac wave in each cycle of 50Hz, in this way they are able to move piston in one direction and then with a force from a spring to other direction and pump a small amount of liquid in each cycle.

So what if we take DC voltage and set it at 50Hz ( 10ms on, 10ms off) and feed it to these pumps ?
Will they work properly ? No side effect ? or Because of square wave we need to reduce on time or voltage ?
 
Last edited:

Thread Starter

Amir Mohammad Soltanpour

Joined Sep 10, 2017
23
What would you control using pwm? As stated above, the pump may be need to operate at a specific resonant frequency. You might also find that below a certain pulse amplitude the pump's internal valves wouldn't open/close properly
Using single MOSFET of IGBT and Turn gate 10ms on and 10ms off should make the pump think this is an AC voltage 50Hz and it will think that the internal diode is causing the 10ms off.

The question is, what is the effect of the square waveform instead of the pure sine wave?
 

kaindub

Joined Oct 28, 2019
53
The point of this thread is to understand that are these bumps sensitive to pure sine wave or will square wave voltage make any effect on their output or life time.
They are rated at AC voltage and as we know main power of buildings is pure sine AC voltage.
I'm trying to use these pumps with DC voltage coming from backup batteries and definitely i don't want to design an expensive and sophisticated pure sine wave inverter.
So the idea is to take DC voltage from batteries and feed it through a single MOSFET or IGBT and control the gate with a microcontroller and make the output voltage exactly at 50Hz or to what ever these pump are rated at.
These pump use built in diode to cut half of ac wave in each cycle of 50Hz, in this way they are able to move piston in one direction and then with a force from a spring to other direction and pump a small amount of liquid in each cycle.

So what if we take DC voltage and set it at 50Hz ( 10ms on, 10ms off) and feed it to these pumps ?
Will they work properly ? No side effect ? or Because of square wave we need to reduce on time or voltage ?
Did my reply not make it clear? It won't work on a square wave dc signal because the magnetic circuit of the internal solenoid will saturate. Im not talking out of my bum. I spent a large part of my career delving into all sorts of devices with magnetics ie motors, transformers and relays. Once you saturate the magnetic circuit with too much flux created by the applied voltage it just looks like a short circuit. You can't use less voltage Because you don't create the initial flux to drive the solenoid.
Build your circuit( a 555 timer and a fet are all you need initially) and tell me the results
 

Thread Starter

Amir Mohammad Soltanpour

Joined Sep 10, 2017
23
Probably if you use a a series power transistor you can make it work with a pulsed waveform. You will need to adjust the frequency and the duty cycle to get the best performance. But if you have the AC voltage available why would you use DC? It will be more complex and less efficient.
I want to use them both with main AC voltage and DC backup batteries.

Also one of these pump are rated at 24 volt 50Hz, So i'm start to thinking why the manufacturer build such a pump ? i mean which part of the world use this voltage ?

So is it possible to using a transformer to make 230 or 110 volt to 24 volt for general use and use 24 volt from batteries for backup?
 

MisterBill2

Joined Jan 23, 2018
8,016
These pumps do not run in resonant mode. You can think of them as being a DC coil that has a 50hz 50% duty cycle signal applied to the solenoid inside.
if you apply an AC voltage of a different frequency the pump will pump more or less depending on the frequency. But in order to keep the magnetic circuit from saturating the voltage would have to vary in proportion to the frequency. this is somewhat ifferent to PWM.
For the reason stated above, the pump probably would not work on a PWM signal as the square waveform would saturate the magnetics and the coil would overheat.
You could vary the output by just applying pulses of AC to the pump. It a basic form of PWM but the frame rate would be in the seconds if not tens of seconds
OF COURSE the duty cycle would need to be quite a bit less than 50%, that has already been stated.
 

MisterBill2

Joined Jan 23, 2018
8,016
I want to use them both with main AC voltage and DC backup batteries.

Also one of these pump are rated at 24 volt 50Hz, So i'm start to thinking why the manufacturer build such a pump ? i mean which part of the world use this voltage ?

So is it possible to using a transformer to make 230 or 110 volt to 24 volt for general use and use 24 volt from batteries for backup?
OK, with this comment about wanting a battery backup the whole question changes and the picture is far different.
LOTS of areas use 24 volts, NOT for distribution but to run things on a voltage that does not pose a shock hazard.
Certainly it is possible to use a transformer to provide the 24 volts AC to run the pump, that is how it is intended to be done. BUT vibration pumps will not run on simple DC from batteries. You would need a circuit similar to a PWM system to delver quite short pulses of DC at the correct frequency.
 

Bernard

Joined Aug 7, 2008
5,744
I would buy the 24 V version as we can then talk about a driver as suggested by kaindub, , 555
with variable duty cycle, not strictly PWM, probably at about 30%. Input will be polarity sensitive
unless diode is removed.
 
Top