# Some questions regarding Pump Soft Starters

Status
Not open for further replies.

#### baileychic

Some questions regarding Pump Soft Starters

I am designing a Pump Soft Starter using PIC16F877A and Snubberless 600V 40A type motor. I am going to use this with a 1.5 HP monoblock pump.

I have seen soft-starters designed using 2 anti-parallel SCRs. According to my understanding the load voltage is controlled by controlling the phase angle of the AC voltage and gradually (2 to 3 sec time) the load voltage is increased to max (say 230 or 240V if mains voltage is say 270V).

1) I want to know can I use single TRIAC instead of using 2 SCRs ?
2) What should be the ramp up delay duration in seconds ?
3) How to find out what is the initial phase angle required to start the pump based on HP rating ?

I want to run the motor even if mains voltage is 270V but only 230 to 240V has to be applied to motor. For this how to calculate load voltage based on Phase angle that needs to be applied at the end of the ramping ?

4) Can I measure the load voltage while ramping using ADC and control the final firing angle which maintains constant 230V for the motor/pump ?

#### KlausST

##### Super Moderator
Staff member
Hi,

I am designing a Pump Soft Starter using PIC16F877A and Snubberless 600V 40A type motor. I am going to use this with a 1.5 HP monoblock pump.
* 1 Phase or 3 Phase?
* how can this be: "600V 40A type motor" <--> "1.5 HP monoblock pump" ?

I want to know can I use single TRIAC instead of using 2 SCRs ?
Usually Yes. In detail it depends on your circuit and the control scheme.

What should be the ramp up delay duration in seconds ?
It depends on the inertia, the expected input current, ...

I expect a pump not having overly increased start up current (compared to the steady current). Are you usre you need a soft starter?

How to find out what is the initial phase angle required to start the pump based on HP rating ?
I don´t think it depends on pump HP rating. Why do you think it should be based on HP rating?

I want to run the motor even if mains voltage is 270V but only 230 to 240V has to be applied to motor. For this how to calculate load voltage based on Phase angle that needs to be applied at the end of the ramping ?
Spend some minutes with Excel and get this:

Can I measure the load voltage while ramping using ADC and control the final firing angle which maintains constant 230V for the motor/pump ?
Not that simple. But possible.
I´ve designed a measurement tool exactly for this. It is for 50Hz and calculates and outputs the exact true RMS value of V and I every fullwave.
(and P and I_avg and V_avg).
But you can try an energy metering IC. Not sure how fast the filters are and the how good they work with phase controlled AC...

Klaus

#### c_mitra

Snubberless 600V 40A type motor. I am going to use this with a 1.5 HP monoblock pump.
600V 40A motor is unlikely to be a single phase motor; the details of the motor are needed before you can plan a reliable soft start.

1.4HP monoblock pump is very likely a single phase induction motor but the nature of the load will decide the soft start time constant.

It is very important to start motors (except ones with less than 1KVA - perhaps) with a lower voltage and bring them up to speed before you apply the full voltage.

It will be messy if the motor is already loaded. You need to observe the motor current and increase the voltage so that the motor current does not become excessive.

(If no external load is present, consider the rotor itself as the load)

Its messy also for synchronous motors.

#### baileychic

Hi,

* 1 Phase or 3 Phase?
* how can this be: "600V 40A type motor" <--> "1.5 HP monoblock pump" ?
1 Phase
Sorry, it was 600V 40A snubberless TRIAC and 1.5 HP Single phase 230V 50 Hz AC Monoblock pump.

I expect a pump not having overly increased start up current (compared to the steady current). Are you ssre you need a soft starter?
Yes

I don´t think it depends on pump HP rating. Why do you think it should be based on HP rating?
Based on HP starting and running currents will come into consideration and we are controlling load voltage which also controls current as resistance and inductance of pump motor winding is fixed. So, Phase angle will control the voltage and also load current (especially startup current). This is my analysis. Correct me if I am wrong.

Spend some minutes with Excel and get this:

Not that simple. But possible.
I´ve designed a measurement tool exactly for this. It is for 50Hz and calculates and outputs the exact true RMS value of V and I every fullwave.
(and P and I_avg and V_avg).
But you can try an energy metering IC. Not sure how fast the filters are and the how good they work with phase controlled AC...

Klaus
Okay.

@c_mitra

Okay.

Actually in India nobody manufactures a single product which includes

Pump Starter
Automatic Water Level Controller (AWLC) and
Pressure Pump control.

Sometimes when AWLC fails working then Overhead tank water level goes low and pressure pump turns ON continuously to maintain pressure but it fails to do so and finally the pressure pump gets damaged and so I want to make a single product which includes all the above three functionalities in one unit.

Last edited:

#### KlausST

##### Super Moderator
Staff member
Hi,

Based on HP starting and running currents will come into consideration and we are controlling load voltage which also controls current as resistance and inductance of pump motor winding is fixed. So, Phase angle will control the voltage and also load current (especially startup current). This is my analysis. Correct me if I am wrong.
How to find out what is the initial phase angle required to start the pump based on HP rating ?
I understand this: phase_angle <--> pump_HP_rating

Example:
I don´t think that one can say "a 1.5kW pump needs 60° phase angle" and a "2.5kW pump needs 70° phase angle"

* The phase control has a known phase_angle to RMS_voltage relation (see chart)
* The motor will have a current to torque relation (maybe the motor datasheet can tell you)
* The pump will have an RMS_voltage to RPM relation .. but this will depend on torque (maybe the pump datasheet will tell you)
* thus the pump will have a phase_angle to RPM relation ... and a phase_angle to current relation .. but for us it´s impossible to tell you.
--> I assume you are the only one that is able to collect the data to find out. Either by using the known operation conditions ... or by test and measure...

Klaus

#### FvM

##### Super Moderator
Staff member
In addition to what's already said, I would doubt that Vrms is the quantity of interest. I think, it's rather magnitude of the fundamental. Harmonic components cause less current due to the inductive motor impedance.

I also wonder if soft start is an actual requirement for water pumps. I never heard about it.

#### c_mitra

I also wonder if soft start is an actual requirement for water pumps. I never heard about it.
We have deep bore wells (100m is common; 200m is not unheard of) that use submersible pumps that use 3 phase motors.

The starter uses a star-delta manual start which is a simple form of soft start. I have not seen the innards though.

However, for single phase monoblock pumps with less than 1KVA ratings do not need or use any soft start features (but I think personally that they can well use a good electronic starter).

In our previous house we had a 1/2 HP monoblock pump that was used to lift water to the 2nd floor tank. Lights used to go dim for a fraction of a second every time the pump was started.

#### baileychic

Hi,
Not that simple. But possible.
I´ve designed a measurement tool exactly for this. It is for 50Hz and calculates and outputs the exact true RMS value of V and I every fullwave.
(and P and I_avg and V_avg).
But you can try an energy metering IC. Not sure how fast the filters are and the how good they work with phase controlled AC...

Klaus
You said measuring phase controlled load voltage is not that simple.

How am I getting correct mains and load voltage with simple circuit and code ?

Find complete unfinished project.

See the simulation result.

In the Lcd display

MV = mains voltage
CA = Conduction angle (incorrect as triac is triggered from external signal and not code)

Interrupt code has some bug and I am trying to fix it. ADC code is clean and working.

In code 5V adc input = 270V AC voltage

Please don't change the POT values. It has been set properly to give 270V AC voltage readings for 5V DC inputs.

#### Attachments

• 72.4 KB Views: 1
• 142.3 KB Views: 5
Last edited:

#### FvM

##### Super Moderator
Staff member
What makes you think that RMS is voltage is the quantity you want to measure and control? This is motor rather than lighting or heater control.

#### baileychic

@FvM

I have to control two things. The startup current of pump and for that I will use ACS758-50 (50A type hall effect sensor). By monitoring the initial current I will gradually ramp up the load voltage within 2 to 3 seconds so that the pump doesn't draw 6 to 7 times of nominal current. For this I have to do some tests on already available pump starters (soft starters).

I am designing soft starters for upto 2 HP pumps both mono block and submersible pumps.

I think RMS load voltage has to be measured because for the other purpose that is to maintain constant 230V to load after pump start if mains voltage is greater than 230V.

- - - Updated - - -

Got the triac firing signals. Bug was in this piece of ISR code.

Code:
if((TMR1IE_bit) && (TMR1IF_bit)) {
TMR1IF_bit = 0;
TMR1ON_bit = 0;
TRIAC_GATE_TRIGGER = TRIAC_GATE_ON;
InitTimer2();
}

Code:
Timer2_Init();
in the above ISR code I had just used

Code:
TMR2ON_bit = 1;
Find attached new code. I still have to code the ramping to get constant 230V if mains voltage is greater than 230V.

#### Attachments

• 78.1 KB Views: 0
• 157.7 KB Views: 1
Last edited:

#### baileychic

It is somewhat working.

If mains voltage is greater than 230V and less than or eual to 270V then load voltage is maintained fixed at 230V. The load voltage is fluctuating between two values after reaching almost stable value that is between 229 and 230V.

Can somebody tell me how to improve the code and get stable load voltage.

Regarding Pump current I have read that startup current will be 6 to 7 times the nominal current. Is this right ? If yes, then

1.5 HP pump = 756W * 1.5 = 1134W

Irms = 1134W / 230V = 4.93 = ~5A

startup current will be 5 * 7 = 35A

So, I have to slowly ramp the load voltage so that the startup load current will be say 3 times the nominal voltage that is 3 * 5 = 15A and then gradually ramp up the load voltage up to 230V within 3 to 4 seconds time during which the load current becomes nominal current of 5A.

Can this be efficiently implemented ?

Here is the current simulation result.

Proteus 8.8 SP1 is needed to simulate Proteus file. If you don't have Proteus 8.8 SP1 and wants to simulate then you can request me for Proteus 7.10 SP1 version Proteus Simulation file so that I can redraw the circuit in Proteus 7.10 SP1 and post it here.

Find the attached project. Experts tell me how to improve the coding.

#### Attachments

• 83.9 KB Views: 0
• 149.2 KB Views: 3

#### c_mitra

because for the other purpose that is to maintain constant 230V to load after pump start if mains voltage is greater than 230V.
Interesting idea. But motors, like LEDs, are basically current devices and it should be prudent to monitor the current rather than the voltage.

Most motors get damaged because of low voltage; they tolerate high voltage rather well.

You should be able to detect stalling (most motors burn out just after stalling) by looking at the current profile.

Most decent motor rated for 230V will run happily ever even at 270V (the current will increase only a little)

For 3 phase motors, phase faults are also a silent killer.

#### KlausST

##### Super Moderator
Staff member
Hi,

I wonder:
What makes you think that RMS is voltage is the quantity you want to measure and control? This is motor rather than lighting or heater control.
I want to run the motor even if mains voltage is 270V but only 230 to 240V has to be applied to motor. For this how to calculate load voltage based on Phase angle that needs to be applied at the end of the ramping ?
So what else "motor voltage" if not "RMS voltage"?

****
Most motors get damaged because of low voltage;
I know that
* DC (overcurrent, overheating)
* missing a phase (out of three. Causing overcurrent, no RPM --> reduced ventilation)
* overcurrent (overheating)
* missing or reduced ventilation (overheating)

****
The load voltage is fluctuating between two values after reaching almost stable value that is between 229 and 230V.
Can somebody tell me how to improve the code and get stable load voltage.
This 1V fluctuation should be more than sufficient.
Mains voltage will fluctuate way more than this and the motor will easily withstand this.

You are concerned about 1V fluctuation. But you measure not RMS but something close to "rectified average".
Because of math you can't get true RMS (especially phase controlled) with your measurement circuit. And I don´t think it is necessary.
Thus - even if you show "230V" on the display - the motor will not behave like at 230V RMS sinewave.

***
I don´t want to say that your solution is bad - I rather think that is an appropriate solution for a pump.
But your requirement for very precise motor and stable voltage can not be achieved with the shown circuit.

--> be satisfied with your solution.

***
But I see one problem. The zero cross circuit is not symmetric. Thus it is expectable that your ignition timing is not symmetric, too. This may cause DC to the motor.
Increased current with increased heating may be the result. But it depends on motor type.

Klaus

#### c_mitra

Well, well. If the voltage falls below a level so that the motor stalls, the back emf disappears and the large current flows (but not enough to start the motor) and the motor dies because of overheating.

This is common for motors with high loads on starting; most common starters, the star-delta types, do apply low voltage during start up. That means the voltage must fall low enough so that it is not able to produce enough torque to start the motor with a load.

A stalled motor can take a lot of current. No, zero V will not do!!!

You have never heard about this because in western countries the line voltage is protected by overvoltage and undervoltage cutoffs; but major part of the world's population lives where these luxuries are non-existent.

Yes, another killer is the phase fault (one phase is lost) and the motor fails to produce enough power and stalls.

Status
Not open for further replies.