V/f profile and current drawing

[abfer]

1- How does the current change with a 3ph induction motor which is drived in v/f principle and unloaded?

2- I'm trying to determine a v/f profile for a delta connected motor. But i couldn't be able to find a good profile. Due to dead time it's not good at very low frequencies. As frequency goes up it smoothes but at some frequencies it begins to vibrate. If i increase voltage it begins to stabilize but current also rises and a noise sounds up like it's forced. If i decrease voltage at the same frequency, it stabilizes again but torque output reduces dramatically.

3-Sometimes when i increase the frequency by 1 Hz, it suddenly stucks and continues to revolve.

What may be the reason of these? How should i create a v/f profile? All my tests are held while the motor is unloaded by the way.

 

[waross]

Calculate the volts per hertz ratio from the nameplate on the motor. If you allow the Hertz to drop much below this ratio you will saturate the magnetic circuit of motor and current and heating will increase disproportionately.
In other words, don't drop the frequency without dropping the RMF voltage in the same ratio.
This is the simplified basics. I am sure you will get a lot of other advice on fine tuning your performance.
yours

 

[Marke]

Hello abfer

At line frequency, you can consider that the unloaded impedance of an induction motor is primarily inductive, being the magnetising current of the motor. If you vary the voltage and keep the V/Hz constant, the current will also remain constant.

As you reduce the frequency, the ractive impedance also reduces. The impedance of the motor seen by the drive is in fact a combination of reactive impedance and resistive impedance (due to the resistance of the winding) At a low frequency, the resistive component of the impedance becomes sufficiently significant that it begins to influence the impedance as seen bey the drive and reduces the current draw if the V/Hz is maintainted. It is therefore necessary to deviate from the V/Hz straight line at low frequencies in order to compensate for the winding resistance. The actual impedance is motor design dependant so there is no magic curve that can be applied to all motors. Each motor has a different curve.

If you use a straight line V/Hz curve with no resitive compensation at low speed, you will reduce the magnetising current. This will reduce the torque capacity.
At low frequencies, you should not have dead band issues as the current should continue to flow and be sinusoidal. If you are haveing dead band issues, it sounds like your sinewave generator is not creating a sinewave and you need to address this. Have a look at the current waveform into the motor.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[abfer]

Yes my current waveform is distorted at low frequencies. But it smoothes as frequency and voltage goes up. So according to your say, i can create profile by observing the current. In my situation i'm trying to keep the v/f constant but as frequency goes up, also current goes up. Right?

 

[waross]

A small detail, As the frequency goes up, you allow the volts to go up. The increased voltage causes the increase in current. If you increase the frequency and hold the voltage constant, the increased inductance may lower the current.
yours

 

[Marke]

Hello abfer

If you keep your V/Hz constant, then as you increase the frfequency, you increase your voltage and the current will remain constant. This is the objective of the V/Hz. Keep the flux in the iron constant.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[abfer]

Ok, i think i got some feel. Increasing impedance has an effect on the current but also there is speed voltage which affects the current. So as slip changes may be straightness doesn't work?

 

[abfer]

I also want to ask you about the shock or very short duration stuck problem? Is 1 hz too much to step? Should i lower the steps between frequencies? Is this shock due to this? Also why does it vibrate?

 

[CJCPE]

The best v/f results in the same value of slip rpm at all speeds for the same torque. That is, if there is 50 rpm of slip at 100% speed and 100% torque then there should be 50 rpm of slip at 50% speed and 100% torque. Slip rpm should change only when the load torque changes, not when the frequency changes.

The vibration you mentioned in item 2 and the sudden problem described in point 3 may be due to some problem with the inverter. What is the carrier frequency? Are there sudden changes in the waveform at various frequencies. Can you describe point 3 in a different way, so that we can better understand what you mean?

 

[CJCPE]

The shock problem could be due to to the 1 Hz step. Inverters are usually designed for very small frequency changes. A 1 hz step would result in a step change in slip of about 60%. The motor would produce about 60% of rated torque in response.

 

[abfer]

Carrier frequency is ~7KHz. Shock problem occurs when i step from one frequency to another. Steps are 1Hz. Motor suddenly jumps or bounces sometimes. But not always at the same frequency, not every time but often enough to bother. I also couldn't keep current constant or near constant. It always goes up as frequency/voltage goes up even when i apply lower voltages.

 

[CJCPE]

When you say you step from one frequency to another do you mean that the output waveform suddenly changes frequency or do you mean that you make a step change to a speed reference input to an acceleration control circuit? Any kind of sudden change in the output waveform can cause a motor torque pulse that would cause the motor to jump if it is not firmly held in place.

How much does the current increase when the frequency increases?

 

[abfer]

Yes output waveform suddenly changes frequency like from 10Hz to 11Hz. Test motor is 3HP. According to the voltage it starts from as low as below 0.4A or 0.8A and goes up beyond 2A and above. I didn't go until the rated frequency or near it. Because it's current goes up and rotation gets worse.

 

[CJCPE]

Are you working with an experimental inverter? Commercial product variable frequency drive inverters do not permit sudden frequency changes. The microprocessor makes very small adjustments under control of a timed ramp function. The step change must be part of your problem. The random nature of the jumping may be related to random position of the magnetic field in the motor compared to change in phase of applied voltage when the step occurs.

What is the rated voltage and frequency of the motor? When you say the rotation gets worse, do you mean it is not smooth? Is there a lot of current variation when the frequency is constant? Are the three phase currents equal?

As Marke suggested, it seems like you may have a waveform problem. There might be some DC offset.

 

[Marke]

Hello abfer

It is usual to slide the frequency rather than step it. If you have any loss of output or loss of synchronisation in the output when you change frequency, you will get a very high torque transient. It is very important that the sinewave curent continues to flow without any interruption or phase change. I would sugest that you need to look hard at your waveform generation.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[CJCPE]

If you have a 230 volt, 3 hp motor, the no-load current should probably be between 2 and 4 amps. The current increase that you have described may be due to v/f too low to overcome the winding resistance at low frequency and approaching normal v/f and normal current as frequency increases.

The problem with rough rotation noise at higher frequency may be due to a waveform problem.

 

[abfer]

Yes 230V, 3HP , delta connected squirrel cage motor. And yes inverter is an experimental one. I'll check the waveforms then i'll inform you. Actually applied voltage seems good. I only checked one phase's current until now. It was distorted at very low frequencies, then it becomes sinusoidal as frequency/voltage goes up. And when i step to a higher frequency, it's distorted again during the step and stabilizes again. And if i remember right, when motor vibrates the current waveform has some changes like it's amplitude is lower in some periods than others. And it was periodic i think. I mean, in one period it's in full magnitude, the other it's not. Anyway i'll reduce steps and check the the 3 phase waveform. Then i'll inform you.

 

[CJCPE]

I believe that you can have DC bus voltage and output current amplitude oscillations if the input power source impedance is too high and the bus capacitor value is too low.

With no load, you should be able to get the motor to run well with constant v/f through the whole frequency range. Just apply rated voltage/rated frequency. You should be able to fully load the motor in the upper part of the frequency range. When you want to load the motor at low frequency, you will need to determine how much v/f increase is necessary to produce torque.

 

[Skogsgurra]

Abfer,

I have followed your "struggle" with some interest and I cannot really understand what kind of work you are involved in.

Is it:

1 Development of a new inverter?

2 A school project?

3 Learning by doing?

4 Building an inverter for a special purpose?

5 Other?

Eng-tips doesn't do complete design work. Nor is it acceptable to help students with things they should be able to find out if they read their books and go to classes. And - if you are designing an inverter commercially - you may find that you are on deep water and may need specialised help.


Gunnar Englund

http://www.gke.org

 

[abfer]

At last i put the things on the way. I had several problems.

1-Vibration problem was due to unsufficient voltage.
2-Low speed problem was due to unsufficient voltage but mainly dead time effect which cause a bad sinus waveform.
3-Stuck or shock problem was due to large steps.

And i'd discovered that i made a mistake which overheads the cpu.

I corrected the software and reduced the steps from 1Hz to 0.1Hz. And create a straight V/F profile. With these corrections, I've been able to go from zero speed to rated speed without a major problem. Only at 1Hz and 2Hz, motor didn't revolve.

But i also realized that 0.1Hz is still a big change in the output and the speed change can easily be noticed.

Nevermind, i reduced the steps to 0.001Hz now. I'll use straight V/F profile again. And i'll add an acceleration/deceleration algorithm. Afer i'll manipulate the V/F profile and i'll try to achieve a reasonable operation.


To skogsgurra:

I don't want anybody to do homework or other job. Only i want to learn something. I'm trying to make an inverter. I don't know whether i'll use it or not. My experience and education wasn't on power electronics or electrical machines. I only got the basic knowledge. I turned that way later. So with only theory i may not be able to do something or get it done correctly. To be sure that i'm doing the things right, i'm asking here. Am i doing something wrong?