Brushless Dc motor does not rotate without hand force

[gsecer]

bldc motor snubber

Hi everybody

I have been looking at this forum time to time but never really needed for a help.

I have built up BLDCM driver circuit which is just like classical structure. I am using IR2110 Mosfet gate drivers.

I have two problems.

1) Motor does not start rotating without hand help.

2) When i apply %100 duty cycle, it rotates at around 13500 rpm while its manual says it should be at around 12000. Moreover, when i apply %50 duty, it rotates around 10000. it behaves like it is non linear.

BTW, i have to say that there is no load connected to motor.

What may be the problem guyz?

I have dealt with this stuff since 2 weeks and i am totally messed up. I have built up this circuit 5 or 6 times.

Any help or suggestion is appreciated.

 

[FvM]

dead time ir2110 bldc

I have built up BLDCM driver circuit which is just like classical structure.

You may want to be somewhat more specific. A classical BLDC should be expected to involve a rotating field and provide a torque at zero rotation speed. Your design is apparently different.

 

[gsecer]

stray losses bldc

What i have meant by classical structure is that;

6 mosfet and 3 ir2110 which are expected to drive them.

Connection of ir2110s is as it is exhibited in its datasheet. I am using two 220nF ceramic capacitors parallel as bootstrap capacitor and UF4007 as bootstrap diode. When i examine my ciurcit i dont see any misconnection or shortcircuit.

Added after 1 hours 27 minutes:

Another problem that i found is a weird but expected issue. When i connect a load to motor and increase load, motor handles it until load reaches a value such that motor cannot handle. At this value of load, motor stops and even if i remove the load, it does not start to rotate again.

Please help

 

[FvM]

what does it mean by dc motor is rotating

You're only mentioning obvious points instead of refering to relevant design details, that may help to understand the said issues.

According to literature, I understand that BLDC designates a motor involving some method for rotor position detection in contrast to a standard synchronous motor without position feedback. As your design apparently isn't capable of providing a torque in some situations, I wonder what's the involved control method?

 

[alexplace2001]

bldc 100% duty cycle bootstrap

Hi gsecer

I'm also developing a BLDC drive using the IR2110.
check:
I've posted my shcematic and many wave forms.

I started by experimenting with a small 24v BLDC motor. At some moment I had the same problem you describe "Dc motor does not rotate without hand force". I started the control algorithm, the motor vibrated and with little help it started rotating.

The problem here was an incorrect control sequence. If you are using hall sensors be shure that they correspond to the programed phases. (by exchanging a the phase wires i solved the problem.)

The same thing reapered, but this time was simply a bad control sequence, so double check it.

From what I've read you can't send a 100% duty cycle when using bootstrap operation. Because you need to refresh the charge of the BS capacitors from time to time.
Also 440nF seems to little for bootstrap capacitance. What PWM frequency are you working with? (be aware that you need to charge the BS capacitors, by activating the low Mosfet of its branch.)

Some of the problems I had, were caused because I didn't have a dc bus capacitor. You should also need a snubber.

Hope this helps.

Alex

 

[gsecer]

hand brashless

Hi again;

My DC Bus Voltage is 24 V (24 V Bldc Motor as well). I use the same configuration in IR2110's datasheet. I am using FPGA for commutation. It checks hall sensors and provide commutation signals which are going into ir2110 at 50 khz and my PWM freq. is 20 kHz.

Yesterday; I got over the first problem which is that "motor does not rotate without hand force" by changin phase wires but now i have a strange problem. I think it should stem from a specific problem but i dont know what it is since the situtation is totally weird.

I apply PWM to the high side of MOSFET bridges and motor's no load speed is 12900 rpm as stated in its catalog.

The problem is that;

When duty is %10 motor rotates at 4500

When duty is %20 motor rotates at 6500

When duty is %30 motor rotates at 7500

When duty is %40 motor rotates at 8500

When duty is %50 motor rotates at 9500

When duty is %60 motor rotates at 10500

When duty is %70 motor rotates 11500

When duty is %80 motor rotates at 12500

When duty is %90 motor rotates at 12900

When duty is %100 motor rotates at 13500

only rpm value for %100 duty cycle appears to be ok but the other ones are all weird.they are like non linear.

Can it be caused by BS capacitors? If not what could be?

ps for Alex:

I am out of ideas just like you said in your topic Alex and my configuration is the same as yours except the resistors you used series with the reverse diode on the way of mosfet gate which is paralel to the ohm gate driving resistor. I used none of them.

I am using 1N4148 instead of SR560.

I am using UF4007 instead of 1N4937.

I am using 22 ohm resistors instead of yours 68 ohm.

Added after 6 minutes:

I am sending a schematic as quickly as i can be in order to clarify some blurry points.

Added after 2 hours 30 minutes:

I dont use any snubber elements in my circuit do I need them?

Added after 1 hours 30 minutes:

Additional info:

I do not have any oscilloscope but I observe the gating signals using a multimeter in voltage duty cycle mode.

For example, fpga sends pulse width modulated signals at %50 duty then when i control the low and high side mosfet's gate to source voltages i see that low side mosfets are triggered with %100 duty cycle high sides are modulated at %50 as i expected. Even for different duty cycle values, gating signals looks ok.

Added after 2 hours 47 minutes:

Using my multimeter, i have made some experiments again. I measured low side Mosfet's drain voltage.

For%10 duty it is 4.9V

For %20 duty it is 7.2 V

For %30 duty it is 8.4 V

For %40 duty it is 9.17 V

For %50 duty it is 10.10 V

For %60 duty it is 10.61 V

For %70 duty it is 11.22 V

For %80 duty it is 11.53 V

For %90 duty it is 11.85 V

For %100 duty only it is 12 V as expected.

However it seems that this voltages have a paralellism with the motor rotation speed naturally and it can be said that higher voltages are applied to the phases of motor.

Now my main problem is that there is over voltage across motor phases?

What is the reason guyz?

Please help :|

 

[alexplace2001]

24v bldc driver

Hi gsecer

I think I was in this situation last week.

Try Adding a 4700uF(50v or +) capacitor parallel to your 24v DC bus. This may solve your problem.

What is happening here is, that the current circulating through the motor inductance is being converted to voltage because of the switching.

I made a couple some tests at 100v and without the capacitor it got over 160v.

I think its a good idea to add snubbers, you only need to add a 0.1uf polyesther capacitor near as possible to your inverter branch. This capacitor will consume the voltage spikes generated by the stray inductance.

If you are so concerned about the speed in your aplication try adding a PI controller, you can get the speed from the hall sensors.

hope this helps

Alex.

 

[FvM]

motor stops at 50% duty cycle

I think, that the nonlinear duty-cycle to speed behaviour is caused by a none-continuous motor current at low loads. This happens, if high and low side switch of each half bridge are not driven by an alternating waveform.

 

[gsecer]

how to fix bldc motor

Firstly, thanks you guys for your replies.

However, I do not understand what FVM said. What do you mean by "alternating waveform". What i am doing is that i apply PWM only to highside MOSFETs whereas logic 1 is applied to lowside ones with respoect to info coming from hall sensors.

BTW, I am going to test the driver again after connecting a high bus capacitor.

 

[FvM]

brushless motor will not run correctly

What i am doing is that i apply PWM only to highside MOSFETs whereas logic 1 is applied to lowside ones with respoect to info coming from hall sensors.

You're most likely not applying "logic 1" (on-state) to a low side driver while the high side is PWM switched... It would mean a bridge short. I guess, it's permanently off, similar to the waveform presented by alexplace2001 in his BLDC thread, which causes the said problem of discontinuous current flow.

To keep the average motor voltage proportional to PWM duty cycle, you must allow a bipolar motor current (respectively 4-qudrant bridge operation). Freewheeling of the opposite bridge switch only allows an unipolar current.

 

[gsecer]

brushless dc motor problem

Here is my commutation algorithm

when "001" =>

ch1 <= "10" ;

ch2 <= "00" ;

ch3 <= '0' & pwm;

when "010" =>

ch1 <= '0' & pwm ;

ch2 <= "10" ;

ch3 <= "00" ;

when "011" =>

ch1 <= "00" ;

ch2 <= "10" ;

ch3 <= '0' & pwm ;

when "100" =>

ch1 <= "00" ;

ch2 <= '0' & pwm ;

ch3 <= "10" ;

when "101" =>

ch1 <= "10" ;

ch2 <= '0' & pwm ;

ch3 <= "00" ;

when "110" =>

ch1 <= '0' & pwm ;

ch2 <= "00" ;

ch3 <= "10" ;

For channels, logic "1" is applied to the low side while pwm goes into gate of high side switches.

The determination criteria of programme is hall sensor outputs. The most significant bit of 3 bits data coming from hall represents hall sensor output 3.

Added after 20 minutes:

Here is my circuit schematic

 

[gsecer]

24v bldc

BTW, alex i examined your topic and you are saying that your bus voltage increases to 160 V but i have no such problem. My bus voltage is constant at 24 V. could it still be a bus voltage capacitor problem?

 

[FvM]

bootstrap capacitor in bldc motor

The drive scheme is as expected, involving free-wheeling of the PWM switched bridge output. That's basically not an issue, but doesn't give a linear PWM to speed characteristic.

 

[alexplace2001]

why my dc motor can not start?

Hi gsecer

Well the voltage rise, were spikes of of 160v, measured with a oscilloscope. A multimeter won't meausre this kind of spikes. You don't loss anything by adding the bus capacitor, give it a try.

Alex

 

[gsecer]

bldc problem

I connected 100 nF capacitors between phases and ground. As a result speed has decreased to 5600 for %25 duty. However, i increase capacitor value by adding another 220 nF polyhester capacitor which is parallel to the 100 nF one but nothing changed.

Is it a way taken for the solution guyz?

I have 700 uF bus capacitor; isnt it enough still? what do you think?

 

[FvM]

capacitor as bldc load

As a result speed has decreased to 5600 for %25 duty.

I guess, due to increasing losses.

 

[gsecer]

dc motor does not rotate

Well, FVM what do you think as a reason?

I am totally messed up?

Should I rebuild the circuit?

 

[FvM]

I already told my interpretation, I think, it's a matter of pwm drive scheme.

when "001" =>
  ch1 <= "10" ;
  ch2 <= "00" ;
  ch3 <= '0' & pwm;

the last line should be changed to

  ch3 <= (not pwm) & pwm;

Possibly, a deadtime generator may be neccessary in addition.

 

[gsecer]

Everything is fixed up. Thank you guys a lot.

I have fixed it up in software doing what FVM said.

 

[gsecer]

There is one thing that i have forgetten to say.

Current flowing on the circuit has increased seriously. I think it is a problem which stems from cross conduction during pwm transitions. If i put a diode across High side Mosfet's source and Low Side Mosfet's Drain as it prevents cross conduction, Does problem disapeear?

On the other hand, Do I need dead time? The diode as i mentioned earlier fix this situation up instead of using dead time?