DcBusComp with big DcBus ripple

Daniel Hofstetter said:

I'm using InstaSpin with DcBusCompensation. But i still see the DcBus ripple on my current Signal. (I have a big ripple (50V) with 300Hz. (3 phase input with low DcBusCapacity))

DC Bus Compensation does not stop your bus from rippling. It keeps the proper voltage output on the inverter regardless of the actual Bus. To explain the effect, if you have a nominal 350V bus, and your algorithm determines it needs 300V, it may produce a duty cycle of 85.7%.  With DC Bus compensation you are measuring the actual bus, so that if the bus is actually 325V, and you still want the 300V output, you produce 92.3% duty cycle.

But as you noticed, we get this effect not by directly changing the duty cycle - which can be unstable and saturate - but by scaling the proportional gain of the current controllers.

4375.Current ripple.tif

Hi,

I know that the DC Bus Compensation does not stop the ripple an the bus. But i expect that it stop the ripple on the phase current.

In my "Current ripple.tif" you can see that when the DC Bus(C1 Yellow) is low, then the Phase Current (C4 Green) is also low. I think with a direct compensation it shoud be better, when it's enough fast.

For my 300Hz ripple i change the DCBus Filter (My Hardware ist fast enough)

#define USER_DCBUS_POLE_rps (1890.0)  // for 300Hz ripple

Any idea why there is still ripple on the current?

Thank you

Dani

Hi,

One possible issue is that the output of the current controllers are already giving the maximum output, so even if it tries to compensate for a drop in DCbus, there is no more voltage left to add to the output.

Is it possible that you capture the output of the Id and Iq controllers when this is happening?

-Jorge

Dear Chris,

I'm running my project that developed from Lab11a. I want to check the DcBus Compensation when running the motor with V/f control to calibrate my hardware before running the system with Vector control (will use instaSPIN's). However I have a problem every enable the DcBus Compensation by the following code:

if (gMotorVars.Flag_enableDcBusComp == true)
{
   oneOverDcBus = EST_getOneOverDcBus_pu(estHandle);
   Vab_pu.value[0] = _IQmpy(Vab_pu.value[0], oneOverDcBus);
   Vab_pu.value[1] = _IQmpy(Vab_pu.value[1], oneOverDcBus);
}

SVGEN_run(svgenHandle,&Vab_pu,&(gPwmData.Tabc));

The problem is: Every setting gMotorVars.Flag_enableDcBusComp == true on debugging watch window, the motor get big swing on the motor speed and the motor cannot run smoothly.  I checked the result oneOverDcBus versus gAdcData.DcBusVoltage and verify that it is true.

Where am I wrong?

Thank you for your help!

Mr. Tran

Dear Chris,

I captured the screen-shot for two cases:

a) At VdcBus = 100 Vdc. Please look the gAdcData.dcBus = 0.1071 (p.u) which is equivalent to 107.1 Vdc, and gOneOverDcBus = 10.666 (p.u)

b) At VdcBus = 330 Vdc. Please look the gAdcData.dcBus = 0.3312 (p.u) which is equivalent to 331.2 Vdc, and gOneOverDcBus = 3.271 (p.u)

Whether I was wrong in somewhere?

Thank you for reply as soon as possible.

Mr. Tran

Dear Chris,

I checked again in my code, and found that, the last report is the result of calculating oneOverDcBus of my method after discussing with you on the topic:

Which is created by implementing the following codes:

	if(gMotorVars.Flag_enableDcBusComp == true)
	{
		// or Calculate 1/VdcBus_pu by my own function
		gOneOverDcBus = _IQdiv(_IQ(1.0), VdcBus_pu);                             //VdcBus_pu = _IQ12toIQ(gAdcData.dcBus));
		Vab_pu.value[0] = _IQmpy(Vab_pu.value[0], gOneOverDcBus);
		Vab_pu.value[1] = _IQmpy(Vab_pu.value[1], gOneOverDcBus);
	}

When running the code using TI's function, I got the result as following:

	EST_run(estHandle,&Iab_pu_fb,&Vab_pu_fb,gAdcData.dcBus,gMotorVars.SpeedRef_pu);
        if(gMotorVars.Flag_enableDcBusComp == true)
	{
		//! \brief using the TI's estimation.
		//! \notes firstly using this function, user need to call the EST_run() before that.
		gOneOverDcBus_est = EST_getOneOverDcBus_pu(estHandle);

		Vab_pu.value[0] = _IQmpy(Vab_pu.value[0], gOneOverDcBus_est);
		Vab_pu.value[1] = _IQmpy(Vab_pu.value[1], gOneOverDcBus_est);
	}

Notice that the result in gOneOverDcBus_est = 100 (in IQ24). This result is different from the result which return from my own method for the same gMotorVars.VdcBus_kV = 200V.

Where did I misunderstand with the concept of 1/VdcBus_pu?

Many thank you for your ideas.

Dear Chris,

Thank you for your advice!

I will test the feature of DC Bus Compensation late when my hardware is available for this. I can ask for your help for this problem again.