DC Bus Compensation

[Marke]

Is there a way to disable the DC Bus compensation on the PDL Elite VFDs and the ATV 61 VFDs?
We have found this can be necessary to prevent oscillation with drives used with harmonic filters on weak supplies.

Mark Empson
Advanced Motor Control Ltd

 

[mikekilroy]

ok, never too old to learn. I'll ask: what is "dc bus compensation?"

 

[Marke]

Hi Mike
DC Bus compensation is incorporated into drive algorithms so that the output modulation is varied depending on the DC Bus voltage so that the motor voltage is less dependent on the incoming supply voltage.If the DC bus voltage increases, the output voltage in an open loop environment would also increase, likewise if the DC bus voltage reduces the output voltage would also reduce. By applying some control based on the DC bus voltage, the output voltage can be regulated by adjusting the output modulation.

When a passive harmonic filter is applied to the input of a VFD on a weak supply, instability can result if the response time of the DC bus compensation is too fast for the time constant of the input circuit including the filter. The solution is to either alter the time constant of the DC bus compensation, or to turn it off.

Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[LiteYear]

Oh, in that case you'll probably find the ATV 61 (I'm not familiar with the PDL Elite, but it's likely similar) doesn't have an equivalent function. What you describe sounds more applicable if the drive were operating in V/Hz mode - that is, controlling for output voltage based on the output speed. In fact the ATV 61 is typically configured in space vector modulation mode, so the control algorithm's goal is to match a reference vector. I doubt it's possible to separate the functions of voltage and phase control. In other words, "DC bus compensation" is "built-in" to the control algorithm.

 

[sibeen]

Pardon my ignorance on this one, but I thought all drives used space vector modulation; it's just that the reference vector depended upon the operational mode?

 

[ozmosis]

Hi mark
I don't think you can do it on the ATV 61, other than switching the drive off.

 

[mikekilroy]

OK, so your comp is a method to reduce overvoltage error. You may also be including a much much more complex algorithm that includes basically 2x vfd drive cost to have input IGBT control rather than diode front end.....

Let's attack each individually....

Complex first.....

If you say this algorithm will artificially or any other way INCREASE output voltage OVER the input voltage, then you jut doubled - plus - the price of the drive. Other than sophisticated Siemens special drives I know of no other that will make a higher output voltage than available from the input. I am not all knowing of course so I can be full of it.

That all said, to have higher output ac voltage than input requires an active front end rather than diodes, in addition to input reactors: the idea is to SHORT out the input choke each of say (as in siemens case) 2khz rate so the inductive KICK is HIGH and gets rectified. result is a higher dc bus voltage than just diode 1.4x input so you can have higher output ac voltage. I really really doubt either brand you mention has this.

so forget making dc bus higher to get higher output ac than input on your models identified.

Now to the simper rEducE output..... ok, so what cause dc bus to go higher than nominal? REGEN. Ie., deceling or overrun motor. If it goes too high then we have possiblility of OVErVOLTS fault. How to reduce dont do the thing that caused the high decel energy, or reduce the decel ramp. AH...... now your onto something... What if we watch the dc bus voltage, and if it starts to go too high and threaten to fault, we instead REDUCE THE DECEL RAMP? Can we call this "dc bus compensation?" Sure, if it helps sell more drives. OK, so let's do it... So any parameters in the equation Sure, we can pick the START point of dc bus BEFORE overvolts fault, we can pick response time like PID stuff.... Enter more modern drives.... Hitachi for one has this. You can adjust the volt sense point as well as the time before response. If you need the B085() etc paramter nos I can look it up for u. Call if "dc bus comp?' sure if it helps sell more drives....

Say it will "adjust the pwm output freq to adjust for it" sure, if it sells more drives....I'll take your order sir!

So is this a function of 'vector drives?' 'space comp drives?' 'v/hz drives?' YES. IT DOES NOT MATTER! this is a totally separate algorithm and has nothing whatso ever to do with motor speed control.

 

[LionelHutz]

mikekiljoy - I don't think you're understanding what it does....

It is not specific to regenerating motors or active front ends. Basically, the duty cycle of the PWM output is varied depending on the bus voltage change from ideal or nominal.

 

[ozmosis]

The question was, can the ATV61 or the PDL drive do this function? I think the answer is no.
The questions about the function itself, can be answered quite simply by saying other drives on the market can do it. We manufacture one of them.
However, in the interest of not promoting ones own product, this is better left where it is.

 

[Marke]

Hello Mikekilroy

Thank you for your long explanation, however as LionelHutz suggested, the issue is not about regeneration or any thing, but it is a pretty standard algorithm built into the waveform generator to keep the V/Hz closer to the ideal when practical, especially at reduced speeds, and reduce the dependence of the V/Hz on the incoming supply voltage. At full speed, overmodulation can be used to increase the average voltage, but this is at the expense of increased harmonic currents to the motor as the waveform becomes squarer than it should be.
As Osmosis has suggested, some drives provide the ability to turn this algorithm off, or to change the time constant to minimize instability. In some cases, the only way to reduce the problem is to reduce the rated motor voltage, reducing the flux in the iron across the whole speed range. Not ideal, but it can help with instability, but increases motor current.

Mark Empson
Advanced Motor Control Ltd

 

[controlsdude]

was using sensorless vector to start

It was unstable when I had a problem with a simple process had to disable Bus regulation mode but other settings too.
try these other settings if you have them in your drive
you dont have this or maybe they need to update the firmware? >>>set to disable Bus Reg Mode

Slip Hertz @ FLA = disable Compensation
Compensates for the inherent slip in an induction motor.

This is a mechanical or electrical compensation setting.
Enables/disables correction options that may improve problems with motor instability.

set your dynamic brake setting to 100% (I did not have a dynamic brake but still changed the setting.
NoProtection” (100% Duty Cycle) Maybe think more about this if you do have dynamic brake!

All else fails set to V/hz
v/hz last chance

 

[mikekilroy]

ozmosis I would like to read a description of a drive with this feature. It would not be self promoting but educational if you would simply provide a link to a drive document that discusses this feature please?

 

[ozmosis]

Mike
Here you go

http://mcliterature.danfoss.com/WebPublish/doc_MG80C402.pdf

Page 56.
However, it is just highlighting that there is a parameter that can be disabled.

 

[ozmosis]

I was thinking more about this topic. Some might be thinking "why the need for this function in the first place?"
The reality is that 6 pulse drives create harmonic distortion. As a drives manufacturer, we need to obviously extoll the virtues and benefits of having a drive in the first place but also be conscious of the impact of fitting these devices too. This is why drives have EMC filters, harmonic filters, high IP ratings, coated boards etc. Hopefully, the benefits outweigh the issues. They will generally outweigh when you know most of the facts behind the installation and the application.
Sometimes, you don't. One of these, and quite understandably, is the fact that purchasers of drives will not always know how 'strong or weak' the supply impedance is. They will know drives create harmonic distortion and we will tell them that there are various ways of mitigating this distortion (active, passive etc). Sometimes, customers will buy a drive and buy a harmonic filter and we have no idea (as they typically don't) about the supply conditions. There is no "correct" solution, it depends completely on the specific installation. Unfortunately, we cannot advise on every sale.
So, in situations where drives and passive harmonic filters are installed on very weak supplies, you can get additional issues with a drive that does not behave as it should do and one of these is the effect described by Marke above. This is why we have to introduce functions to support an installation that respond to problems 'after the event'. The farmers in NZ generally don't take too kindly if you say to them after they have invested in drives, filters etc "why didn't you tell us before, what the supply impedance was..?" Usually the second word response is "..off!"
Hence the need to have functions that can correct certain issues we find out afterwards. We call it a hindsight parameter.

 

[mikekilroy]

Ozmosis, thank you for the reply! Unfortunately the link says not foundhere in Ohio

http://mcliterature.danfoss.com/WebPublish/doc_MG80C402.pdf

I search the site for a model MG80 but do not find one.

Please list maybe a drive part no. (& maybe that parameter no) so I can download its manual maybe?

thanks!

 

[mikekilroy]

found doc. tnx

 

[Marke]

One of the problems associated with DC Bus compensation, is that it can cause instability when the VFD is used with a tuned trap harmonic filter on a weak supply, but also, it can cause a significant error in the V/Hz ratio if the calibration of the DC Bus voltage measurement is incorrect.
I recently came across a situation where the magentising current is increased by about 35% due to incorrectly calibrated DC Bus monitoring. I suspect that the VFD was calibrated without a DC Bus Choke and then this was added without recalibration.

Mark Empson
Advanced Motor Control Ltd

 

[mikekilroy]

Trying to get a handle on this feature.

how much does dc bus comp adjust the output voltage? 2%? 5%? 50%?

Has anyone ever seen it on a drive running in vector mode (i.e., not v/hz)? If so, what brand & model?

Is it more common on very large (say >500hp) drives?

Marke, you seem to see it a lot; what brands & sizes of drives was it on?

 

[Marke]

Hi Mike

Most of the well known drives that I see, have this feature in one form or another when running in V/Hz mode. It is often hidden with no way to control it and if you do not have problems with it, you would not know that it is there.
Ideally, if the input voltage is low, then the DC bus voltage will be low and under true open loop conditions, the output voltage will also be low. Adjusting the modulation level enables the output voltage to follow the correct V/Hz curve up to the supply voltage. Over modulation (adding output harmonics) can be applied to further boost the output voltage when the supply voltage is low. Likewise, if the supply voltage is high, the output waveform can be under modulated to reduce the output voltage so that the correct output voltage is met.
The major issues occur if the DC Bus measurement calibration is incorrect, or if the control loop time constant is too fast. Incorrect calibration will alter the V/Hz and can, as I have recently seen, overflux the motor, or if the loop is not slow enough, the output voltage can go unstable on a weak reactive supply.
There is little if anything in the manuals describing the feature. The waveform generation, (commonly space vector modulation) is able to set both the frequency and the output voltage

Mark Empson
Advanced Motor Control Ltd

 

[Marke]

Hi Mike

Most of the well known drives that I see, have this feature in one form or another when running in V/Hz mode. It is often hidden with no way to control it and if you do not have problems with it, you would not know that it is there.
Ideally, if the input voltage is low, then the DC bus voltage will be low and under true open loop conditions, the output voltage will also be low. Adjusting the modulation level enables the output voltage to follow the correct V/Hz curve up to the supply voltage. Over modulation (adding output harmonics) can be applied to further boost the output voltage when the supply voltage is low. Likewise, if the supply voltage is high, the output waveform can be under modulated to reduce the output voltage so that the correct output voltage is met.
The major issues occur if the DC Bus measurement calibration is incorrect, or if the control loop time constant is too fast. Incorrect calibration will alter the V/Hz and can, as I have recently seen, overflux the motor, or if the loop is not slow enough, the output voltage can go unstable on a weak reactive supply.
There is little if anything in the manuals describing the feature. The waveform generation, (commonly space vector modulation) is able to set both the frequency and the output voltage. The output voltage is set to the motor voltage at rated frequency and reduced from there as the frequency reduces with the measured DC bus voltage being part of the equation.

Mark Empson
Advanced Motor Control Ltd