VSD rated motors 1

[d23]

All:

I have a few questions about VSD’s that is the result of reading other threads referring to motors needing to be VSD rated. I have not worked with drives in several years and realize that the technology has changed. Please correct me with my outdated assumptions.

I have seen several references to the output voltage being 1.414 higher than the output RMS volts. In the past with PWM type drives the actual overshoot would be in the order of 3 to 4 times the bus voltage at the time of the semi-conductor switching. Even then to get numbers that low you would need an extremely well designed drive. The only way to prevent this overshoot was to build an elaborate capacitor filter to be placed between the drive and the motor. In simple terms this overshoot would literally try to shoot holes in not only the motor insulation, but the cable insulation too.

Q:
Has drive technology improved enough that the expected overshoot is not that high?

The other point about using “inverter duty motors” was the expected vibration or mechanical stress on the motor winding themselves that was caused by the constant switching or pulses.

Q:
Do VSD rated motors have additional blocking on the windings to prevent them from moving (walking) during inverter operation?

Old technology was that the pulses from a PWM drive would try to shoot holes in the insulation of both motors and cable.

Q:
Were there ever any recommendations about upgrading the cable insulation on the output of a VSD?

Once again I am very far removed from the drive industry. It is my understanding that drives are now operating with a sinusoidal output. The idea of operating a transistor in a linear fashion is amassing. Heat dissipation must be a big problem.

Q:
Are drives being made that can do that? If so what type transistor is being used? With a drive like this do you need a VSD rated motor?

Thanks for any input you may offer.

David

 

[d23]

All:

A couple more questions about motors and drives.

Q:
The limit I have been told for a two pole motor, 100+ HP is 64 hertz due to the rotor design. Is this still the upper limit for most two pole VSD rated motors?

Q:
It use to be that PWM drives had a limit to cable length between drive and motor. At some point you would have to start de-rating the drive. Is ringing caused by cable length still a problem for drives?

Thanks for your help!

David

 

[dadfap]

d23,

Do you still need help here?
Are you refering to drives used on submersible motors?

All the best

 

[GusD]

Hi D23
1-The voltage spikes can be as high as 2.5 times.
At times depends who you ask.This value is most often heard .Transistors (IGBTs)are the most common today.Yes they require enormous cooling to operate reliably.
2-To my knowledge,the windings of an Inverter duty motor mechanically are no different than other motors.The bracing of coils and support of end turns are much the same.
3-As for the upgrading of motor feeder cables for VFD application,iI don't believe that is the case as long as the cable is rated for the application.
4-the sinuzoidal output is probably a little stretch.The wave still formed from a bunch of square pulses and it may approximate a sinewave,it still only an aproximation.
5-As for the Inverters shooting holes thru cables and windings ,it is part and parcel of inverter swithching.High Vd/Vt can raise havoc with the in coil turns and unless special protection is applied to these coils,failures are common.Today, we have a too ready a
tendency to blame all winding failures on inverter duty applications.
I have seen Random winding failures that did not need a VFD application to fail,they were just poorly rewound and would have failed prematurely in any application.
David ,I hope it helsp,a little bit anyway.

GusD

 

[CB2]

d23,
1 - The improvements made in VFDs during the last few years is the problem. Older VFD designs did not have the advantages of today's very fast and efficient IGBTs. The votage rise time (dv/dt) can cause problems with motor windings. Primarily seen in small motors (<50 Hp) and applications of motor cable in excess of about 100'. The cable itself has not been a problem. Improper grounding practices and poor motor junction box connections have been repeated problems.

2 - In agreement with GusD, inverter duty motors are mechanically built the same as standard duty motors. Inverter duty motors are different because they use a high voltage winding wire insulation (1800 V) and address cooling at less than full speed.

3 - Standard cable is suitable for the VFD load side. There is special cable available for dealing with common mode voltage problems.

4 - The short answer, with an output &quot;sine-filter&quot;, the motor will see a very clean sinusoidal wave form. An expensive filter, used for very quiet motor operation, very long motor leads, common mode voltage problems, and when retrofitting old motors.

Heat dissipation is an issue with VFDs. From a different perspective, the VFDs today are 98% efficient at full load/speed, not bad when compared to older obsolete variable speed drives.

5 - If you check with the motor manufacturers you will find that most any standard motor is able to run up to 90 Hz with maybe some consideration for the bearings.

6 - The &quot;ringing&quot; of &quot;standing wave&quot; problem you refer to is a motor problem, the VFD does not care. There is no reason to derate the VFD due to long motor leads.

We are having fun
CB2

 

[d23]

All:

First I sincerely appreciate your input(s). Most of this post is over, but I am still interested in transistors that operate in a linear fashion. Is drive technology to the point of linear or (somewhat) sinusoidal output without capacitive filters or is it still PWM? I have heard that AB can go up to 500 HP in a linear fashion. Is that correct?

I would still like an email address for dadfap.

Thanks for all your help!

David

 

[dadfap]

d23

dadfap@hotmail.com

regards

 

[dadfap]

d23

The motor ringing issue can be dealt with in two ways.
1. Use of a filter, strongly recommended by the likes of Toshiba for ANY motor where there is an extended cable between the motor & VFD. They presented a paper at last year's ESP roundtable.
2. Use a bus follower style VFD where the output is a quasi sinusoidal 6 step output. I know there will be arguements about the effects of heating on the motor, but this applies to both styles of drive

All the best

 

[CB2]

dadfap

Can you provide more info on your post #2. I would like to know what it is.

cb2

 

[UKpete]

d23 - I actually wrote my reply a few days ago before anyone else had responded, then had problems getting on-line. But here it is anyway, it may fill in some of the gaps.
Q:
Has drive technology improved enough that the expected overshoot is not that high?
- No I don't think so. Modern drives are PWM which have high frequency content and potentially plenty of overshoot. The 1.414 factor you mention is simply the ratio of peak/rms for a sine wave, it doesn't directly describe the ratio between the voltage at the VFD terminals and the motor terminals - this depends on the motor size, cable length, VFD design (e.g. DC link voltage) etc.

Q:
Do VSD rated motors have additional blocking on the windings to prevent them from moving (walking) during inverter operation?
- Again I don't think so. Movement of end-windings is particularly due to high currents including fault currents; high frequency components will not induce strong forces. Most VFD motors will have VPI (vacuum-pressure impregnation) anyway so the windings will be less susceptible to vibration.

Q:
Were there ever any recommendations about upgrading the cable insulation on the output of a VSD?
- I don't know for sure but I think the main problem was always the enamelled wire in the motor winding, not the connecting cables which always had significantly more insulation on them and also run a lot cooler. A VFD supplier would be able to tell you of course.

Q:
Are drives being made that can do that? If so what type transistor is being used? With a drive like this do you need a VSD rated motor?

- The only pure sine wave VFD drives I've come across are for very high speed machines working at high frequency (say 1kHz). You may be thinking of PWM drives which switch at frequencies say 100x the fundamental frequency, these give a more sinusoidal current but there are still a lot of voltage harmonics there, with fast rise times. It is not just voltage amplitude that does the damage, it is the rise time (measured in V/micro-seconds).

Q:
Are drives being made that can do that? If so what type transistor is being used? With a drive like this do you need a VSD rated motor?
- not transistors (i.e. bipolar). Most VFDs have IGBTs these days, these have replaced SCRs which were difficult to switch off. Low power VFDs use MosFETs, very high power ones use GTOs. The IGBT has very fast rise times, and it was this that was causing motor insulation failures; so whilst the VFD has got simpler, cheaper and more compact, it has actually resulted in more stress on the motor!

Your comment about the maximum speed - for a particular motor this is determined by ability of the rotor to withstand the rotating forces, so particularly with the higher powered machines the higher speeds are difficult to achieve. So the motor manufacturer must specify what the maximum speed is.

As far as I know, the only differences between an inverter-rated motor and a standard motor are the insulation system, particularly the wire enamel which has special additives to resist partial discharge damage - for example Essex Ultrashield Plus and Phelps Dodge Quantum Shield (these are easily recognizable, they have a green or brown enamel). Also the use of insulated bearings (at least at the non-drive ends) as bearing currents are another VFD induced problem.

There is a good NEMA guide on VFDs that has been posted before, this is free on:

http://208.156.24.64/index_nema.cfm/1427/0DDF742A-8790-4A21-88D139CFC069648B/

(it is the application guide mentioned under the heading &quot;Complimentary Documents&quot

UKPete

 

[Skogsgurra]

Yes, there is a VFD with pure sine-wave output. Visit USPTO and search for patent number 4,947,309. It describes a way of switching with whithout producing HF components. The drives have an absolutely clean sine-wave output and the efficiency is comparable to an ordinary PWM inverter. Also, visit

http://www.nfodrives.se

and select the english version. There is a description and wave-forms that shows the actual output.

 

[dadfap]

cb2

These drives use a controlled rectifier on the input to control the DC bus, the level of which is dependant on the output voltage.

The inverter section is IGBT switched to give a quasi sinusoidal output.

These have been used for years from 60Kva to 1000Kva on submersible motors in the oil field. Cable ranges vary from a few hundred metres to 1500+ metres.

Personally, I have run 330HP submersible motors as high as 81Hz.

You can go to

http://www.centrilift.com/CDS/Electrospeed_GCS.htm

or

http://www.centrilift.com/CDS/Electrospeed_ICS.htm

all the best

dadfap

 

[UKpete]

skogsgurra - thanks for that.

The high-speed drive I referred to was also from NFO I think, except it was for a PM synchronous motor. They still need a filter, although I accept it is actively integrated into the drive. Perhaps this limits the maximum rating that is offered, 11kW. Above this, you are limited to PWM drives.

UKPete

 

[jbartos]

Suggestions to d23 (Petroleum) Jan 19, 2003 marked ///\\I have seen several references to the output voltage being 1.414 higher than the output RMS volts.
///AMPLITUDE=Sqrt2 x RMS=1.414 x RMS for sinusoidal waveforms.\\ In the past with PWM type drives the actual overshoot would be in the order of 3 to 4 times the bus voltage at the time of the semi-conductor switching.
///These would be more on spike levels than on overshoot level.\\ Even then to get numbers that low you would need an extremely well designed drive.
///PWM modulation is good modulation. However, there are other principles and better ones.\\ The only way to prevent this overshoot was to build an elaborate capacitor filter to be placed between the drive and the motor.
///Not necessarily. This depends on the principle of operation. There are several of them.\\ In simple terms this overshoot would literally try to shoot holes in not only the motor insulation, but the cable insulation too.
///Yes. Voltage spikes are dangerous to insulations.\\Q:
Has drive technology improved enough that the expected overshoot is not that high?
///Yes. There are different principles of operation, E.g. Zero crossing switching, matrix converters, cycloconverters, etc.\\The other point about using “inverter duty motors” was the expected vibration or mechanical stress on the motor winding themselves that was caused by the constant switching or pulses.
///Yes, there are higher harmonics, electrical and mechanical. The mechanical harmonics or vibration would have to experience resonances; else, the higher frequency vibrations do not have that much power in them.

Q:
Do VSD rated motors have additional blocking on the windings to prevent them from moving (walking) during inverter operation?
///Better produced inverter duty motors have very carefully wound coils and have them carefully embedded in slots.\\Old technology was that the pulses from a PWM drive would try to shoot holes in the insulation of both motors and cable.
///Yes, the spikes were damaging insulation of conductors and winding.\\Q:
Were there ever any recommendations about upgrading the cable insulation on the output of a VSD?
///Rather than upgrading cable insulation, the remedy of higher harmonics by filtering, slower switching devices, and different principles of VFDs took place.\\Once again I am very far removed from the drive industry. It is my understanding that drives are now operating with a sinusoidal output. The idea of operating a transistor in a linear fashion is amassing.
///Nothing new. The electronics have been using linear portions of transistors since their beginning.\\ Heat dissipation must be a big problem.
///Yes and no. The heat sinks are literary the science nowadays.\\
Q:
Are drives being made that can do that?
///Yes, there is a fierce competition and noticeable progress in terms of switching devices, topologies, and principle of operations.\\ If so what type transistor is being used?
///Power transistors, IGBTs and IGCTs mostly. Others are coming.\\ With a drive like this do you need a VSD rated motor?
///VSD rated motor is much better quality motor than a standard or ordinary motor.\\

 

[madddog]

DADFAP you refer to Baker Hughes Centrilift and the fact that you have experience in pumping power to a 330HP submersable pump. I need to find characteristic curves on any VFD rated ac motor rated to 100HP. If they are available from the Baker website I sure was not able to find them. Any manufaturer will do at this point. I just need to see a curve given a constant ac voltage and varing the frequency, how the horsepower varies.

 

[CB2]

Alot going on in this post.

To clarify some of it, the Centrilift applications have one unique problem to overcome in powering their pump motors, VERY long motor leads and thus significant voltage drops. Up until recently the only VFD technology viable for this was the old VVI &quot;6 step&quot; drive. Key to the aplication is the ability to output the higher than line voltage to compensate for the voltage drop at the motor. In the general drive market, this type of VFD has not been offered by any manufacturer for close to 10 years.

Recently, the current technology, voltage source PWM drives have moved into the down hole pump market. Centrilift (their own), REDA (Toshiba), ABB and a few others. This is just another example of where the PWM VFD/induction motor combination is forcing the old mechanical and electrical variable speed drives out of most markets. Only the niche players currently survive.

I also checked the NFO Drive web site that Skogsgurra suggested, alot of marketing but not much technical. Smells like a VFD with an output filter.

 

[dadfap]

CB2

You are correct, but all of the manufacturers almost insist on operating the ESP motors on a PWM VSD with an output filter.

To do otherwise is courting disaster and these add cost, footprint and complexity. Much of the move to PWM in this market segment was driven more by the PWM drive manufacturers wanting to enter this market than a strong technical arguement for their use.

maddog

The Baker website will not be able to help with this request. All of the motor curves from Centrilift will be at 50 / 60Hz. The way these motors are run on a VFD has the voltage and frequency variable, hence the use of the v/Hz ratio. Voltage is dependant on the output frequency of the drive.

dadfap

 

[Skogsgurra]

UKpete, your posting april 26

Well, yes - and no.

There is a distinct difference between a filter where the &quot;reject&quot; is filtered away and other topologies where there are no rejects to filter.

If you study the NFO Sinus inverter carefully you will find that their solution is a soft switch/energy store and not a filter.

The difference is not only semantic - it is very real. The rejects that are filtered away in an ordinary PWM inverter with an output filter are dumped to PE while the energy store type has no connection at all between electrically active circuits and PE. And, as you know, PE is supposed to be clean - not a sink for HF pollution.

This has some important inplications, while ordinary inverters with motor filters have leakage currents in the &quot;many hundreds mA&quot; region the energy storage inverters have less than one or two mA low-frequency leakage. Which makes the use of RCDs possible, even with 10 or 30 mA trip current.

For drives that need good dynamic performance (lifting devices, robotic applications et cetera) a sine filter is disastrous - many inverters cannot even auto-tune the motor parameters if a sine filter is fitted. The energy storage technology makes a big difference and dynamic response with an ordinary sensor-less induction motor can easily make a controlled 20 percent speed step in about 7 milliseconds, even around zero RPM.

The EMC level that you can reach with PWM inverters plus filters (line and motor filters) plus screened cables is often way above the EU EMC directive limits. With the soft switching energy storage technology you are BELOW the EMC directive limits - without any filters and without any screend cables and special EMC glands.

In fact, the Swedish National Electrical Safety Board has conducted tests that show this. The test report is available at the Board as Document # 84-99-273. It is yours for the asking - that is how the Swedish open society works!

Comments invited.

 

[jbartos]

Suggestion: Visit

http://scholar.lib.vt.edu/theses/available/etd-102497-12366/unrestricted/final.pdf

for:
Soft Switching Versus Hard Switching Efficiencies
ZVT-Zero Voltage Transition
ZCT-Zero Current Turn-off
ZVT Inverter
etc.

 

[Skogsgurra]

Thanks jb!

Interesting reading, the difference between inverter types is not extremely big. But an important aspect for under-the-hood applications is how much cooling will be needed - and that has been very well covered in this thesis.

There is a little problem with this paper, however. Have a look at figure 14 and judge for yourself. How well does the author know power electronics?