OT: Motor / VFD question

[DickDV]

At the risk of adding complexity, there is another type of load out there which increases torque as the speed goes down. Examples would be center-driven winders, center-driven unwinders, and many types of machine tools such as lathes, mills, and drill presses.

These are difficult loads to handle with variable speed because the load behaves exactly opposite of how the motor cooling behaves. A motor with a separate auxiliary cooling system is almost always required for these. The motor hp is also usually much higher than the actual load hp because the motor has to develop high torque at low speed and also high speed for the top speed end. Since the high torque and the high speed never occur at the same time, the motor hp is never fully utilized.

Another tough load to handle is one that cycles constantly from motoring to braking. Stamping presses and oil jacks are examples of these.

While a challenge, they are what makes this field so much fun. Even for an old **** like me! Looking forward to doing this for a few years yet!

 

[rdrast]

We also found that there is pitting of the bearings in non-inverter motors on VFDs (in our application). Changed to inverter rated motors and the problem went away.

Even with inverter rated motors, we still install shaft-grounding brushes here. They aren't the most inexpensive things in the world, but they seem to be excellent secondary protection for motor bearings.

 

[leitmotif]

Dick

Not only did it do it for me it hit the nails on the heads.
To summarizie
It is an attempt to describe motor self cooling ability at lower speeds

The Variable Torque rating is for variable toque loads such as pumps. So I would want this value to be very high ie 100 or preferably 1000

The constant torque would be the value applicable vehicle drive again I would like a higher value ie 20 as opposed to 10.

As soon as it is explained well it is obvious.
Thank you for clearing the fog.

Dan

 

[DickDV]

Actually, Dan, even tho the motor would cool properly on a variable torque load down to a very slow speed, it is rare that a centrifugal pump is run slower than half speed because it stops pumping and just slips liquid by the impeller. A fan also rarely goes below 40% speed because it stops moving air.

So, while the motors can go slow, they would not normally be set up to do that.

 

[plc noob]

In a 460VAC environment, motor leads on a 10hp motor of 60 feet or more and on a 100hp motor of 250 feet or more result in voltage pulses often approaching 1400 volts

Dick why would the voltage pulses reach as high as 1400 volts ?

Also can you provide some links or info to some detailed info on these subjects.I really want to know drives more in depth but i have found it very difficult to find this info.

Most info i have found is on drive basics and setup and not application engineering such as you have described here.

Dick is there a book you would recommend ?

Thank you for the replies of such valuable info.I know most of it is most likely from true field experience.

 

[naegely]

Thanks again. This discussion got over my head very quickly, but I learned quite a bit. I am really humbled every time I post a question on this site. Very grateful for your time.

Rob

 

[leitmotif]

Dick

I suppose that like with any other piece of machinery when you get too low on the curve they just dont work that well. The only real hands on I have done with this is at Boeing where we put in a "super slow" option on spray booth fans to provide some ventilation while parts dried. It kept the painters happy which is the reason we did it. Going this way instead of adding another fan and ductwork saved $30K or so I was told. Copied this from Navy submarine experience where main sea water pumps had 3 speed from a 2 speed motor and a piggy backed motor with drive belt to get super slow. Never did measure flow vs RPM so I do not know how less efficient either setup was in super slow.

Looking back we should have used VFD at Boeing instead of a belt drive but I had not learned about VFD at that time (1991 or so)

Dan

 

[milldrone]

Dick why would the voltage pulses reach as high as 1400 volts ?

Dick may have a better explanation. but here is an excerpt from
http://216.85.60.10/qd/Applicat.nsf...11794a749120b819862566ac0055c8ed?OpenDocument

IGBT's

The relatively recent availability of high voltage, high current IGBT's has led to the wide use of these devices as the main switching element in the D-C to A-C inverter section of 1-phase and 3-phase AC Pulse Width Modulated VFD's. Virtually all of the manufacturers of these types of power conversion circuits have developed, or are developing, product lines that utilize these relatively new devices. One of the main reasons for the widespread use of these devices is their extremely fast switching time. This results in very low device transition losses and, therefore, in highly efficient circuits. In addition, a fast switching time allows drive carrier frequencies to be increased above the audible range. (Slower switching topologies operating at a range of 1 to 2kHz often induced irritating mechanical noise in a motor.)
The Reflected Wave Phenomenon
Voltage wave reflection is a function of the voltage rise time, (dV/dT), and of the length of the motor cables which behave as a transmission line. Because of the impedance mismatch at both ends of the cable, (cable-to-inverter and cable-to-motor), some portion of the waveform high frequency leading edge is reflected back in the direction from which it arrived. As these reflected leading edges encounter other waveform leading edges, their values add, causing voltage overshoots. As the carrier frequency increases, there are more leading edges present that "collide" into one another simultaneously, causing higher and higher voltage overshoots. If the voltage waveform was perfectly periodic, it might be possible to "tune" the length of the wire. However, since the width of the pulses varies throughout the PWM waveform, it is not possible to find any "null" points along the lead length where the motor may be connected
without the fear of damage.

I do not know if this next quote is true but it makes sense to me.

Insulation Punch-Through Failures

Seldom, if ever, do large motors fail due to insulation punch-through. This is because they are usually "perfect" wound, which means that the location of each turn of wire in the phase winding is precisely controlled. Therefore, the level of voltage from turn to adjacent turn is controlled. In smaller motors, however, the wire size is quite small and the number of turns is large. Usually, these motors are "random" wound and do not lend themselves to control over the proximity of adjacent turns. Therefore, it is quite possible to have two turns of wire next to each other with a high voltage potential that is close to the maximum allowable limit of the insulation system. Even in the absence of an overshoot voltage, when a high dV/dT is applied, the insulation components may experience punch-through, causing motor failure. Normally, these types of failures occur within hours or weeks of start-up.

This part especially

Normally, these types of failures occur within hours or weeks of start-up.

This explains to me anyway, why some installs work for years even though they violate the motor lead length ideals. (Someone just got lucky)

 

[DickDV]

There may well be some good info on this subject in this board's FAQ section or you could Google using "VFD reflected wave" or maybe "VFD long lead problems". I have not found in my own travels a very practical explanation of the long lead problem.

At the risk of raising more questions than answers, the square pulses that the drive sends out to the motor are full of high frequency components, some as high as 1 MHz. Since all conductors including straight wires have distributed inductance along their length, the longer the lead, the more the drive looses control over the higher frequency components of the pulse. As a result, on the very fast pulse rising edge you start to get overshoot. The longer the lead, the more the overshoot until, at the lengths I specified, the overshoot extends from the 650V pulse up over the pulse to 1400 or so volts. You get a similar undershoot when the pulse falls back down to zero.

The phenomenon is much like holding a 1/4" steel rod in your hand horizontally. If you have a rod only 2 feet long and you abruptly switch its position up and down repeatedly a distance of 6 inches, the loose end of the rod will pretty much repeat the movement of your hands. But, if you switch to a rod 10 feet long and again move your hands abruptly up and down 6 inches repeatedly, the far end of the rod will whip wildly, far beyond the 6 inches movement of your hands. That same "whipping" of the signal occurs as the drive/motor leads get longer and longer.

I know, the illustration is pretty poor but, if anyone wants to substitute something more understandable, please do. Trying to explain this mathematically causes most peoples' eyes to glaze over.