Minimum HP

Status
Not open for further replies.
weressl said:
How much short circuit can an ASD deliver? I bet no properly sized OCPD would trip before the ASD's own overcurrent limit would cut the power supply.
While I don't disagree, I still see no allowances for that fact in the NEC. As I see it, you still need to have the SCPD sized for the motor circuit.

I disagree that one can connect any number of motors on a single drive. Drive characterization portion of the software will try to calculate the motor parameters, including the cable parameters and optimizes the power supply parameters accordingly. Different cable lengths to the individual motors will result in overdriving certain motors and underdriving others in the same group while everything is running. When you start taking motors out the situation gets even more complex, especially if it is desirable to have the motors running in synch.
Nothing wrong with what you are saying here, except that this is true only with vector drives that do the motor circuit calculations and adjustments you are speaking of.

So this points out that I failed to mention that you MUST put the VFD in open loop V/Hz (a.k.a. Scalar) operating mode in order to run multiple motors from one VFD. You cannot run them in any kind of vector control mode, open or closed loop. All of the VFD manuals will tell you that as well. As Besoeker said, you will lose the accuracy and torque performance that vector drives are capable of, but it will work to drive all of the motors at the same commanded speed. In the conveyor system, as a single load moves down the conveyor and loads the individual motors differently, the load speed may not remain the same, but in a relative sense it will work. Roller conveyor manufacturers do this all the time.

Besoeker,
Its an experience thing, but I think the extra harmonic problems come from the fact that each individual motor circuit will be a different length of cable and no two motors are exactly alike, so different total circuit impedances, cable capacitances, ringing effects etc. Only 2 motors on one drive is not usually a problem, 9 on one drive is. I have done numerous tunnel freezers where we had upwards of 15 motors on one drive and the motors would fry at less than 1/2 load, which can only be attributed to harmonics. I have also done this on utility power transformer cooling fans where we had only 4 fans on each drive, still a problem. Both situations were solved with reactors. But I have done dozens of pump stations with only 2 motors on one drive with no reactor and had no problems. So the crossover point is somewhere between 2 and 4 motors on one drive, but I can't say for sure what all of the circumstances are, so it's just prudent to always put the reactor in.
 
I can't recall ever having seen 9 motors run off one VFD.

Off hand other than the problems of providing appropriate short circuit and OCPD for each motor, it probably can work.

I would be a little more nervous of the thing was spread out all over the place and the motors were hundreds of feet apart.

usually, you can easily couple sections of conveyor together mechanically. That might be a better choice.
 
petersonra said:
...I would be a little more nervous of the thing was spread out all over the place and the motors were hundreds of feet apart.

usually, you can easily couple sections of conveyor together mechanically. That might be a better choice.

1st motor is 30', 2nd 60', 3rd 90,...9th 270'. Picture a chain of xmas lights.

I imagine they can add or remove a section of conveyor now, in less time than what it would take to take the chain guard off a mechanically linked section.
 
Jraef said:
Besoeker,
Its an experience thing
Quite so.
I've been playing around with VSDs for forty some years (I'm the Victor Meldrew* of VSDs). But 40 years It still isn't enough.:grin:
[/QUOTE]
*You probably don't know who Victor Meldrew is. He was one of the main characters in the UK comedy series "One Foot in the Grave".
He's the original grumpy old man. Irascible, cantankerous, and a grouch.
Describes me perfectly!
 
Jraef said:
While I don't disagree, I still see no allowances for that fact in the NEC. As I see it, you still need to have the SCPD sized for the motor circuit.

I agree, the NEC still requires it. The only thing I am trying to point to that even though you would need to put it in it functionally absolutely useless, it does not accomplish what the NEC intends to provide, it is a waste of money for no benefit.


Jraef said:
Nothing wrong with what you are saying here, except that this is true only with vector drives that do the motor circuit calculations and adjustments you are speaking of.

So this points out that I failed to mention that you MUST put the VFD in open loop V/Hz (a.k.a. Scalar) operating mode in order to run multiple motors from one VFD. You cannot run them in any kind of vector control mode, open or closed loop. All of the VFD manuals will tell you that as well. As Besoeker said, you will lose the accuracy and torque performance that vector drives are capable of, but it will work to drive all of the motors at the same commanded speed. In the conveyor system, as a single load moves down the conveyor and loads the individual motors differently, the load speed may not remain the same, but in a relative sense it will work. Roller conveyor manufacturers do this all the time.

Hm, I don't recall any ASD manuals refering to multi-motor applications. Then again, I only know about 5 in detail.

One of the main differences between the control modes that in the scalar mode the algorithm uses a 'worse case' constant and in the flux and vector control modes it actually 'learns' the parameters and tries to calculate what part of the impedance belongs to the cable, eg. constant and what belongs to the motor. Since it still works within some parameter limits, the multiple motor scenario could be outside of the base parameters yet still within the acceptable results. The summary of Xd and R and L' of motors and cables are different than a single motor and cable so the results could be problematic. It will not necessarily result in immediate failure, but shorten the motor and cable life.

I still don't see how regular thermal overloads can adequately protect drives. It is outside of their design parameters.

In summary, I think when one goes beyond two motors they should look if the economics still make sense. I did a number of dual motor installations and they seem to be working fine in the Scalar mode.

Anyone remembers Parajust?
 
weressl said:
Anyone remembers Parajust?

Used to work for Parametrics out of Seattle in the Parajust GX drive days, right before Asea bought them. They essentially invented the modern PWM drive as we know it (before my time there). But it took ABB a long time to live down the legacy of the bad feelings they left behind in the marketplace by releasing new products before they were ready. Classic example of a company run by engineers!
 
weressl said:
I still don't see how regular thermal overloads can adequately protect drives.
You would use them to protect the individual motors, not the ASD.
weressl said:
In summary, I think when one goes beyond two motors they should look if the economics still make sense.
That's a fair point.
 
Besoeker said:
You would use them to protect the individual motors, not the ASD.

That's a fair point.

Sorry, I meant to say DRIVERS, eg. motors. The thermal overload will not protect the the motors, that was my point and I am sticking by it.:grin:

Those are designed to parallel the thermal behavior of the winding at full and SINUSODIAL FREQUENCY and VOLTAGE.
 
Jraef said:
Used to work for Parametrics out of Seattle in the Parajust GX drive days, right before Asea bought them. They essentially invented the modern PWM drive as we know it (before my time there). But it took ABB a long time to live down the legacy of the bad feelings they left behind in the marketplace by releasing new products before they were ready. Classic example of a company run by engineers!

That is an erroneous conclusion.

A company run by engineers would never be able to release a product, as an engineer would continue to tweek the item, until its perfect.:grin: They would go through the greek alphabet for pre-release trials, then start with the Chinese.

A company run by beancounters, would release a product as soon as it can be 'sold'.
 
weressl said:
Sorry, I meant to say DRIVERS, eg. motors. The thermal overload will not protect the the motors, that was my point and I am sticking by it.:grin:

Those are designed to parallel the thermal behavior of the winding at full and SINUSODIAL FREQUENCY and VOLTAGE.
Thermal overload units (and I mean thermal, not electronic) respond to RMS current and thus are waveform independent. But, if they are CT driven, I agree that the CT won't work at low frequencies. They do, however, work pretty well on non-sinusoidal waveforms at rated frequency.
Many DC variable speed drives use CTs on the AC input to derive current feedback. The input current is far from sinusoidal. The output from the CTs is rectified and the resulting output is is the same as the drive DC output but scaled by the current transformer ratio.
 
weressl said:
That is an erroneous conclusion.

A company run by engineers would never be able to release a product, as an engineer would continue to tweek the item, until its perfect.:grin:
Also known as the creeping improvement sickness.:grin:
 
Besoeker said:
Thermal overload units (and I mean thermal, not electronic) respond to RMS current and thus are waveform independent. But, if they are CT driven, I agree that the CT won't work at low frequencies. They do, however, work pretty well on non-sinusoidal waveforms at rated frequency.
Many DC variable speed drives use CTs on the AC input to derive current feedback. The input current is far from sinusoidal. The output from the CTs is rectified and the resulting output is is the same as the drive DC output but scaled by the current transformer ratio.

The thermal dissipation model of the thermal overloads is designed for motors operating at 60Hz or 50Hz, depending in which country are you in, therefore it will NOT protect as designed at different frequencies, regardless of harmonic content. Harmonic content and nonsinusodial waveform only adds to the problem.

Again it may work, UNTIL a problem occurs. Then you can go all to the manufacturer to claim that they device does not work they will tell you; you were operating it outside of it's design parameters.
 
weressl said:
The thermal dissipation model of the thermal overloads is designed for motors operating at 60Hz or 50Hz, .
If the operation of the overload realy is thermal rather than electronic it is RMS that matters.
RMS is independent of frequency.
 
weressl said:
Why don't you ask one of the design departments of the manufacturer of the units? See if they concur with your concept.
I have but, admittedly, not recently.
So, I will pose that specific question to our incumbent supplier and post their response here.
 
Besoeker said:
I have but, admittedly, not recently.
So, I will pose that specific question to our incumbent supplier and post their response here.

Make sure that the answer is from the design engineer not sales or marketing.

Please post the question as well.

Like polls, I can get any answer I want just depends how I pose the question.:D
 
Over a fairly wide frequency range, resistance 'heaters' in overload will respond truthfully to the RMS current flowing though them. When the frequency is very low the temperature of the 'heaters' will actually fluctuate with the portion of the AC cycle, and thus no longer be responding to the RMS.

However the thermal characteristics of the overload 'heaters' are supposed to be a model of the motor thermal characteristics, and the motor thermal characteristics will change significantly as motor speed changes. If you don't properly maintain flux density, you may get increased core heating with the same RMS current. At low speeds you will have reduced fan cooling, etc.

The net result is that the heaters will respond to the variable frequency current, but that response will not necessarily match that of the motor.

IMHO, if you use conventional overloads, but with a conservatively rated (should I say 'derated') motor, you should be able to provide correct overload protection given a variable frequency source.

-Jon
 
AB states which overloads to use for the protection of motors on the load side of a drive. They had two types listed in the FAQ page I read this morning. One was a melting alloy type with a W heater for class 20 protection, the other was an IEC style, if I remember correctly. Electronic overloads were not to be used.
 
Status
Not open for further replies.
Top