Minimum Drive Frequency

[Beggar]

I've just been told that motors run through a VFD shouldn't be run under 20 Hz... No reason provided, just the advice.

I can't think of any good reason except that the cooling drops with the speed. In my case, I'm typically drawing less than 50% of FLA (though in a continuous-duty application with a moderately variable load) and the motor casing doesn't get too warm.

Can anybody validate or debunk that advice, along with justification either way?

At this point, I'm not at liberty to start fiddling with the drive components to change shaft-to-output ratio - I've just got to go with the VFD.

I just don't want to do something stupid and destroy my drive or motor.


--------------------
How much do YOU owe?

http://www.brillig.com/debt_clock/

--------------------

 

[itsmoked]

Cooling is the biggie. You can, at some point, have cogging that could cause resonances but this would probably be at lower yet frequencies [<5Hz]. The motor's output HP is also dropping.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

- kcress@<solve this puzzle>

 

[Warpspeed]

Agree, cooling is the demon to watch.

With some loads the torque requirement reduces along with speed, and you can sometimes get away with it. If full rated torque is still required at extremely low speed, external independent forced air cooling will probably be required.

 

[Skogsgurra]

Yes, cooling. But if you have enough cooling, there's nothing preventing you from running close to zero.

http://www.reliance.com/prodserv/motgen/b7097_2.htm

discusses some of the issues.

There used to be a problem - and still is with some drives - when running at low speeds. The first/is torque. It simply vanished at low speeds in the earlier (non-vector) drives. So, the recommendation may be from those days. With a rather thick margin added.

The second problem at low speeds was vibrations due to voltage being applied in chunks rather that PWM. That made the motor rather jerky.

I think that you can forget about that today. But keep an eyay on the temperature. Modern drives usually go below 5 Hz with ease.

Gunnar Englund

http://www.gke.org

 

[ozmosis]

It depends on the application. Not sure what a moderately variable load is but if it's a centrifugal fan or pump then yes, it will not draw as much current as a constant torque load (and therefore heat in the motor) but the reality is that the flow from the fan/pump is so insignificant that it is often pointless running below this level anyway.

 

[Beggar]

Well, the motor's been running for several hours and the temperature's not an issue. I think I will add that to the operator's hourly walk-through inspection.

The comments above about resonance are interesting for one of the system characteristics that has "chased" me down to this speed is trying to cope with feedback vibration at a slightly higher speed (~23 Hz with a 50 Hz motor).

Perhaps the root cause of the whole vibration problem is just what you're talking about - trying to run the motor too slow in an application with a highly variable load (variable in both frequency and magnitude).

I know this is taking the thread in a new direction but would you expect the vibration to diminish at a higher motor speed under the same loading conditions?

--------------------
How much do YOU owe?

http://www.brillig.com/debt_clock/

--------------------

 

[motorspert]

depends what the vibration source is. If its simple inbalance then vibration increases with speed. But it will rise through a resonance and drop off as it comes out the other side.
If force has a magnetic source then the magnitude will be dependant on the machine flux which is probably constant with speed.

 

[waross]

What is your load? There are probably people here who can give you specific recomendations for specific applications.
yours

 

[moltra]

On my motors that run below 20hz I install a seperate powered cooling fan on the motor to keep it cool.

 

[Barry1961]

How slow you can run will depend on the type of VFD and the load. A vector drive can run at a much lower speed than a standard volts/hertz drive.

One problem is that if you maintain the volts/hertz ratio the voltage becomes very low at low hertz. With a 240V/60hz motor, 4:1 ratio, the voltage would be 20 volts at 5 hertz. The drives boost voltage and do other black magic stuff at low hertz to counter this. The better the black magic the slower you can run at load.

Barry1961

 

[jraef]

One thing to be careful of is that rotor heating can be much worse than stator heating at low speed, and feeling or measuring the frame is not a good indicator of rotor heat because by the time the rotor heat gets absorbed by the stator and transferred to the frame, the damage is done. So even if it "feels" OK to your operator, that doesn't mean the rotor isn't under thermal tress. I would recommend a separately powered cooling fan unless you have a reasonable method of measuring rotor heat. They are far far less expensive that a rebuild!

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[Skogsgurra]

Absolutely so, Jeff!

I have had bearing failures du to high rotor temperature. An effective way to detect such problems is to check the shaft temperature. Try to keep it below 90 C. The grease starts to drip out of the bearings above that - if you do not have HT grease.

Gunnar Englund

http://www.gke.org

 

[DickDV]

jraef, TENV motors are available in the US at 30hp and smaller. I have used these at full load and zero speed without any apparent problem. Thermal detection has been by klixon switches in the stator windings.

Your point about rotor heating raises my curiousity a bit. The motor manufacturer (Marathon) doesn't offer much in a way of caution about rotor heating. Am I missing something here?

 

[Warpspeed]

I am intrigued by this too.

A totally enclosed fan cooled motor is essentially sealed, and there is no obvious direct path available for rotor cooling. Internal rotor windage, and internal radiation are the only cooling modes that I can see. If approximately half the motor losses occur due to rotor slip, how does the heat normally escape from the rotor?

It must surely mostly be internal windage and subsequent transfer to the outer casing, or am I missing something ?

 

[itsmoked]

I suspect the radiant percentage is actually rather large with the close proximity and cylindrical surround.

Q = rhoT⁴

Q = heat transfer
rho = Stephan-Boltzman constant
T = Absolute temp Rankin (F + 460)

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[Warpspeed]

If in fact radiation is the most significant cooling mechanism, then heat transfer should be quite independent of rotor Rpm.

But if windage is a significant cooling contributor too, any large speed reduction could spell trouble. Fascinating.

 

[jraef]

Well, first off we can't assume he was using a TENV/TEFC motor, he never said. While they are designed to radiate the rotor heat through the shafts and frame (mainly by having longer core lengths and thinner laminations that give up heat faster), the cooling of that transferred heat is still provided by fan assisted convection on the outside, not just radiance from the fins. So my point was if he waits for the frame fins to feel hot when the fan has not been helping, that still means the rotor got a lot hotter than it should have been.

DickDV, does Marathon say that you can run that motor like that? Just because you did and got away with it doesn't mean it was designed for that. In any application, the load is what determines heating anyway. If, as Warpspeed said early on, there is little load at that low speed, then cooling may not be an issue at all.

Anyway, go ahead and run your TENV motors at zero speed and no fan cooling. If it works, great. If not, it will end up being a lot more expensive than that small fan. I just recommend erring on the side of caution if it's important to operate at low speeds like that for prolonged times.

Eng-Tips: Help for your job, not for your homework Read faq731-376