No-Load Soft Start Operation 2

[sry110]

I'm a gear-head dabbling in soft starters and I need help!

We make a Turning Gear System which consists of the following: a 7.5 HP AC motor (460VAC / 3-ph / 60Hz, 1800 RPM) driving into a double-reduction worm gear reducer with 173.1:1 overall speed reduction ratio, with a shear pin coupling (mechanical fuse) and an overrunning clutch on the output shaft of the speed reducer gearbox. The clutch ultimately connects / disconnects our system from the customer's shaft, which is a very large diameter (i.e. high intertia load) induced draft fan in a power generating station.

The system is used to start the customer's shaft turning from rest (i.e. breakaway from rest). The Clutch we use requires 15 degrees of shaft rotation before it engages. Through our 173.1:1 gear ratio, that is 7.2 revolutions of the 7.5 HP motor. During this 7.2 revolutions the motor is essentially unloaded (loaded only by the torque required to turn our gear elements, which is negligible) so it is up to full speed and when the clutch engages, it slams into engagement with the Customer's stationary shaft which creates a shock load that damages components in our system.

So, we need a soft-start motor controller that will force the voltage (and therefore torque) to ramp over a set period of time regardless of the load on the motor. This way we know that when the clutch engages, while it may be at full speed it is only at a fraction of full torque. I know most soft starters will go to across the line voltage as soon as the motor is up to full speed and no load, but that does us no good since the first 7.2 revolutions are unloaded. We need it to always go through its voltage / torque ramp.

Does such an animal exist?

 

[itsmoked]

If you tell a softstarter to take 20 seconds before DOL time will that give you enough time?

Otherwise just use a VFD and let it creep the motor right up to the no slack point then smoothly accelerate from there. You can even get VFDs that have mini PLC front-ends that would let you use logic inputs or timers to run the motor in different modes. Like you could have the motor start at 20RPM for the first 10 seconds then execute an S-curve run up to full speed.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[sry110]

Hmm ok, so that takes care of the speed issue, but what does the VFD do in the way of voltage / torque limiting (if anything?) Would I then be hitting the clutch engagement point with fractional speed and full locked rotor torque (or whatever the full rated torque is at that given speed on the Torque vs. Speed curve)?

 

[itsmoked]

Many drives have adjustable torque limits too.

Does it actually matter if the motor is delivering full torque if the speed is only a few percent of running speed? The inertial impulse seen by the load will be very low at a low speed. You may actually need available torque to provide some breakaway for the machine. With a VFD you probably could ditch the clutch and the shear pins to boot. Let the drive deal with jams.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jraef]

It's true that most voltage ramp soft starters will automatically go across the line when the motor gets to full speed, so if it is unloaded you will likely be in trouble by the time your clutch engages. That is essentially inherent in the design of most voltage ramp controllers. They "push" voltage down (retard the phase angle) and monitor current in a feedback loop, decreasing the retardation over time. If the current drops before the time expires however, i.e. the motor gets to speed, they stop retarding the phase angle and go X-line. If they don't, the motor will oscillate at the end of ramp, sometimes to the point of failing to fully accelerate. Years ago some soft starters had this problem, most of them went out of business. The survivors all adopted what was called an "anti-oscillation" circuit, which was that feedback loop I just described. As far as I know, all voltage ramp soft starters on the market now have this (they are all offshoots of the same few survivors, but that's another story).

A soft starter that uses PF instead of current in the feedback loop however will not necessarily do that because the PF stays low when the motor is unloaded. They tend then to be better at remaining linear in their ramp than a voltage ramp starter. Not perfect, but better. This is often referred to in marketing materials as "Torque Ramp". A soft starter with that feature may work for what you want. Not all soft starters are capable of this, you have to ask. Unfortunately the quality of one mfrs "torque ramp" circuit may not be the same as another, so it may take some trial and error to determine a workable solution. Make sure you explain exactly what you expect and insist that if it doesn't work, they will take it back. Reputable suppliers will agree to that.

There is also a way to "force" some high-end soft starters into a purely linear ramp regardless of load, but it involves using a tach feedback loop. The combination of that external hardware along with the high-end soft starters that can do that would make it more expensive on a 7-1/2HP motor than a VFD as Keith suggested.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[Marke]

Hi jraef

voltage ramp soft starters will automatically go across the line when the motor gets to full speed

There are two modes of phase control used in soft starters, 1) phase control relative to the voltage zero crossing and 2) phase control relative to the current zero crossing.
Both methods have their strengths and weaknesses. Some soft starters begin in voltage zero crossing reference and then switch to current zero crossing referenced.
The simplest system is the voltage zero crossing only and most of the early soft starters used this method only. The problem with voltage zero crossing is that close to full speed, there can be some major oscillations and in order to overcome these, many suppliers step to full firing.
Soft starters that are set to current zero crossing during the later part of the ramp do not suffer this problem and do not step to full firing, but ramp for the full period.
This technology is also more effective at soft stop.

There are certainly soft starters available that will operate as required without stepping.

Best regards,
Mark.

Mark Empson
L M Photonics Ltd

 

[jraef]

Mark,
If ever I talk myself into getting into the business of making a soft starter on my own, I'm going to hire you.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[waross]

One point to mention; I believe that it is the inertia of the motor rotor at full speed that is causing your shock loading. Your solution is not to try to limit torque but to limit the speed ramp. Believe that this has been assumed y previous posters.
As has been suggested, VFD and set the speed ramp.
Or
Use a primary resistor starter with a high resistance in the first step. The motor will start to roll over until it takes up the slack in the drive train and then stall. Shortly thereafter it will go DOL.
Possible but not suggested.
And, some older shrimp sizing machines had issues with starting shocks eventually stretching the drive chains and butting the sizing roll out of time. The issue was completely resolved with VFDs. These were about 2 HP.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[DickDV]

As already described, a softstarter will limit torque when starting but, in your case, even with very low torque for the first few revolutions, the motor will tend to run up to full speed causing a slam-start condition when the clutch engages. The rotating mass of the motor rotor thru the gear reducer will provide plenty of torque to stress the components.

At only 7.5hp, I would use a VFD instead and start with a very low fixed speed for a set time (enough time to engage the clutch) and then run up to whatever running speed is required. Being a draft fan, I would think that full speed would not always be required and significant energy savings would result by using the VFD to slow down when demand is low.

That way you get a nice soft start and energy savings too. If you really need full speed redundancy, you could bypass the drive as well but, at 7.5hp, the bypass would probably cost more than another drive.

Hope this provides some food for thought.

 

[waross]

Hi Dick;
This is turning gear, the fan is turned very slowly, about 10 RPM, probably to bring the lube oil up into babbit bearings with oil rings or chains before going to full speed with the main drive motor. Possibly for a cool down cycle.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[DickDV]

Hi, Bill. So, clearly, energy savings is not practical but, at the low price of 7.5hp drives, I'd still use one to hold the low speed until the clutch engages.

 

[LionelHutz]

It's a 7.5hp motor. Buy a decent VFD and you will get a device that has all the features necessary to do what you want. A good VFD will aallow you to limit the torque and allow you to set a slow speed run and then ramp up to full speed. You can also conbine a little cheaper VFD with a cheap smart relay (SquareD Zelio or Siemens Logo?) to toggle the digital inputs (jog, torque limit etc) on a timed basis to get it going.

A motor with no load will tend to quickly accelerate towards full speed when started with a soft-starter, no matter what type or the settings you try. It will be very difficult to get a slow rotation for the first 7 revolutions and then crank the torqueto accelerate the fan. The only possible way I see this working is a unit with a true jog or slow speed cyclic control that would run the motor at a very slow speed for the first bit. At any rate, any good soft-starter will be similarly priced to a VFD.

 

[sry110]

Thanks everyone for all of your helpful replies. We put a few of these out in the field and I think our findings align well with what you are all saying: when the motor is unloaded, the Torque Limiter (dummy soft-starter) does not behave as advertised - it basically just gets up to full speed. We put an oscilloscope on the leads and did some test runs and found that once the clutch engaged and the load was seen by the motor, the TL unit started to behave as expected, i.e. ramp the voltage over time. But we are still impacting the shear pins at full speed. Our field guy actually observed that the pins yielded slightly which I believe is due to the inertial impact.

So now I'm considering a VFD. I'm convinced now, from a mechanical standpoint, that for my system I need to control speed and not torque. The torque is calculable, because I know the Locked Rotor and Breakdown torque output of the motor, and I know the efficiency of my gearbox. What I don't know is how much "Impact Factor" to multiply the torque by to account for full-speed engagement.

From what I have read here, it seems that a VFD will control the speed regardless of motor load. For example, given the 7.2 revolutions (45.3 radians) of motor shaft rotation needed to engage the clutch and neglecting motor accelerating time (i.e. assume constant speed from time zero), and if I set the VFD to run at 200 RPM, I would need to let it run for 2.16 seconds until it engages. So to build in some margin I would set it to 2.5 seconds, then ramp quickly to full speed / full voltage to achieve breakaway. Seems to me that this will remove almost all of the impact from the system because I will be going 1/9th the speed, and since flywheel energy goes as the square of the speed I'm getting 1/81st of the impact torque (I think).

Sounds reasonable as far as whether the VFD will support this operating scheme?

 

[itsmoked]

I would first weld that idiot clutch solid. And try your existing SS. If it doesn't cut it then I would use a VFD knowing that it would absolutely work. Then leave the motor connected and spinning. If you then want to "disconnect" just have the VFD stop driving the motor. Leaving the motor spin won't hurt anything. If you want to use the motor to stop the fan, just use a VFD that can "catch" a spinning motor and return to control. Include a breaking resistor if you want to brake the fan.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

Keith,

Without the clutch the main drive will try to drive the 10rpm worm reduction output at whatever speed the main motor rotates at, which is probably not mechanically possible without destroying the gearbox. The clutch will have to stay!


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

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

Dang!

Thanks Scotty.

How big is this fan?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Hevs]

most inverters have a soft start, ramp up and ramp down programme built in, maybe try this.

Hevs
Conveyor Manufacturers

http://www.conveyor-manufacturers.co.uk