Brushless servo drives and MCR contactors


Thread Starter

Bill Sturm

I have always used a contactor to cut the power to a servo drive in an E-Stop situation. I am questioning the need to do this with an AC or
brushless DC drive. If the drive fails, will it run away? If an output transistor is shorted closed, the motor will fail to commutate and will
lock into one position. The older brush style motors could certainly run way if the drive failed, but I think that BLDC motors are much safer.

I would prefer to only disable the drive with my e-stop circuit instead of cutting power to the drive. The drive could be programmed for a controlled fast stop when it is disabled. I think that this is safe, and I am generally pretty cautious.

I have seen CNC systems without the traditional contactors and the vendor stated that with modern drives and digital electronics with error correction and watchdogs that these systems are not prone to run away, even if a drive fails. Is this true?

What are others doing in the same situation?


Bill Sturm

Bouchard, James [CPCCA]

The contactor is used not only to avoid run away situations but to make sure that when you press the E-Stop the motor stops. In relying on the electronics you cannot be sure that removing the enable will stop the drive ( I have see situations where the drive continued to run, under
control, when the enable was removed ) We require contactors on all drives and I would be very uncomfortable if asked to remove them. It takes only one incident to ruin your life. In some cases where we want to make use of electronic braking, we use a time delay on the contactor to drop it out after the drive has come to a stop ( usually only a few seconds ) the time delay is done through a safety relay that is designed for this type of application.

James Bouchard

Stefaan Van Eenoo

<P>In such a situation I would use the disable contact of the drive to do a controlled (fast)stop of the engine. For safety reasons, I also cut the power of the drive with two serial place
Of-Delay timers.</P>

----||----------------------------------( )---
E-Stop | Off
| Delay 1
----------------------------( )---
Delay 2

----||--------||------------||----------( )---
E-Stop OF-Delay 1 Of-Delay 2 Relay
The reason I work like this is that if the operator is coming in a place where
he can touch the motor of can be injured by the motor will come in a situation
which created a E-Stop ( for instance if you open a monitored door ). This way
my motor will do a fast stop and afterwards will be cut of voltage.</P>

<P>Please feel free to comment my safety-feeling</P>

<P>Kind Regards,<BR>
Stefaan Van Eenoo<BR>
Best practice and CE requirements are that you use a contactor to open the drive power in the event of an e-stop condition. All estop functions must be hardwired, and in some case self checking, etc. The best way to estop a drive
is to use a safety relay with a built in time delay function. Doing so allows one contact to trigger the drive to fast-stop in a controlled manner and the delayed contacts open a short time ie. 3-5second later triggering the contactor to open. This is the best situation for the drive as well as safety concerns. It should be noted that you cannot use just any timing relay, but need to use a safety relay with the checking, etc built in. Most major electrical manufactures SquareD, AB, Etc have a safety relay for this use.
If not because of transistor shorts or drive failure, brushless motors can still runaway for others reasons, especially when driven with digital drives with hybrid Hall/encoder commutation and controlled by integrated or
external motion controls. Some reasons:
- Swapping of the sequence/phase sensitive Hall or motor phase wires
- Swapping of the encoder polarity (wires or parameter) as seen by the motion controller (positive feedback runaway)

Example: user panics, starts swapping parts and makes a re-wiring mistake; or someone changes the encoder polarity parameter.

You should get key advice from the manufacturer of the specific drive & controller model under consideration. If the drive has a way to do
controlled internal braking on a runaway condition, this may be an alternative.

I'm not a machine designer, but maybe OSHA etc. have specific requirements for E-stop and exactly what it does during a dangerous condition too.

Bob Peterson

What I have done is to have the estop circuit cause a controlled stop as fast as possible, and after a few seconds trip the power contactor to the drive. I have generally used a pneumatic off delay timer, set for a few seconds (however long a controlled stop is likely to take). The machines are guarded to prevent personal injury, and a lot of times an uncontrolled stop is not an
ideal situation.

You still get the benefits of a hard stop if the controlled stop fails (a highly unlikely, but still possible situation). I tend to agree that modern drives are probably not going to cause a runaway problem, and in some cases machine damage can occur from uncontrolled stops. And the machine may actually come to a stop faster with the drives live then with them unpowered.

I consider this akin to a tool relief situation where a tool is partially withdrawn when the estop is tripped, and power drops off 100 or 200 msecs later. Otherwise the tool might get embedded in a part.

Bob Peterson

Michael Griffin

Whether or not it is safe or "safe enough" depends upon the design of the drive, and the risks involved in the application. I don't think you can make a blanket statement about all drives. Rather, you would need to know whether
the particular one you are using has a fail safe means of disabling it, and if not, then what the consequences are of a failure in your application.

If your requirement is to allow a controlled stop (rather then just saving a
contactor), then most safety relay manufacturers have relay versions with timed circuits. These give you both immediate and delayed drop-out for
this application. The amount of time you have to stop the drive will depend upon the distance from the guard to the point of hazard, or you can install a solenoid lock.

Michael Griffin
London, Ont. Canada


James Ingraham

As everyone else has said, DON'T DO IT. Always, always, always cut the power to the motor when you are in e-stop. The pneumatic time delay is a good option; that's what my company uses as well. With most drives, the motor power is seperate from the control power so that you don't have to re-home the motor everytime you e-stop. There is no reason to keep power on the motor in e-stop.

Sage Automation, Inc.
Actually, I have always been safe and used an input contactor to my servo drives, however the drives that I am using shut down completely which
forces me to re-home the system after an e-stop. I was hoping to avoid this, but the advice from the list is unanimously in favor of using the hard wired power contactor, so of course I will.

I have used the pneumatic delay in the past on high speed lines with much inertia, this gives a chance to fast stop before cutting power. My current project moves slowly with enough friction that cutting power will cause a quick stop. There is actually plenty of residual power in the bus to decel even after the power is cut. I plan to skip the time delay circuit in this case. I will simply cut power and disable the drive upon e-stop.

I remember the first time I designed an AC drive circuit and had an input MCR contactor and no fast stop input or drive disable upon e-stop. The motor continued to run for several seconds after I cut the power to it. That was interesting...

Thanks for all of your advice,

Bill Sturm
Is there any reason not to put the contactor between the drive and the motor, so that E-Stop kills the actuator but not the feedback? That's what I thought the "DON'T DO IT" comment meant.

Paul T
Good question,

In the past, I worked with large DC drives and we usually put an output contactor on the drive. In AC drives and servos, I usually see input
contactors. I am afraid to open up the output circuit under load, it may cause damage ??? It is also more expensive because you need one contactor per drive.

Bill Sturm
It's definitely more expensive if you need one contactor per drive. As far as opening the drive under load, you can use an aux contact on the
contactor to disable the drive output. I'm interested in this from a safety standpoint. NFPA 79 9.6.3 requires that an E-stop have "only hardwired electromechanical components." An electromechanical component is one in which "electrical energy is used to magnetically cause movement." Now, 9.6.3 also forbids dependence on electronic logic, either hardware or
software. Yesterday, as a matter of fact, I was looking at some Reliance drives (SP-120's) that have five programmable inputs. One of those can be
configured as a stop; I haven't found an enable. If I want to meet NFPA 79, how do I configure the motor circuits for E-stop? With servos, we always
try and use a contactor between the drive and the motor so that E-stop kills the actuator but not the encoder feedback. Even tho the VFD doesn't
have an encoder, I would still like to leave the drive electronics powered but disconnect the motor. Does anyone have any thoughts on normal practice for E-stops controlling VFD's? This application is for a large conveyor system - about 120 VFDs. Fortunately, they are primarily fractional HP motors... but there are a lot of them <g>.

Paul T