High Voltage Stepping Motor

ld up current fast in the winding, for good transient and speed performance )

uld need something like a 35V stepping motor, but it needs to be able to ha ndle peak voltages of 400V (functional isolation)

The limitation seems to be the enamel insulation on the wire used to wind t he motor - typically about 500V.

The motor manufacturer should be able to tell you.

Back in the late 1970's I frightened some mechanical engineers half to deat h by proposing to use 60V to drive the windings of a nominally 5V stepper m otor.

It took quite a while - and a meeting with the agents who sold us the motor - to persuade them that 5V was what would cook the motor if you hooked one or more windings to a power supply and made sure that the rotor wasn't rot ating and generating back-emf.

Anybody who wants to get useful power out of a stepper motor drives the win dings at a much higher voltages than the nominal rating.

If you are pulse width modulating a 400V source, it would probably be a goo d idea to put some inductive filtering close up against the source. Otheris e the motor leads will transmitting a +/-400V signal at a couple of hundred kilohertz, which can worry the neighbours.

uild up current fast in the winding, for good transient and speed performan ce)

would need something like a 35V stepping motor, but it needs to be able to handle peak voltages of 400V (functional isolation)

the motor - typically about 500V.

Yes. The wire is going to be really thin, so it may cost more to produce du e to lower yield and production challenges

ath by proposing to use 60V to drive the windings of a nominally 5V stepper motor.

or - to persuade them that 5V was what would cook the motor if you hooked o ne or more windings to a power supply and made sure that the rotor wasn't r otating and generating back-emf.

indings at a much higher voltages than the nominal rating.

We're doing thay already, driving the 2.5V stepper at 30V. There's no other way getting it to work

ood idea to put some inductive filtering close up against the source. Other ise the motor leads will transmitting a +/-400V signal at a couple of hundr ed kilohertz, which can worry the neighbours.

Yes. On the upside, the wiring is encapsulated in metal housing, so it may no be a problem at all, but the common mode noise is probably higher since the capacitive coupling is simelar to the low voltage version, but the need ed dV/dt is higher.

On the positibe side, the removal of the front end SMPS removed one noise s ource entirely

Cheers

Klaus

Steppers can produce torque up to a point where the back EMF exceeds the supply voltage, or the inductance prevents the coil current from ramping up before the current needs to be reversed for the next step, whichever comes first.

Raising the supply voltage helps with both of these limits, thus allowing the torque knee to be moved up. But, you also have to consider heating. There is copper loss, mostly due to plain DC resistance, but also a bit of eddy loss in the wires. Then there is iron loss, and the more times the current in the coils rises and falls per unit of time, the more heating you get. Stepper motors are really bad in this regard. Spin a typical stepper motor with an electric drill, and you will detect iron loss, even WITHOUT any coil current! This limits the upper speed, especially in jobs with the motor spinning a lot of the time.

So, while you might get somewhat better performance at 400 V, without a lot of sophistication in the chopping scheme and current recirculation, the motor might run extremely hot, maybe even without moving at all! So, raising the voltage in great leaps will have diminishing returns.

Jon

Can you elaborate on the current recirculation?

And the reason why the higher voltage would motor would have larger loss?

I mean, the voltage is higher, but the current is lower, so switching losses of the driver is probably about the same. The resultant field strength is also the same. Please correct me if I am wrong...

Cheers

Klaus

motors respond to current, so that's no less. Delivering a given i*t in a narrower pulse increases copper loss and iron loss.

NT

I have played with steppers in the past (got big 1.4deg steppers running at 1500rpm) but nowadays I'd use a servo motor and encoders. That solves a lot of problems regarding missing steps and speed.

If you have to put more turns on the motor windings in order to run from a higher supply voltage, then you will not have any speed advantage because the inductance will go up with the square of the turns count.

The only possible saving when you add turns and run straight from 400V would be the possibility to run straight from rectified mains and save on a power supply. I think this will not be worth it except for very large motors, because the extra winding insulation in the motor that is needed to deal with higher voltages will reduce the motor performance somewhat.

If you are talking about running the 2.5V motor from 400V, that would likely heat the laminations of the motor excessively because the eddy currents in the laminations are proportional to the voltage induced in the loops that can be drawn within the lamination, and that will be proportional to the voltage applied to the windings. Although you would only need to apply very narrow pulses of voltage to the windings to achieve the rated current, the losses in the laminations would be excessive because the loss is proportional to the square of the induced voltage in the iron.

Chris

The high voltage is just used to increase the current rise time, not to increase the RMS current. There may be iron losses, of course, but the high speed that is possible makes it easy to add an effective fan...

The high voltage is just used to increase the current rise rate...

I mistakenly posted 'increase rise time', it should be 'decrease rise time' or 'increase rise rate', of course.

uild up current fast in the winding, for good transient and speed performan ce)

would need something like a 35V stepping motor, but it needs to be able to handle peak voltages of 400V (functional isolation)

You could improve the situation considerably by low-pass filtering the very narrow pulse outside the motor. They'd become wider, lower-voltage pulses and the induced voltages in the "iron" - not much used in small stepper mot ors - would be correspondingly smaller.

It's a good point, but an intelligent designer would do something about it before a motor got melted, and most designers would melt only one.

build up current fast in the winding, for good transient and speed perform ance)

I would need something like a 35V stepping motor, but it needs to be able t o handle peak voltages of 400V (functional isolation)

d?

ry narrow pulse outside the motor. They'd become wider, lower-voltage pulse s and the induced voltages in the "iron" - not much used in small stepper m otors - would be correspondingly smaller.

t before a motor got melted, and most designers would melt only one.

Well, if you low pass filter the the pulse, then the current won't rise fas t and the maximum speed of the motor is reduced

Cheers

Klaus

Yes, me too if possible. I had an application where a big stepper needed a lot of torque but only for a small proportion of a rotation. Trouble was, you needed enough current to cope with this peak all the time, there was no easy way to switch it.

The factory manager was less than impressed by the smoke.

Cheers

to build up current fast in the winding, for good transient and speed perfo rmance)

o I would need something like a 35V stepping motor, but it needs to be able to handle peak voltages of 400V (functional isolation)

oad?

rom

is

ld

ed

very narrow pulse outside the motor. They'd become wider, lower-voltage pul ses and the induced voltages in the "iron" - not much used in small stepper motors - would be correspondingly smaller.

it before a motor got melted, and most designers would melt only one.

ast and the maximum speed of the motor is reduced

At first sight it looks self defeating, but it could be used to enable a di rect from mains drive.

NT

to build up current fast in the winding, for good transient and speed perfo rmance)

o I would need something like a 35V stepping motor, but it needs to be able to handle peak voltages of 400V (functional isolation)

oad?

rom

is

ld

ed

very narrow pulse outside the motor. They'd become wider, lower-voltage pul ses and the induced voltages in the "iron" - not much used in small stepper motors - would be correspondingly smaller.

it before a motor got melted, and most designers would melt only one.

ast and the maximum speed of the motor is reduced.

Well, the context is very narrow pulses. All you care about is the integrat ed voltage over the time you need the current to change, and if you low-pas s filter the pulses only to the point where they overlap appreciably, the i ntegral of the voltage peaks doesn't change.

The rate of change of current in the coils does, and the currents induced i n the magnetic core, but only down to the point where you get the change in magnetic flux you want in the period over which you want it.

The post I was responding to was talking about 400V pulses into a 2.5V moto r.

Filtering that kind of pulse before it gets into the leads to the motor wou ld be a good idea. Don't generate any more electromagnetic interference tha n you have to.

Some people have the idea that motor coils are pretty inductive, so it's st upid to pay for extra inductance in an external filter, but it can turn out to be a worthwhile investment.

(to build up current fast in the winding, for good transient and speed per formance)

So I would need something like a 35V stepping motor, but it needs to be ab le to handle peak voltages of 400V (functional isolation)

road?

from

t.

ave

t is

ce

ld

dy

in

ould

uced

e very narrow pulse outside the motor. They'd become wider, lower-voltage p ulses and the induced voltages in the "iron" - not much used in small stepp er motors - would be correspondingly smaller.

ut it before a motor got melted, and most designers would melt only one.

fast and the maximum speed of the motor is reduced.

ated voltage over the time you need the current to change, and if you low-p ass filter the pulses only to the point where they overlap appreciably, the integral of the voltage peaks doesn't change.

in the magnetic core, but only down to the point where you get the change in magnetic flux you want in the period over which you want it.

Yes, that's a nice idea, provided the cost of the filter won't outdo the po wer supply price

Cheers

Klaus