Stepper Motor Controller design problems

[Ben]

Hi,

I need help. I have been in electronics design for almost 20 years but this
is getting me down.

I had to design a stepper controller for a linear actuator from Haydon. It
has a worm screw and a nut and it is used to lift a panel to exact
positions.
I have a controller/driver chip like PBL3717 and I drive it from a
Micdoprocessor .
The controller chip uses a chopper drive with the 5V motor running off 40V @
I max=560mA
Motor is 200 step/rev bipolar.

I am stepping in fullstep mode with 2 coild nergized at all times.
The 4 step phases are:
A1A2 B1B2
0 1 0 1
0 1 1 0
1 0 1 0
1 0 0 1

I am stepping batches of 400 steps at a time and I repeat this over and
over.

I use a accelerate and decelerate routine as follows:

Step 0-23 = accelerate 0 to 635 stpes/sec
Step 24 - 374 = run at 635 steps/sec
Step 374 - 399 = decelerate to - and then stop.

Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,
suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.
This is NOT caused by the drive sequence or anything in the pulses. In fact,
there is no electrical difference at all in the driving. The motor just does
his. The next batch will be ok again. Sometimes it works fine for 30 an
dmore batches and suddely it does it again.

Has anyone come across a problem like this and how is it best solved?

I tell you, stepper motors are not simple when resonance, acceleration and
exact position all come into it simultaneously!


Regards
Bernt

 

[Paul Burke]

Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,
suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.
This is NOT caused by the drive sequence or anything in the pulses.

Stepper motors don't do ANYTHING without the pulses. Have you got high
frequencies on the pulse outputs, superimposed on the normal stepping
sequence?

 

[[email protected]]

Hi,

I need help. I have been in electronics design for almost 20 years but this
is getting me down.

I had to design a stepper controller for a linear actuator from Haydon. It
has a worm screw and a nut and it is used to lift a panel to exact
positions.
I have a controller/driver chip like PBL3717 and I drive it from a
Micdoprocessor .
The controller chip uses a chopper drive with the 5V motor running off 40V @
I max=560mA
Motor is 200 step/rev bipolar.

I am stepping in fullstep mode with 2 coild nergized at all times.
The 4 step phases are:
A1A2  B1B2
0  1     0   1
0  1     1   0
1  0     1   0
1  0     0   1

I am stepping batches of 400 steps at a time and I repeat this over and
over.

I use a accelerate and decelerate routine as follows:

Step 0-23 = accelerate 0 to 635 stpes/sec

This would seem to be excessive, typicaly a motor this size can reach
speed in 5 steps. When you shorten the steps remember the acceleration
profile should not be linear.

Step 24 - 374 = run at 635 steps/sec
Step 374 - 399 = decelerate to - and then stop.


Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,

suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.

Sounds like resonance, do you have the load connected? You must not try
to run the system with the motor lying loose on the benchtop.

 

[default]

Hi,

I need help. I have been in electronics design for almost 20 years but this
is getting me down.

I had to design a stepper controller for a linear actuator from Haydon. It
has a worm screw and a nut and it is used to lift a panel to exact
positions.
I have a controller/driver chip like PBL3717 and I drive it from a
Micdoprocessor .
The controller chip uses a chopper drive with the 5V motor running off 40V @
I max=560mA
Motor is 200 step/rev bipolar.

I am stepping in fullstep mode with 2 coild nergized at all times.
The 4 step phases are:
A1A2 B1B2
0 1 0 1
0 1 1 0
1 0 1 0
1 0 0 1

I am stepping batches of 400 steps at a time and I repeat this over and
over.

I use a accelerate and decelerate routine as follows:

Step 0-23 = accelerate 0 to 635 stpes/sec
Step 24 - 374 = run at 635 steps/sec
Step 374 - 399 = decelerate to - and then stop.

Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,
suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.
This is NOT caused by the drive sequence or anything in the pulses. In fact,
there is no electrical difference at all in the driving. The motor just does
his. The next batch will be ok again. Sometimes it works fine for 30 an
dmore batches and suddely it does it again.

Has anyone come across a problem like this and how is it best solved?

I tell you, stepper motors are not simple when resonance, acceleration and
exact position all come into it simultaneously!


Regards
Bernt

Mechanical resonance will cause that - particularly at close to
maximum torque - running backwards. Try to change the load or
damping.

 

[[email protected]]

Hi,

I need help. I have been in electronics design for almost 20 years but this
is getting me down.

I had to design a stepper controller for a linear actuator from Haydon. It
has a worm screw and a nut and it is used to lift a panel to exact
positions.
I have a controller/driver chip like PBL3717 and I drive it from a
Micdoprocessor .
The controller chip uses a chopper drive with the 5V motor running off 40V @
I max=560mA
Motor is 200 step/rev bipolar.

I am stepping in fullstep mode with 2 coild nergized at all times.
The 4 step phases are:
A1A2 B1B2
0 1 0 1
0 1 1 0
1 0 1 0
1 0 0 1

I am stepping batches of 400 steps at a time and I repeat this over and
over.

I use a accelerate and decelerate routine as follows:

Step 0-23 = accelerate 0 to 635 stpes/sec
Step 24 - 374 = run at 635 steps/sec
Step 374 - 399 = decelerate to - and then stop.

Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,
suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.
This is NOT caused by the drive sequence or anything in the pulses. In fact,
there is no electrical difference at all in the driving. The motor just does
his. The next batch will be ok again. Sometimes it works fine for 30 an
dmore batches and suddely it does it again.

Has anyone come across a problem like this and how is it best solved?

I tell you, stepper motors are not simple when resonance, acceleration and
exact position all come into it simultaneously!

Do you know what the resonant frequencies for the motor are? Most of
the motors I worked on has a first resonance - the product of the
moment of interia of the rotor and the spring constant of the magnetic
field - at around 100Hz, comfortably below your 635 steps a second.
There is almost always a second resonance due to the spring constant of
the shaft connecting rotor to to its load - here, your worm screw. In
some linear actuators the rotor is the nut driving the worm screw, and
the spring constant of the worm screw would then be the corresponding
factor - complicated by its tendency to change as it feeds through the
motor.

In one application, I ran into another interesting trap for the unwary
- the motor was being asked to run fast enough that the back-emf in the
windings was almost as high as the drive voltage, which produced some
very interesting coil current waveforms.

We identified the problem by driving the stepper motor shaft with a
variable speed electric dril and monitored the coil voltages with an
oscilliscope. Solving it was tolerably straightforward - the nominally
5V motor was replaced by the 2.5V version, and all the power tranistors
in the drivers got changed for appreciably beefier parts ...

Good motors show the back-emf factor in the data sheet, and good
designers read the whole data sheet - but "goodness" seems to be as
rare in the data sheets as it is in the designers.

It might be worth checking to see if a 40V drive is high enouhg to
drive your motor at 635 step/sec.

 

[John B]

Do you know what the resonant frequencies for the motor are? Most of
the motors I worked on has a first resonance - the product of the
moment of interia of the rotor and the spring constant of the magnetic
field - at around 100Hz, comfortably below your 635 steps a second.
There is almost always a second resonance due to the spring constant
of the shaft connecting rotor to to its load - here, your worm screw.
In some linear actuators the rotor is the nut driving the worm screw,
and the spring constant of the worm screw would then be the
corresponding factor - complicated by its tendency to change as it
feeds through the motor.

In one application, I ran into another interesting trap for the unwary
- the motor was being asked to run fast enough that the back-emf in
the windings was almost as high as the drive voltage, which produced
some very interesting coil current waveforms.

We identified the problem by driving the stepper motor shaft with a
variable speed electric dril and monitored the coil voltages with an
oscilliscope. Solving it was tolerably straightforward - the nominally
5V motor was replaced by the 2.5V version, and all the power
tranistors in the drivers got changed for appreciably beefier parts
...

Good motors show the back-emf factor in the data sheet, and good
designers read the whole data sheet - but "goodness" seems to be as
rare in the data sheets as it is in the designers.

It might be worth checking to see if a 40V drive is high enouhg to
drive your motor at 635 step/sec.

You could try half-stepping the motor or adding a damper to the shaft.

This is an excellent reference book for all stepper motor problems:

<http://www.amazon.co.uk/Stepping-Motors-Practice-Control-Engineering/dp
/085296417X/sr=8-1/qid=1160141406/ref=sr_1_1/026-2838258-5495621?ie=UTF8
&s=books>

Rather long, I hope it wraps OK.

 

[colin]

Hi,

I need help. I have been in electronics design for almost 20 years but this
is getting me down.

I had to design a stepper controller for a linear actuator from Haydon. It
has a worm screw and a nut and it is used to lift a panel to exact
positions.
I have a controller/driver chip like PBL3717 and I drive it from a
Micdoprocessor .
The controller chip uses a chopper drive with the 5V motor running off 40V @
I max=560mA
Motor is 200 step/rev bipolar.

I am stepping in fullstep mode with 2 coild nergized at all times.
The 4 step phases are:
A1A2 B1B2
0 1 0 1
0 1 1 0
1 0 1 0
1 0 0 1

I am stepping batches of 400 steps at a time and I repeat this over and
over.

I use a accelerate and decelerate routine as follows:

Step 0-23 = accelerate 0 to 635 stpes/sec
Step 24 - 374 = run at 635 steps/sec
Step 374 - 399 = decelerate to - and then stop.

Problem:
Motor starts runs and stops correctly but as I repeat the batches of 400,
suddenly, RANDOM (!) motor starts and runs at a higher (audible buzz) speed
in the opposite direction.
This is NOT caused by the drive sequence or anything in the pulses. In fact,
there is no electrical difference at all in the driving. The motor just does
his. The next batch will be ok again. Sometimes it works fine for 30 an
dmore batches and suddely it does it again.

Has anyone come across a problem like this and how is it best solved?

I tell you, stepper motors are not simple when resonance, acceleration and
exact position all come into it simultaneously!

You sometimes have to identify where the resonance is and avoid stepping at
that frequency,
It might be that you are accelerating through the point of resonance too
slowly.

Colin =^.^=

 

[Johnny Boy]

40V

You sometimes have to identify where the resonance is and avoid stepping at
that frequency,
It might be that you are accelerating through the point of resonance too
slowly.

Colin =^.^=

I was going to suggest that maybe he's ramping the motor up too quickly, but
you have a good point. Ramping through the point of resonance more quickly
has cured my troubles in the past. But I've never had a motr run backwards
under these conditions. They've just stopped spinning and sat there buzzing.
.... Johnny

 

[colin]

I was going to suggest that maybe he's ramping the motor up too quickly, but
you have a good point. Ramping through the point of resonance more quickly
has cured my troubles in the past. But I've never had a motr run backwards
under these conditions. They've just stopped spinning and sat there

buzzing.

Yes it is a bit strange but possible, if its suddenly running backwards at
twice speed it means it has plenty of power to spare to accelerate more
quickly.

Colin =^.^=

 

[Tim Wescott]

buzzing.

Yes it is a bit strange but possible, if its suddenly running backwards at
twice speed it means it has plenty of power to spare to accelerate more
quickly.

Colin =^.^=

I've had the worm-gear type steppers run backwards like that if they
were lifting the load -- the upset of the step sequence on the coils
(and possibly some harmonic of said sequence) would make the thing go
essentially frictionless and fall down.

It was a consequence of trying to push the motor too fast, with too
little time in the project to get things right.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google? See http://cfaj.freeshell.org/google/

"Applied Control Theory for Embedded Systems" came out in April.
See details at http://www.wescottdesign.com/actfes/actfes.html

 

[colin]

I've had the worm-gear type steppers run backwards like that if they
were lifting the load -- the upset of the step sequence on the coils
(and possibly some harmonic of said sequence) would make the thing go
essentially frictionless and fall down.

It was a consequence of trying to push the motor too fast, with too
little time in the project to get things right.

yes step too fast too quickly and it becomes useless,
however it seems as though his is actually 'stepping' backwards,
theres a lot to be said for positional feedback servos.

Colin =^.^=

 

[Jim Thompson]

[snip]

I've had the worm-gear type steppers run backwards like that if they
were lifting the load -- the upset of the step sequence on the coils
(and possibly some harmonic of said sequence) would make the thing go
essentially frictionless and fall down.

It was a consequence of trying to push the motor too fast, with too
little time in the project to get things right.

I've not done any stepper controller designs, but about ten years ago
I did a design for a 3-phase EC motor for Bosch (Mercedes A/C blower).

The squirrel cage would spin backwards when the outside vent was open
and the motor was not driven. Thus it would run backwards when
powered up under those conditions.

I came up with a cute way to cure that. I'll look through my notes.
(I even have a video tape of me driving the cage backwards using a
leaf blower, then applying power, and it slows down to a stop and
starts running in the correct direction.)

...Jim Thompson

 

[Tim Wescott]

yes step too fast too quickly and it becomes useless,
however it seems as though his is actually 'stepping' backwards,

That's exactly what ours did -- because 'useless' in this case meant
'fall', and it was trying to lift.

theres a lot to be said for positional feedback servos.

Hear hear!

Colin =^.^=

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google? See http://cfaj.freeshell.org/google/

"Applied Control Theory for Embedded Systems" came out in April.
See details at http://www.wescottdesign.com/actfes/actfes.html

 

[Nico Coesel]

In one application, I ran into another interesting trap for the unwary
- the motor was being asked to run fast enough that the back-emf in the
windings was almost as high as the drive voltage, which produced some
very interesting coil current waveforms.

Powering the motors from a current limited power supply solves that
proplem. About twelve years ago I build something that required a 1.8
Ohm / 2.5 Amp stepper motor to run at 5000 steps/s. At top speed, the
supply voltage was around 80V.

 

[Joerg]

Hello Jim,

[snip]

I've had the worm-gear type steppers run backwards like that if they
were lifting the load -- the upset of the step sequence on the coils
(and possibly some harmonic of said sequence) would make the thing go
essentially frictionless and fall down.

It was a consequence of trying to push the motor too fast, with too
little time in the project to get things right.

I've not done any stepper controller designs, but about ten years ago
I did a design for a 3-phase EC motor for Bosch (Mercedes A/C blower).

The squirrel cage would spin backwards when the outside vent was open
and the motor was not driven. Thus it would run backwards when
powered up under those conditions.

I came up with a cute way to cure that. I'll look through my notes.

Haven't done much with steppers either but one thing I did to avoid back
travel was to turn on one winding at a substantial DC current. That
stopped and held it.

(I even have a video tape of me driving the cage backwards using a
leaf blower, then applying power, and it slows down to a stop and
starts running in the correct direction.)

At Bosch? I didn't know they had those noisy leaf blowers in Germany.
Well, maybe now they do.

 

[Jim Thompson]

Hello Jim,

[snip]

I've had the worm-gear type steppers run backwards like that if they
were lifting the load -- the upset of the step sequence on the coils
(and possibly some harmonic of said sequence) would make the thing go
essentially frictionless and fall down.

It was a consequence of trying to push the motor too fast, with too
little time in the project to get things right.

I've not done any stepper controller designs, but about ten years ago
I did a design for a 3-phase EC motor for Bosch (Mercedes A/C blower).

The squirrel cage would spin backwards when the outside vent was open
and the motor was not driven. Thus it would run backwards when
powered up under those conditions.

I came up with a cute way to cure that. I'll look through my notes.

Haven't done much with steppers either but one thing I did to avoid back
travel was to turn on one winding at a substantial DC current. That
stopped and held it.

(I even have a video tape of me driving the cage backwards using a
leaf blower, then applying power, and it slows down to a stop and
starts running in the correct direction.)

At Bosch? I didn't know they had those noisy leaf blowers in Germany.
Well, maybe now they do.

Nah. This was in my office where I was _breadboarding_ the chip
design. (Electric blower.)

...Jim Thompson

 

[[email protected]]

Powering the motors from a current limited power supply solves that
proplem. About twelve years ago I build something that required a 1.8
Ohm / 2.5 Amp stepper motor to run at 5000 steps/s. At top speed, the
supply voltage was around 80V.

This ought to be obvious. The point I was making was that you have to
make sure that the supply voltage can go high enough to exceed the
back-emf from the motor - and even stepper motors have back-emf - as
well as coping with the resistance and inductance of the coils.

It is a point that I've seen ostensibly professional engineers miss.