stepper motors

would anyone know of a way to turn a signal meant to drive a stepper
motor into a pulse witdh modulation duty cycle (mapped to the start/end
parameter of the original stepper controls

 

One way is to use a small microcontroller as the glue. Stepper motor
commands in one port, PWM out another.

 

Sound great, I understand a little (which can be a dangerous thing!)

I have had a look in maplin, but am unsure which one to go for because I
am as yet unsure of the nature of the stepper signal.

any ideas on what controller would do the job?

cheers
bern

 

Subject: stepper motors

That's a pretty interesting problem, mostly because there are only a few
circumstances in which the solution you want would be useful. Steppers give
angular control of motor shaft position by counting steps. Most applications
using stepper motors are actually trying to provide positional control of some
kind. In order to do that with another kind of motor drive, you would need
some kind of feedback from the motor shaft. That feedback is usually provided
with a resolver or an encoder, and the whole closed-loop system is called a
servo. It's necessary because motor speed curves are non-linear, and are
usually bent way out of shape by load torque requirements. That kind of thing
is not for the faint of heart, and requires more expertise than you'll probably
get from free advice on a newsgroup. Your statement at the end of your post
about mapping to the start/end parameter of the original stepper controls
indicates that you'll probably need a servo-type solution.

Packaged servo systems exist which accept the step/direction-type signals
common to stepper motor drivers. They're not cheap, though. You'll have to
purchase a motor with feedback encoder built-in along with your servo driver.
The application$ people who $ell the $ervo $ystem to you will be happy to help
with $etup and any nece$$ary fiddling (their $uper application$ help i$
$sometime$ built into the price).

If your control system is just controlling motor speed rather than position,
with neither angular shaft position or number of turns being important, you
might be able to do this. You'll also need some latitude as far as response
time (frequency-to-voltage converters have inherent lag). The easiest solution
would be reading the pulse frequency and outputting PWM with a cheap
microcontroller. There are also a number of analog methods for accomplishing
this, but from your description, they don't sound like what you need. If you'd
like more advice, a good description of your control system would be in order.

Good luck
Chris

 

Is this "signal" a "step signal" or is it the three/four phase signal
driving the stepper motor itself ?

The PWM signal would need to know what a single step is.
Some sort of feedback would be required to know when the step should stop.

More information would be nice.
Your question may be clear to you, but out here it make little sense.

 

It depends (as Chris points out) on what exactly the job is. A minimally
complex system could probably be run with a little 8-pin AVR ATtiny15L.
If you need more sensors or feedback or "smarts" then add program memory
and I/O width as required. Maybe an ATmega8 as a swag for a mid-range.

 

thanks for the response folks...

What I am trying to do is this.

I have a lpg closed loop control system (there are a few types on the
market) that uses a stepper to meter gas, according to lambda, tps,
temperature. This hardware has a fairly nice simple interface for engine
tuning. If you put you foot down, the stepper will jump and open up,
closing again due to lambda being rich, hence underfuelling moentarily
to create the fluctuation needed to acheive lambda..
The stepper has 255 steps, and most cars run a default level of 80-120
in the range.
I fit a lot of multipoint injection systems for gas, and for the older
cars, doing this mod to the old controller would be more than adequate
for the smooth running of the engine.
I know I am a little off topic(considering the gas blurb), and I
understand the motor has latency issues, although by fitting a device
that has PWM, would that latency be less?

How would I test the stepper to verify what type it is. Does it goe by
the number of wires going to it?.
so the crux of it is this.

I have a stepper with 255 steps, and I require an injection pulse up to
10 ms , extrapolated from the stepper signal.

I have faith in you all

cheers
bernie

 

[snip...snip...]

Not necessarily. You'd need to run the numbers.

The canonical web reference is http://www.cs.uiowa.edu/~jones/step/. You
can usually figure out the type by ohming out the lead combinations.

Your best bet would be to redo the control loop from scratch based on
the new drive characteristics.

Next best would be to get upstream of the point where the steps are
initiated in the current controller and re-interpret the control signal
in terms of a PWM value.

It could also be done by interpreting the step commands after they've
been sent by (waves hands vigorously in the air) tapping at the base
leads in the drive circuit, then having the firmware keep a count of
where the stepper is supposed to be and commanding the appropriate PWM
value.

 

thanks a lot for that, i will have a look.
back soon
bern

 

Greetings Bern,
After reading more of your posts one solution comes to mind.
Geckodrive (www.geckodrive.com) makes a device that accepts step and
direction pulses and outputs the servo drive. It uses an encoder to
position the servo. For $114.00 it's a good deal. An encoder can be
had for about $30.00 from U S Digital. For less than $200.00 you can
get a surplus servo, xmfr, cap and rectifier, the Geckodrive, encoder,
and 5 volt regulator and have a system that works well and right away.
It may be more than you want to spend but is an option.
Cheers,
Eric