Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

stepper too slow !!!

Status
Not open for further replies.

dreamproject

New Member
i am using an uln 2003 to control a 7 kg-cm torque stepper . my problem is it's too slow. i intend to use it to drive the steering of a real car. i am using Qbasic to generate the pulses. Is it a characterisic feature (slow speed) of steppers (can anything be done , i spent a ton on these motors) or do i have to look to other kinds of hardware for this feature ( 20 degrees / sec with good torque) . thanks in advance.
 
dreamproject said:
i am using an uln 2003 to control a 7 kg-cm torque stepper . my problem is it's too slow. i intend to use it to drive the steering of a real car. i am using Qbasic to generate the pulses. Is it a characterisic feature (slow speed) of steppers (can anything be done , i spent a ton on these motors) or do i have to look to other kinds of hardware for this feature ( 20 degrees / sec with good torque) . thanks in advance.

Steppers are slow, they don't actually 'rotate', only move a single step at a time - it seems a really bizarre choice for what you are doing?. I would consider the 'proper' way to do it is with a DC motor servo system, this way you also have precise control of the steerings absolute positioning, plus much more power (and much less money!).

You can increase the stepping speed of steppers, by increasing the supply voltage and using current limiting resistors - but it doesn't look a very good choice to me!.
 
I'm with Nigel - a DC servo motor system is the only way to go.

You can increace the speed of steppers but it increaces the likelyhood of losing counts. You could mount an optical encode on your stepper to ensure you don't loose counts but a DC motor would still be a better solution.
 
the fastest way to drive a stepper is with a constant-current driver. these are commonly called "chopper" drivers. essentially you drive the motor at way above its normal voltage rating (like running a 6v motor at 30v) and switch it with a transistor, but have a feedback resistor in the current path of the motor coil... this feedback resistor drives a comparator, the output of which controls the base/gate of the transistor, and the other input of which is connected to a reference voltage. in this way, the motor receives a high voltage to help quickly overcome the inductance of the coil and get current flowing, and then when the current reaches the right value, the transistor is cycled on and off to maintain that current. this turns the coil on as fast as possible with a given supply voltage, and you use diodes to collapse the field in the coil as fast as possible after the coil is turned off, to help it de-energize as fast as possible. this way you can push your maximum step speed way up without sacrificing torque, so you are less likely to skip steps, especially at high torque levels on the output.

doing it this way is the next step past using a higher voltage with a series resistor, and gives about the best possible performance. the only way to do it better is to use an essentially infinite supply voltage so that the coil can energize and de-energize instantly... but this is about as you can get (reasonably) in the real world.

here's a page with some info on that for you:
**broken link removed**
 
dream project , what are you doing for feedback?
and what if the steering wheel is a little out of balance and decides to turn by itself..?
i am not trying to poo poo your project, just trying to give ya something to think about
 
Dreamproject,
Unless you have a high power steering assist a 7Kg-cm servo will hardly turn a car wheel. In foot and finger units you have a 6 in-lb motor. Have you ever felt how much force it takes at the steering wheel rim to turn a car? On a 12" (300mm) steering wheel you can only apply 0.5 lbs (0.467 Kg under 1g acceleration).
Another problem will be hunting. stepper motors work in discrete steps so your vehicle direction will be proportionally the same. Any road tilt or vehicle assymetry will make the car wander with your motor only being able to discretely correct for it.
Hopefully you are not depending upon the paint lines for guidance.
Interpreted BASIC probably runs way too slow. You might as well be using a Commodre 64 to drive it. Why not compiled C or assembler?
 
Quite worked out

hi,
thanks to every one for their input . I am planning to use a joystick to control the steering. Since I have geared the motor (7 kg-cm) with a 2:1 reduction ratio, I should get about double the torque . The motor (with gears) connects directly to the steering column . With a centre to centre distance of about 3 cm (steering column centre to motor shaft centre) , i should be able to get atleast 2 - 3 kg force , which is more that what a 640 kg hatch-back with power steering would require. :wink: (hopefully). With built-in self locking (reduction gears) , I do not think the steering would turn by itself based on road variations . I have damped the input (pot in the joystick) sufficiently to compensate for hunting.
 
sorry if am out of line here but driving a car with a joystick does this mean back seat driving will become a reality :shock:
 
the future concept cars from a great number of manufacturers show that the cars of the future will be operated with something simillar to a joystick. and some inventors are thinking of cars needing no driver at all. the roads will be like big conveyer belts. the computers will do all the steering and acceleration.
 
i used servo, it's extremly precise and fast (two seconds from full left to full right):
 

Attachments

  • steering_jig.gif
    steering_jig.gif
    169 KB · Views: 620
Re: Quite worked out

dreamproject said:
hi,
thanks to every one for their input . I am planning to use a joystick to control the steering. Since I have geared the motor (7 kg-cm) with a 2:1 reduction ratio, I should get about double the torque . The motor (with gears) connects directly to the steering column . With a centre to centre distance of about 3 cm (steering column centre to motor shaft centre) , i should be able to get atleast 2 - 3 kg force , which is more that what a 640 kg hatch-back with power steering would require. :wink: (hopefully). With built-in self locking (reduction gears) , I do not think the steering would turn by itself based on road variations . I have damped the input (pot in the joystick) sufficiently to compensate for hunting.

I thought the original question was how to make it faster?, yet you've now slowed it down by a factor of two?.

Presumably this is to give enough torque to move the wheel?, at the expense of speed of movement.

However, I'm somewhat dubious about the amount of torque this will give, it might give enough to move the wheel under minimum load conditions, but I suspect it will fail under some actual load conditions - you need a far higher safety margin than 'just enough'.

You have never stated that you have any kind of feedback arrangement?, so there's nothing to keep the wheel where it is, and it's VERY likely to turn itself under varying road conditions. It's essential to have some feedback mechanism, so you know at all times EXACTLY where the wheel is.

I've only just noticed, you are using a ULN2003 to feed the stepper!, this is a fairly puny IC, perhaps suitable for moving a printhead in an inkjet printer, but not for steering a car!.

It looks to me like the entire scheme is misconceived, a small underpowered stepper isn't a suitable device for this role! - you really do need a servo motor system (and I don't mean the little ones in R/C models). It will give you masses more power, faster response times, and accurate feedback about it's position.
 
Hi all,
Thanks Nigel , Panic , Sam , bryan et al for your valuble inputs . Point taken period . Nigel , r u suggesting that I hook up a powerful DC motor with a feed back system or a servo motor with built in encoder (futaba ,etc are not available here ). I have a biq question on DC motors . If i power them , are they not going to rev off even before the encoders (or feedback mechanism) realises . How do we brake them ? . I selected steppers for another reason (holding torque) . Is this possible with DC motors .Please shed some light . :cry: Thanks
 
System in a picture above was designed for industrial application
that needed to run at very high speed, torque and precision.
Torque was limited to max. 22Nm to prevent damage to the car.
Normal operation torque was ca 15-18Nm. Torque was this high
because operation happens before power steering is functional
(and before steering wheel is installed, that's why the unit has
handles so operator can manipulate the unit and mount it on
steering column of tested vehicle).

I was using Mitsubishi components because that was customer requirement:
Servo Drive MR-J2S-20B1
Servo Motor HC-MFS23K (200W, built in high resolution encoder)
Motion Controller QD75M1 (installed in Q PLC) etc.
Motor was coupled with 30:1 gear ratio (10:1 gearbox plus
additional 3:1 gears since the lightweight gearbox wasn't available
with higher ratio).

One of the limitation was weight (it had to be less than 6.5kg for
whole unit) and you can see that big part of the system is made
out of aluminum and composite materials.

For your application motion controller might be an overkill.
DC Motors are great for torque applications. To control them
precisely you will need proper Drive. Drives with feedback
are available...
I am not sure what is your project requirement except to control
it by joystick. What is the driveing speed? How sharp turns are expected?
Is this supposed to be off-road vehicle?
 
Also, how many steps per revolition your motor has and what is the max step frequency (pulse rate) you use to run it? I was using steppers few times (quite limited performace imho, since I'm used to servos).
Last one had some 200 steps and it would run up to 2kHz with proper ramp up (and if torque requirement was low). Thats ca 10rev/second or 600RPM. (this was in machine that cuts and assembples little brushes for Xerox machines).
 
dreamproject said:
Hi all,
Thanks Nigel , Panic , Sam , bryan et al for your valuble inputs . Point taken period . Nigel , r u suggesting that I hook up a powerful DC motor with a feed back system or a servo motor with built in encoder (futaba ,etc are not available here ). I have a biq question on DC motors . If i power them , are they not going to rev off even before the encoders (or feedback mechanism) realises . How do we brake them ? . I selected steppers for another reason (holding torque) . Is this possible with DC motors .Please shed some light . :cry: Thanks

A servo motor will give far greater holding torque than a stepper, the servo mechanism prevents them moving from where they are told to go.

Speed isn't a problem, as the motor is geared down greatly in a servo motor, and the system can be designed to slow down as it gets close to it's settling point.

The small R/C ones (Futaba etc.) are a good cheap example, but too small for what you need - but the principles are exactly the same!. I can't imagine model servo's like Futaba aren't available where you are, they are such a common component - I'm sure you can get them in any country?.

Panic Mode has listed what he used, basically industrial sized servo components - it looks like exactly what you are trying to do?.
 
Thats the way to go

Nigel et al ,

I think SERVO is the ONLY way to go . Point Taken . I'd better get started on that . Any useful links on PID control of DC motors , encoder construction (ready made ones are not common) , and other updates to that effect would be greatly appreciated . And by the way , Nigel , (no Radio Shacks down here in India (futaba) , though i wish they were..... )

And panic , it is an engineering marvel , your design , :wink: , but I cannot use any components / ideas that you suggest because of the prohibitively high prices.

And the answer to the question by Panic , it is planned to be installed in a normal road - use car (suzuki hatch-back) . It would be expected to negotiate all obstacles that normal usage in city traffic would put forth .
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top