Why such high voltage?

[sp1nm0nkey]

The power supply design PDF said 20 times the rated voltage on the motor. I'm designing a stepper controller myself, and I'm wondering if the geckos have some sort of power supply in them that drops that down or if that 20x the rated voltage is really what the motors see.

[ikke1983]

The motors do see that voltage, but not constantly. The current to the motors is constantly switched on and off at a frequency up to 20 khz(pwmulse width modulation).
Because the coil of a stepper motor reacts slow on current changes, the average curent will be the motor rated current. The high voltage is needed for good performance at higher speeds.
When you are running at high stepfrequencies with a low voltage the current in the coil of the stepper wont have the chance to reach its nominal value before the current is switched off to energize the next coil (high step freq.).
To get a full understanding of this I sugest that you have a look at the following text:
http://www.cs.uiowa.edu/~jones/step/

[Eclipze]

The detail behind this is complicated, however without going into the detail... the windings are inductive. If you apply a constant voltage across them, the current ramps up exponentially from zero to full current over a period of time. This time period reduces the maximum power you can get into the coils, because time is wasted getting up to the rated current. However if you can apply a higher voltage, this ramp up occurs faster.

With higher voltage, the current can go too high, so current limiting is required. This is where the power is PWM'd to coil, which is achieved by sensing the current going through the coil.

Higher voltage means higher speed. Higher voltage does not necessarily mean more torque though. As slow speeds, it's easy to get the power into the coil and achieve maximum current throughput. However as speed increases, a lower voltage supply cannot get the current into the coil and it stops PWMing the output and runs in voltage controlled mode. So higher voltage doesn't change low speed torque, only the maximum torque at faster turn speeds. If you go too high in voltage, it becomes difficult to manage the PWMing to the coil, because the duty is so low.

Gecko controllers are very sophisticated, as they also incorporate mid-band resonance damping, which prevents the stepper from stalling at high speeds... morphing to full step when required.

I understand all the theory (at least to a certain point) and quite experienced with electronics. I was so impressed with the Gecko G203V's, I dropped my owm design in progress and placed an order.

[Robin Hewitt]

The big volts are to overcome the inductance of the stepper coils. The inductance resists when you try to reverse the direction the current flows in to get a step. More volts overcomes the inductance faster, if you didn't have the big volts you wouldn't get any kind of speed out of it.

The motor rated voltage is simply the volts you would put across the coil to get the required amps through the coil. Amps=volts/resistance.

To use big volts you need stepper drivers that limit the amps to something the motor can handle without overheating and eventually burning out.

[SyS-e]

All motors are designed with a power factor rating usually expressed as 0.8pf . You can usually find it on the metal ident plate. The general idea thats simple to understand, is if you run a motor at 80% of its power rating, it will live longer, and not overheat and go into meltdown. Also an electric motor reaches it full potential conversion factor at 80% or its rating, and is able to take sudden surges without overlaoding. The stepper coils allow full voltage to be applied to steppers and servos while limiting current. S0 you want a motor 20% bigger not 20 times, than your controller coils limit. eg. 5 amp coils, 6amp motor.

[sp1nm0nkey]

So if I add a shunt resistor to do current sensing and regulate the current within the range of the motor with pwm I should be okay?

[pminmo]

Originally Posted by sp1nm0nkey So if I add a shunt resistor to do current sensing and regulate the current within the range of the motor with pwm I should be okay?

Maybe, it's more complicated than if your doing a bipolar driver and want to do microstepping. Decay modes become important and need to be addressed.

As was addressed by others the reason for higher voltage is to increase the speed attainable by the stepper. If you look at a stepper motors torque verses step rate, you will see it's torque drops significantly as step speed increases. As you increase the motor supply voltage, the rate of drop slows.

[sp1nm0nkey]

Sorry, I'm somewhat of an electronics newbie... I know this may be a bit ambitious. Anyway, what are decay modes and how do I account for them? I wasn't planning on doing micro stepping, but if I did I figured maybe I'd just check the current when one coil is full on (100% duty cycle), and then chop the supply voltage at high frequency based on the sensed current to get it within the motor's range.

I finished a simple stepper driver using tip122s and totally noticed how as the speed went up the current draw went down and the torque went straight down. I guess it just gets complicated sensing current when speeds change fast (can't integrate...) and pulses are short.

[Robin Hewitt]

Originally Posted by sp1nm0nkey I wasn't planning on doing micro stepping, but if I did I figured maybe I'd just check the current when one coil is full on (100% duty cycle), and then chop the supply voltage at high frequency based on the sensed current to get it within the motor's range.

Sounds good, you'd probably get nigh on 2/3 the top speed of a Gecko and you'd have complete control of the coils from your software.

I went the commercial driver route when I had a mains voltage surge and all 3 axes failed the smoke test simultaneously

If I was determined to do it myself again I might rummage that ghastly auction site for some of these...

http://www.national.com/mpf/LM/LMD18245.html

[sp1nm0nkey]

Oh hey... http://www.stepperworld.com/Tutorial...rostepping.htm

That was a huge help.

I think I'll just have a comparator on a shunt between the steppers and ground and a pot connected to a reference. I'll put a buffer between the fets and the uC and switch the buffer with the comparator. Quick and easy fast decay. It may be loud and inefficient but it should work

[Robin Hewitt]

My first attempt turned the power on when the clock went low then turned it off when the it reached the desired power. My second attempt used D type flops, I clocked them at 20kHz and put the output from the comparators on the data in lines. That way everything had a chance to settle down before I read the sense resistors and I no longer suffered from transients.

An old style moving coil meter is a real bonus when it comes to adjusting the current, these new fangled digital meters can get really confused when the line is clocking

[rdpzycho]

decay modes are not as simple as some literature put it. an oscilloscope is a good help when playing with them. on my previous job, I was designing DC to AC inverters and it's hard to debug hardware problems without a scope.

I am using PWM controllers (as they already have much of the needed hardware) and HIP4081 when designing my own controllers.

the supply voltage can actually go anywhere from 5x to 20x or even more. they are rough estimates on what can give you good performance on your motor. you can even use a non-regulated supply, but performance will vary depending on current consumption of the load. if you are already using the controller, try to measure (full rated current on stopped motor) the winding voltage and you will see the rated voltage of the motor.

using LC filters before the motor windings sometimes helps in reducing motor heating. they reduce HF AC current into the winding thus reducing eddy currents.