powering a stepper with an ATX PSU

[ryckmans_t]

My question is about powering a stepper motor, and what type of inductive quickback a stepper could generate and if this could interfere with the power supply.

I have purchased big stepper motors – so I can build something else later with them. They are 3.78V 4.2A 85 N*m with specs here: http://www.motioncontrolproducts.co....products_id=11

I am trying to power these using an ATX 300W power supply, using either the 3.3V or the 5V supply – both can deliver 20Amps. The ATX is working correctly, but every time the motor takes a step, the ATX is shutting down.

I tried the following things:

-Connecting one phase of the stepper to the 3.3V or the 5V output does not stop the ATX, so it’s not shutting down as a response to a high current surge by shutting down
-Shorting (briefly!) the 3.3V or the 5V to ground: this does not switches off the ATX – the fan just spins faster.

This means the ATX is not powered off by a short or a surge. I tried to add a capacitor (1000uF) across the 3.3V supply to see if I could suppress any spike – no change. Someone told me it could be a "ground bounce". This is where your ground (or return) wire from the GND on the controller to the GND on your power supply is not low enough resistance.* This can cause a higher voltage at the ATX supply end causing it to shutdown.

I tried using much thicker wire to connect the ATX to the controller, but this did not work either… tried using more ground wires from the ATX to the stepper bee+ ground (7 for them!) but no changes – the stepper takes one step and the ATX shuts down…

-what is the inductive quickback from a stepper motor? Could this be the issue? On the other hand ATXs should be powering the steppers (albeit smaller ones) running the hard disk… so there must be a way.

I could use the following things (or a combination of these)
-some diodes to block a negative spike due to inductive quickback. Will need the diodes to pass 4.2Amps and dissipate 4.2x1.5Voltage drop = 6.3 Watts. Will have to use the 5V supply because of the voltage drop across the diode…

-a giant smoothing cap across ground and 3.3V supply….
-a 3.3 Zener to bypass any positive spike due to inductive quickback
-A LC filter
-A MOV
… any ideas welcome!!

I’d be really grateful for your insight on this – how do you power your steppers suing a cheap PSU?

many thanks - Thomas

[happytriger2000]

Hello Thomas,
I'm having the same problem, I built a atx psu modified to a lab psu and I thought I could use this PSU to power my stepper motor at 17V, where -12v +5V = 17v but somehow the power shut down. I think the current is not hight enough to supply the motors. However if you check this link out: http://www.thebackshed.com/cnc/RouterA2.asp, he uses 2 x AT psu to achieve 12V at 8amps, so I might be right about the Amperage value.

Here is a photo of my little project: please open Attachment.

If you compare the power supply unit sold online or any other store cnc related projects, the size of the power unit is big and heave able to supply 48v, 10Amps......where as ATX psu....mm......slightly small in size, but not sure about the power, I guess dependes on PSU type.

Cheerio

Freddy

[happytriger2000]

Hello again,
Got problem uploading pictures in the previous reply, so I'm gonna attach the picture in this reply.



Freddy

[kreutz]

Hello Thomas,
I'm having the same problem, I built a atx psu modified to a lab psu and I thought I could use this PSU to power my stepper motor at 17V, where -12v +5V = 17v but somehow the power shut down. I think the current is not hight enough to supply the motors. However if you check this link out: http://www.thebackshed.com/cnc/RouterA2.asp, he uses 2 x AT psu to achieve 12V at 8amps, so I might be right about the Amperage value.

Here is a photo of my little project: please open Attachment.

If you compare the power supply unit sold online or any other store cnc related projects, the size of the power unit is big and heave able to supply 48v, 10Amps......where as ATX psu....mm......slightly small in size, but not sure about the power, I guess dependes on PSU type.

Cheerio

Freddy

12 Volts and 5 Volts share the same ground, if you try to wire them in series you are short-circuiting the 12 Volts rail, hence the shutdown.

Kreutz.

[kreutz]

You will be very disappointed even if you get the steppers to work with any 12volts power supply, unless you are aiming for very low rpms.

Did you try using a 5 ohms 10 Watt resistor (with heat-sink or attached to the chassis) across Ground and the 5 Volts output of the ATX power supply?, switching mode power supplies have a minimum load requirement.

Kreutz.

[ryckmans_t]

kreutz - yes I tried with a load (various power resistors across 5V-ground or 12V ground, drawing 0.3 to 1 amp) but it is still switching off everytime the stepper takes a step.
>> You will be very disappointed even if you get the steppers to work with any 12volts power supply, unless you are aiming for very low rpms.

do you imply I should go for higher voltage, with a current limiting resistors? At this stage I don't mind low rpm - I just would like to understand why the ATX does not work!

many thanks

[george4657]

If the ATX power supply is shutting down it means that there is a short in the circuit. Check your connections to the board and make sure they are correct.
The 3.3 volt supply is usually not rated at a high voltage so use the 5.0 volt.
I would put a amp meter in series with the power supply and measure the current being drawn.
First though I would hook up only one motor and I would put a power resistor in series with the power supply. With no load the motor should still turn even with the current limited too far less than spec. It should turn but have no power.
When you get it to turn lower the value of the limiting resistor or try with no resistor.
Your driver board looks like it has 4 FETS to power the motor. At 4.2 amps you will need heat sinks on those chips.

Hope this helps.

George

[kreutz]

kreutz - yes I tried with a load (various power resistors across 5V-ground or 12V ground, drawing 0.3 to 1 amp) but it is still switching off everytime the stepper takes a step.
>> You will be very disappointed even if you get the steppers to work with any 12volts power supply, unless you are aiming for very low rpms.

do you imply I should go for higher voltage, with a current limiting resistors? At this stage I don't mind low rpm - I just would like to understand why the ATX does not work!

many thanks

Show us schematics of your drive, there is something weird going on. I suppose part of your circuit is hooked to +5V (logic) while motor windings are connected to the +12V (for unipolar motors) by means of (two or four) power resistors calculated to limit the current to the maximum specified motor current, the other end of the windings to the respective N channel Mosfet drains, all voltages sharing the same ground. Why are you adding Capacitors to the 3.3V power supply then? You should add capacitors to the 12 Volts supply instead.

Best regards,

Kreutz.

[ryckmans_t]

george - the ATX still runs even with a short - see original post:
>-Shorting (briefly!) the 3.3V or the 5V to ground: this does not switches off the ATX – the fan just spins faster.
kreutz - the stepper are 3.78V and I am using 3.3 V from the ATX (tried the 5V as well but this made no difference). Nothing is connected to the 12V, but I have tried a 15ohm power resistor (no difference again). I am using the controller here:
http://www.pc-control.co.uk/stepperbee_plus_info.htm which is using a USB
I have not used capacitors - I am just asking if this would be a solution. Tried a flyback diode today - did not work. The ATX still shuts off when the stepper takes its first step

many thanks.

[kreutz]

george - the ATX still runs even with a short - see original post:
>-Shorting (briefly!) the 3.3V or the 5V to ground: this does not switches off the ATX – the fan just spins faster.
kreutz - the stepper are 3.78V and I am using 3.3 V from the ATX (tried the 5V as well but this made no difference). Nothing is connected to the 12V, but I have tried a 15ohm power resistor (no difference again). I am using the controller here:
http://www.pc-control.co.uk/stepperbee_plus_info.htm which is using a USB
I have not used capacitors - I am just asking if this would be a solution. Tried a flyback diode today - did not work. The ATX still shuts off when the stepper takes its first step

many thanks.

I was reading your "stepper bee" manual and think you might have damaged it. It says it works from 12 to 24 volts motors but only 500 mA absolute maximum per winding. It also needs what they call transient suppression (not specified). There is not much real information, what is clear is that is was made for really small motors (specified 12V 200 mA per phase winding) nowhere near the power and low coil resistance of yours. It seems that the little bee really stung you

Kreutz.

[ryckmans_t]

Hi Kreutz - I am using the stepper bee+, which can operate up to 7A per winding at up to 24v. Also, If I use a 5V 5A source, it works fine, so it looks like the controller is fine...

[happytriger2000]

You will be very disappointed even if you get the steppers to work with any 12volts power supply, unless you are aiming for very low rpms.

Did you try using a 5 ohms 10 Watt resistor (with heat-sink or attached to the chassis) across Ground and the 5 Volts output of the ATX power supply?, switching mode power supplies have a minimum load requirement.

Kreutz.

Hello Kreutz,
Yes I have, but not a 5 ohms, I used a 10ohms, 10watt ceramic resistor.

freddy