Stepper Motors

[Neil WyattModerator @neilwyatt]

I want to make a simple stepper controller, nothing fancy, half-step mode is plenty for what I need and the step rate will be about 1 or 2 pulses per second.

I have unipolar a Minebea stepper of 0.4 nm at 2000 pulses per second, plenty for the well geared down application (telescope mount).

As it's unipolar, it will be easy to drive. What I am struggling with is knowing the 'right' driving voltage.

The motor is rated as 24V but the coils are only 9R8 DC resistance (the datasheet says they should be 5R, but this may be a slightly different model). Thing is, they are rated as 80mA max – and 24V would give me 2.5 Amps.

So two questions:

Is the 24V for when you just apply a voltage to move the motor and don't apply a holding current?

If I want to use a holding current (I do, as I want to half step) should I instead use about 8V to give me 800mA, or ean even lower voltage, say 5v?

I should point out that the load on the motor will be quite small.

Neil

[Neil WyattModerator @neilwyatt]

[Les Jones 1Participant @lesjones1]

Hi Neil,
You would only need to use the high voltage (24V) if you need to step it fast. This is because to get the current to rise quickly you need the higher voltage due to the inductance of the coils. For your application driving them directly with about 0.8 volts but it would probably be easier to use 5 volts with a resistor of about 52 ohms (5V/0.08A = 62.5 ohm -9.8 = 52.7 ohm. ) in series with each coil. Most stepper drivers control the current using PWM to save wasting power in the dropper resistor.

Les.

[Neil WyattModerator @neilwyatt]

Thanks Les, will a voltage as low as 0.8V really be enough?

PWM is no problem – I have googled that the rate shoudl be >100KHz, but found nothing on duty cycle. It sounds like about 5% might be a good starting point?

I suppose I may want to upgrade the unit to 'go to' function later which would mean much faster stepping. I also have the 24V+5V PSU that came out of the same printer as the stepper, although a12V solution would be best for me, so it can be portable.

[Les Jones 1Participant @lesjones1]

Hi Neil,
The PWM does not have a fixed duty cycle. The stepper driver chips (Such as TBA6560, TBA6600 etc.) have a current sense resistor which is used to form a closed loop constant current driver. When the drive voltage is first applied to one of the windings the initial current will be low so there will be only a small voltage developed across the sense resistor. At this point the feedback system will give a near 100% duty cycle. At some point in time the current will reach its set value. When this happens the duty cycle will be reduced so the current does not exceed its set value. If you look at these controllers you will find that there is a group of switches that set the desired drive current. You will probably find it cheaper to buy one of these controllers on eBay than to buy the chip and build your own. For the present telescope application I think just using resistors is good enough.

Les.

[jason udallParticipant @jasonudall57142]

Google audrino stepper sheilds…simplw drivers expecting ttl pulse train for step and direction. ..as have been said…cheaper than the chips in one offs

[jason udallParticipant @jasonudall57142]

And possibly spell shield right….

[MuzzerParticipant @muzzer]

The Arduino etc "shields" simply contain an off the shelf stepper driver IC but then require a micro to generate the drive pulses. I did this myself not so long ago with an MSP430 starter kit but it's cracking a nut with a sledgehammer.

You could get your own stepper driver board and simply drive it with a pulse train eg from the old 555. Something like this. The clever bits are already taken care of and they cost peanuts on ebay.

The high voltage is required to allow a sensibly fast speed from the motor. If you put resistors in series, you'll end up with no torque – it's not the right solution. There's really no point trying to drive the motor directly or making your own driver circuit.

Murray

[jason udallParticipant @jasonudall57142]

Murry..I suspect that the “accuracy” of the pulse train from a 555 astab will not meet the telescope drive requirements. ( sidereal time base? )..
And using a pxo and stepper drive would indeed be all that is required…..the micro derived pulse trail at least allows for “tunning” of a basic 10-30 ppm crystal time base….

[Les Jones 1Participant @lesjones1]

Hi Neil,
This information on stepper motors (From the link on Arc Euro's website.) is worth reading.

Les.

[Martin KyteParticipant @martinkyte99762]

Will you not get too much shake from your Stepper by driving it in half step. You really need to supply two phase ac to get a smooth run. Close approximation could be

Farnell

• Order Code: 2424511
• Manufacturer Part No STK682-010-E

running in microstep mode.

Depends if you plan to use a MK1 eyeball or astrophotography really.

regards Martin

[MuzzerParticipant @muzzer]

Jason – sounded as if Neil was planning on developing his own driver but it's not simple. Stepper drivers are switched mode constant current drivers integrated with gating logic and an up/down counter, so pretty sensible to buy it rather than develop it yourself.

If this is for tracking planets, the idea of using a piezo or crystal controlled micro makes sense and a single micro board can be had for peanuts in whatever flavour you like.

BTW, Minebea are the Rolls Royce of small BLDC and stepper motors. The life of steppers and fans is determined by the quality of the bearings and seals (and when the oil runs out). IIRC, Minebea (Nippon Miniature Bearings) actually make and supply the bearings to the majority of the other motor manufacturers.

Murray

[Neil WyattModerator @neilwyatt]

I'm aware that stepper motors have the reputation of being a black art, these responses are proof!

This is initially a low-speed tracking mount for astrophotography. I have a light 70mm scope and a bridge camera with a 1000mm equivalent lens. The camera only does 8-second exposures, but even that will benefit from some tracking. If things go well, I will add a webcam mount and maybe even invest in a second0hand DSLR body with a 30-second (or better) exposure time. This may seem excessive for a small scope, but as results to date with the bridge camera have been very rewarding further good experiences will give me an excuse to go to a 6 or even 8 inch reflector.

A half-stepping controller using an AVR will be an easy enough exercise, it's just putting a bit-pattern at the right speed; I will also fit a few switches, a serial link and a LCD readout. Why? well the sun and the moon move at noticeably different rates from the rest of the sky. These will need to be fine tuned so it will be useful to have a readout of the 'theoretical' speed to compare with what I'm seeing. And also to confirm I have the right setting, once the speeds are nailed. Also, I can readout where the scope is (supposed to be) pointing.

A serial link will let me implement a go to system, the eq mount for the scope has the same 90:1 worm drive for both axes so if it works on the one, I can add a second unit. Then all I have to do is send two numbers from a laptop and the scope will go to that point.

It's getting the drive current right that I'm exercised about.

The tutorial linked to by les is very good, especially as it points out that unipolar steppers act as auto-transformers so I need to chuck in an extra set of diodes to protect the switching transistors.

One issue is whether or not it's worth using PWM instead of resistors for current limiting. I'm comfy with PWM for ordinary motor control. The big advantage is that PWM can be experimented with sitting at the computer, and if successful could be extended to give more sophisticated drives.

Another is will I be able to use the same set up to go to a particular direction, or does it need to be more sophisticated. If I set the drive at approximately one half-step per second, then to complete a full rotation in 30 seconds (which is pretty fast really as most moves will be of the order of 45-degrees) I will need to drive at 1500Hz (full step rate), which is pretty low for this stepper, and probably will work without complex ramp up and ramp down of the step rate, although ramp up/down shouldn't be hard to implement.

After saying all this, I have just discovered I have an Allegro A3984 somewhere… this will drive a bipolar stepper at 2A/35V so well up to the stepper I have, with 1/16 step microstepping I think it's a no-brainer to wire it up and see what happens!

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 09/01/2015 12:12:56:

…After saying all this, I have just discovered I have an Allegro A3984 somewhere… this will drive a bipolar stepper at 2A/35V so well up to the stepper I have, with 1/16 step microstepping I think it's a no-brainer to wire it up and see what happens!

.

Neil,

Now I'm confused … I think you said originally that your stock motor was unipolar.

MichaelG.

[Neil WyattModerator @neilwyatt]

> Now I'm confused … I think you said originally that your stock motor was unipolar.

You can drive a unipolar stepper as a bipolar one by ignoring the centre tap.

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 09/01/2015 16:34:14:

> Now I'm confused … I think you said originally that your stock motor was unipolar.

You can drive a unipolar stepper as a bipolar one by ignoring the centre tap.

.

Just checking

MichaelG.

.

For my pennance, here is a link to one implementation of an Allegro device.

Edited By Michael Gilligan on 09/01/2015 17:17:10

[jason udallParticipant @jasonudall57142]

In that case why not use a

A4988 POLOU or similar module which takes STEP DIR RESET and micro step ratio inputs

About 10 quid for five. ..

Swings upto 35V and onboard current limit ( trim pot to set) of up to 2A

[martin perman 1Participant @martinperman1]

Neil,

A very interesting read, a friend of mine designed and built a telescope tracking device years ago, I am seeing him on Tuesday night so will ask himwhat he did.

Martin P

[LimpetParticipant @limpet]

Neil if you're thinking of doing astro imaging even the commercial goto mounts drift after a few minutes the method used to drive the mount is by focusing on a bright star (with a light splitter or another scope and camera – web cam) and driving the mount to keep the star in the same position relative to the scope It will then follow all night or until the clouds come

Lionel

[Michael GilliganParticipant @michaelgilligan61133]

Posted by jason udall on 09/01/2015 17:10:36:

In that case why not use a

A4988 POLOU or similar module which takes STEP DIR RESET and micro step ratio inputs

About 10 quid for five. ..

.

Jason,

Where do you find them at that price

MichaelG.

[jason udallParticipant @jasonudall57142]

http://www.amazon.co.uk/s/ref=nb_sb_noss_1/275-5056453-2897412?url=search-alias%3Daps&field-keywords=a4988&sprefix=a4988%2Caps

**LINK**

[Michael GilliganParticipant @michaelgilligan61133]

Thanks, Jason

… Interesting range of prices !!

MichaelG.

[Neil WyattModerator @neilwyatt]

> For my pennance, here is a link to one implementation of an Allegro device.

£18 to cover conversion from SMT to DIL (retrograde step if you ask me) and add four resistors and a capacitor!

The A4988 is an uprated version of the chip I already have. Picked up five, in case I ever do a reprap…

Neil

[jason udallParticipant @jasonudall57142]

Neil the "reprap" I am currently working on uses 9..

4 for axis 5 for extruders…

.

[Neil WyattModerator @neilwyatt]

Oh well, first thing to do will be to print four more

Neil

[Author]

Posts