Stepper Motor voltage and power

[Iain DownsParticipant @iaindowns78295]

Hi, all.

I have built a mill power drive with a stepper motor and Arduino.  It works.  Sort of.

The problem is that it struggles towards the ends of travel of the mill table, particularly to start off (in fact it won’t start and won’t move – I have to add some human torque).  The mill is a VM32L with table 840mm long.  So quite heavy and easy to see how it needs a bit more effort at the ends.

The motor is an ACT motor 23HS2442J3 (about 1Nm torque – dropping with speed, of course) and the driver is TB6600 set to 4.2A.

However, the power supply is a 24V supply, but I have read that a higher voltage is better to push the amps through the coils.

I was wondering if swapping out the PSU for a 40V one would make enough difference to be worth while.  My guess is that it would need 1.5 to 2x the torque to work well at the ends.

Also, I think I have the microstepping set to 8 – should I drop that down?  I seem to recall from early trials that the movement is noticeably less smooth.

My suspicion is that the extra voltage won’t get me where I want to be, but would appreciate some input from those who may know.

Many thanks

 

Iain

[John HaineParticipant @johnhaine32865]

Iain, more volts helps when making the motor run faster and smoother but the current is set by the driver.  From limited data on Amazon the coil resistance is 0.9 ohms so would take nearly 30A from a 24V supply if the driver let it!  The same data says that the holding torque is 2.8Nm?  If you have access to a power supply then trying more volts might help.

Have you tried increasing the microstep ratio.

I use a somewhat smaller motor on my VMB mill power feed driven from a DRV8825 driver on 24V.  It also complains at one end of table travel because the table friction increases, but the coil current is something like 1A or less.  Some adjustment of the gibs might help?

Do you have any mechanical reduction to the feed screw?  I have a 2:1 timing belt drive.

The TB6600 drivers don’t have a great reputation and apparently there are ones around that look the same but have different less capable driver chips inside.

[Michael CallaghanParticipant @michaelcallaghan68621]

It’s torque you need, not more volts. There are two ways of adding torque. The first swap the motor for a higher rated motor. The second is to use belt drive and add the power of pulley ratios. On my warco conversion I went for pulley ratios to increase the available torque of the motor. Just adding more voltage will not help much as the real strength is in the amount of amps your system is rated at, most are between 4 to 6 amps. Cheapest way is too swap the motor.

[Andy_GParticipant @andy_g]

As above, more volts helps maintain torque at speed (to maintain current against the back EMF generated by the running motor). I don’t believe it will do much for starting torque.

That’s already quite a powerful motor for the NEMA23 form factor, so it may be that the torque required is too great, and you will need to add reduction gearing (2 or 3:1 timing belt).

However…

You say you’re using a ‘TBA6600’ driver – is it one of the generic ones that are all over Amazon, Ebay & Aliexpress?

If it offers 32:1 microstepping, it hasn’t got a TBA6600 chip inside – likely a TB67S109AFTG instead which, despite what the printing on the case says, is limited to 3 amps output.

If so, you could get a 40% increase in torque by running a drive that allowed you to run the motor at its rated 4.2 amps.

A fairly cheap drive (that uses the TBA6600 chip) is sold as the ‘DIV268N’ – billed as a 5A drive.

I bought a couple from ‘Maker’s Hut‘ and they seem OK, but they’re old fashioned and crude compared to more expensive drives.

 

 

[Martin ConnellyParticipant @martinconnelly55370]

If you don’t want to go through the process of making pulley based speed reduction system then adding an inline planetary gearbox is an option. The RPM a leadscrew needs to run as fast as is necessary is quite low compared to the top speed of a stepper motor so any form of speed reduction between the motor and the leadscrew will work for greater torque without losing the ability to move the machine as fast as you need or want.

Martin C

[Iain DownsParticipant @iaindowns78295]

Thanks for the input guys.

I had expected the motor to be powerful enough and it’s not particularly hard to turn the wheel by hand even at the ends.

The driver I have does indeed have a 32 microstep option so is likely the cheap and not so cheerful option.

The one you pointed me to together with a 48V power supply ( which I think I have knocking about somewhere) looks like a good option.

I will give that a go.

 

Iain

[Iain Downs]

On Iain Downs Said:

together with a 48V power supply

I’d be wary of that – it would be close to the voltage limit of those drives with no leeway for the supply voltage being over driven due to inertial loads, back-driving the motor, etc.

Might be worth sticking to 24V unless you need higher speeds – what is the maximum RPM you expect the motor to be able to reach?

[HuubParticipant @huub]

You could:

• Set the microstepping to 2 or 4 for more torque.
• replace the driver by a digital driver like the DM556. These drivers change to full step when the motor is running at a reasonable speed. This results in more torque.
• replace the driver by a TMC2209 stepstick driver. I use these on my mill and I am going to use them on my 2 CNC lathes. In my experience, the TMC2209 driver delivers 4 times more torque than a TB6600 driver and they run very quit.

[Ady1Participant @ady1]

Fitting ballscrews would probably help but a faff to fit

less friction plus it almost eliminates backlash

I got an xyz set for 100 quid delivered from amazon but not fitted them yet, it’s a pretty big job and I bet there will be milling required to get the job done… once I’ve got my mill all in bits

[Iain DownsParticipant @iaindowns78295]

The power feed will run at up to about 12mm / second through 70% of the mill travel. the speed which it will start it drops off markedly near the ends of the beds.  the acceleration is currently set lower than I would like – it takes a few mm to get up to even a lowish speed.

I’ve got it connected to the leadscrew with  a 1:1 ratio (timing belt).  I don’t really want to use 2:1 as it will reduce the top speed of the traverse and that is half the reason for installing it!

The DM556 looks like an option, but the TMC2209 doesn’t look beefy enough – the TB6600 should already be providing over the peak current of the TMC2209.

However, it may be worth looking more widely for a driver – I believe Clough42 did some comparisons for his electronic leadscrew project.

Adv1 – whilst I can see ballscrews in my future, it’s not a project I’m prepared to tackle now (I have a dozen or so backlogged!).  In any event the X axis travel is very smooth and easy through most of the travel, it’s just the ends that are problematic – so ways not screws I think!

Ta all.

 

Iain

[JasonBModerator @jasonb]

You really need to decide what half you want. A rapid return or a usable rate of feed. If rapid is more important then a motor with more grunt to run 1:1. If feed rates are more important then gearing/belting down would be the best bet.

What size and material (hss or Carbide) cutters do you tend to use and what sort of cuts, work out the chip loads and hence the feed rates you are likely to want and then work back from that.

Do you notice an increase in the amount of effort needed to turn the handwheel by hand, if so look at the gibs and check the leadscrew end bearings line up with the nut as tightening at the ends of travel can be a sign the screw is being bent at a steeper angle a sit gets close to an out of position bearing.

The Powerfeeds I have on the Sieg mills max out around 450mm/ min (7.5mm/sec) and that is pushing it when cutting and I seldom wind it all the way upto that for the return. CNC is a lot faster but so is the spindle speed so feed rates are therefore higher.

This is 450mm feed rate

[Iain DownsParticipant @iaindowns78295]

Thanks, Jason.

I’ve not tried to cut at the top end of the speed, but have cut at around 4-5mm/sec with a 63mm face mill on steel (somewhere under 1mm DOC).

Yes it is a bit harder to turn the wheel by hand at the ends, but nothing like a struggle.  It’s very free near the centre.

I will have another look at the X Gib and see if that makes a difference.  I’m not sure how to check the bearings.

Iain

[Iain Downs]

On Iain Downs Said:

The power feed will run at up to about 12mm / second through 70% of the mill travel. […]

 

I’ve got it connected to the leadscrew with  a 1:1 ratio (timing belt).  I don’t really want to use 2:1 as it will reduce the top speed of the traverse and that is half the reason for installing it!

Has your mill got 0.1″ pitch leadscrews?

If so, 12mm/s corresponds to ~280 RPM. The stepper should easily be capable of at least 3x that speed with reasonable torque. If you gear it down you will be able to set faster accelerations and you will probably be able to achieve a higher traverse speed, too (especially if you up the supply voltage).

The DM556 is the pick of those drivers, but note that it is advised not to exceed 45V (despite 50V spec).

 

[SillyOldDufferModerator @sillyoldduffer]

Some confusion here, this is a stepper motor with an electronic drive,

• Only the absolute maximum speed of the motor is determined by the supply voltage.  Applying more volts allows the motor to accelerate faster without losing steps, but motor speed isn’t proportional to voltage.
• The working maximum speed of an adequately powered motor is determined by the input pulse rate and the mechanical limits of the motor.    This type of motor is designed to step moderately accurately at lowish RPM and to hold position when stopped.  They aren’t boy racers!
• Power (rate of work) and Torque (turning power) of steppers depend on Amps.  Always look at the amps, because volts are less important.
• The motor is pulsed with an electronic driver that can be set to step the motor at 200 steps per rotation, giving the fastest coarsest RPM, or micro-stepped at up to about 60,000 steps per rotation (much smoother and far slower).  When the motor is stopped, the driver provides enough power to hold the rotor in position.   The driver is fitted with a limiter than prevents too many amps being pumped into a held motor  – 4.2A is a reasonable limit for this motor, more would overheat it.

What power supply do you have Ian, Volts, Amps and Type?

The PSU has to provide enough amps to drive the motor.  24V at 6A should do, but the type matters too.   I always use regulated LED power supplies for this, because they’re common as muck and cheap, not because they’re best practise, and one day I shall come unstuck.  Regulated supplies aren’t ideal for driving stepper motors because a regulator designed for a steady load may not cope with a stream of nasty short sharp on/off pulses.  If the regulator is discombobulated by a pulsed load, the power output available at the motor will be throttled.  The ideal power supply for this stepper would be 40Vdc at 6A unregulated, not easy to find.

For winding a milling table, where accuracy and repeatability don’t matter much, I’d go for speed, setting the driver to 200 steps per revolution.  Above all the power supply has to get at least 4A into the motor.   Arguably for this application something like a dimwitted windscreen wiper motor with a PWM controller and a reduction belt drive is a better bet.  Smart steppers are good for rotary tables, where the ability to turn and set an accurate angle is essential.  Wiper motors are hopeless for spinning rotary tables, but good at fast crude movements. Unlike steppers, more volts causes more amps to flow in a wiper motor, and their speed and power increases as expected.   Steppers don’t behave the same way.

Dave

[SillyOldDuffer]

On SillyOldDuffer Said:

Some confusion here, this is a stepper motor with an electronic drive,

• Only the absolute maximum speed of the motor is determined by the supply voltage.  Applying more volts allows the motor to accelerate faster without losing steps, but motor speed isn’t proportional to voltage.

The supply voltage to a stepper motor driver determines how torque falls off with speed. Higher supply voltage will allow higher speed running *at a given torque*.

The motor that the OP has selected has reasonably low inductance so *should* be capable of considerably faster rotation than he’s currently using, allowing the option of gearing it down to increase torque without compromising top speed.

Sorry if my posts above have confused you.

[Iain DownsParticipant @iaindowns78295]

Thanks Dave – all of that made sense.  My PSU is switched mode (pretty much all of the conversions seem to use these).

I’ve ordered a 36V 5A one from china.  You can apparently get a bare bones (no case) one for about 7 quid, but it takes longer to arrive, so I splashed out.

What I will try and do is to get a 20T timing pulley to run at 2:1 and drop the microstepping.  See how that all goes.

this will take some time, both for parts to arrive and me to find the odd day when it’s warm enough to get out to the shed and do things (today is about 8 degrees and that’s just a little too cold for me!).

 

Iain

[john fletcher 1Participant @johnfletcher1]

Dave, picking your brain reading and not understanding much about steppers, or their power supplies either. Would a transformer with say 30/35 volts output, with a rectifier and large smoothing capacitor giving 40 volts at 6 amps be suitable ? John

[John HaineParticipant @johnhaine32865]

Especially with a higher supply voltage, the current consumed from the supply  is significantly less than the coil current you set since the driver operates as a switched mode power supply. The input power equals the output power plus a bit for efficiency.   Easy to over spec the supply!  John your transformer rectifier proposal would be just fine, I used exactly that in my lathe conversion. I would avoid any drive with a  tba6600 chip, they are out of the ark. I’m not sure what arduino code you use but GRBL is free and handles all the acceleration stuff.

[Iain DownsParticipant @iaindowns78295]

Hi, John

I’ve rolled my own.  Well, I used the FastAccel library – I’d rolled something totally my own but I could get my head around more advanced cases (accelerate to a velocity then accelerate again and so on).

Also, John, with an unregulated supply is there not a risk of the voltage rising somewhat above nominal if there is no load?  So if the transformer is specced as 24V, which is what it would deliver under load, under no load it could rise to quite a bit more (off hand about 1.4 times as much (inverse of root mean square)?  That could be over the driver’s allowed voltage…

Dave – I picked a stepper motor with the idea of moving to CNC at some point.  In hindsight I wished I’d picked the windowscreen wiper route 🙁

 

Oaom

[Iain Downs]

On Iain Downs Said:

Hi, John

…

Also, John, with an unregulated supply is there not a risk of the voltage rising somewhat above nominal if there is no load?  So if the transformer is specced as 24V, which is what it would deliver under load, under no load it could rise to quite a bit more (off hand about 1.4 times as much (inverse of root mean square)?  That could be over the driver’s allowed voltage…

 

 

Oaom

In your example the transformer gives 24v rms, which is 1.414×24 = 34v peak which is what the smoothing cap will charge to off load.  The cap needs to be sized so the load current will not cause an unacceptable ripple between charge pulses from the rectifier.   But that off load calc gives the max voltage which you need to be less than the max driver supply.

[Robert Atkinson 2Participant @robertatkinson2]

The statement “The supply voltage to a stepper motor driver determines how torque falls off with speed. Higher supply voltage will allow higher speed running *at a given torque*.”

Is not the whole story. Stepper torque is only related to the current in the coil. The voltage is irrelevant to torque as liong as the current flows. The reason voltage does mattter is that to produce current you need a voltage differential across the coil.
With the motor static the situation is simple, the current is proportional to the applied voltage divided by the coil resistance (Ohms law). So for a 1 Ohm motor and 5 Volt supply you will get 5 Amps. When you want the motor to move there are two primary effects that reduce the current as speed increases with voltage constant. First is back electro motive force (back EMF) This is the voltage generated in the coils by the rotating magnetic field. It’s polarity opposes the applied voltage (right hand rule for generators and left hand rule motors). As the speed rises effective voltage across the coil decreases so the torque reduces. This is exactly the same effect that causes the speed of a permanent magnet DC motor to vary with applied voltage.
The second effect is the inductance of the coils. This opposes any change in the current flowing and is inversely proportional to frequency. So as the step rate (frequency) increases the current increases.
Increasing the voltage overcomes both these effects but is not that simple. If we take our 5V 1 Ohm motor example if the voltage required for the operating speed is 10V then when stopped the current will be 10A causing the the motor to overheat. The answer to this is using an electronic circuit that keeps the current constant. The input supply voltage to this circuit does however have to be high enough to overcome the inductance and back EMF of the motor at the required speed.

[Author]

Posts