[john_m0ers]

Hi,
I'm having a problem with a stepper motor lashup hopefully someday becoming a cnc mill. When I connect the power supply, a 36v 11A switched mode, as I increase the stepper speed through a dedicated stepper driver the voltage measured across the power supply (and the stepper driver) begin to rise towards the maximum allowed voltage of the driver. The motor is a nema 23 4nM motor. As far as I can tell I have all the dip switches in their correct positions for current and step.
Is there anything I can or need to do to correct this, as I would not want to damage an eighty pound driver?
Thanks
john

[ger21]

What is the max voltage for the drive?

[dharmic]

How are you measuring the voltage? I'm going to guess with a DC voltmeter (or multimeter on VDC setting).

That setting will average out the voltage of the pulsetrain coming out of the driver. What'll be happening is that the inductor of the motor coil will be ringing, ie causing a voltage spike. This may be exacerbated by the switchmode trying to keep up with the inrush current at the start of a pulse and overshooting, too. When motor's going slowly, the spikes will be averaged out and the reading will be lower. Go faster, it spends a greater proportion of the time in a spike, so the average goes up.

I'd suggest you try lay hands on an oscilloscope and take a snapshot of the waveform, send it to wherever you got the driver from and ask if it's ok. It probably will be. Otherwise you're looking at a snubber circuit across the stepper which will smooth out the spikes at the penalty of slowing the response time of the stepper down a bit.

[OpinionCalf]

Depends what is the maximum voltage for the drive. But anyhow, interesting explanation you got here. Thanks for sharing !

[A_Camera]

Hi,
I'm having a problem with a stepper motor lashup hopefully someday becoming a cnc mill. When I connect the power supply, a 36v 11A switched mode, as I increase the stepper speed through a dedicated stepper driver the voltage measured across the power supply (and the stepper driver) begin to rise towards the maximum allowed voltage of the driver. The motor is a nema 23 4nM motor. As far as I can tell I have all the dip switches in their correct positions for current and step.
Is there anything I can or need to do to correct this, as I would not want to damage an eighty pound driver?
Thanks
john

Which voltage are you measuring? The power supply to the driver? That voltage should be fairly constant and should not change with stepper speed. Also, the driver should have protection for back EMF, so any voltage generated by the motor should be stopped long before even reaching the driver. Of course, if those protective diodes are faulty then anything might happen...

[KOC62]

... When I connect the power supply, a 36v 11A switched mode, as I increase the stepper speed through a dedicated stepper driver the voltage measured across the power supply (and the stepper driver) begin to rise towards the maximum allowed voltage of the driver. ...

He's measuring the power supply voltage.

[A_Camera]

He's measuring the power supply voltage.

Yes, I did read the OP, and fully understand what he is writing but I am not sure he measured it right. Such behaviour is not normal. Just checked my CNC, and regardless how fast I am driving any axis the power supply voltage remains the same.

[john-100]

its possible the switch mode power supply is at fault

not all switchmode power supplies are stable when used to power CNC motor drives
resulting the output not being regulated as it should be

either no longer the voltage expected - being under or over the set voltage
or the required voltage most of the time but with bursts of the maximum unregulated voltage available - killing your drivers

at first sight a £20 or £30 400W power supply from the likes of ebay looks like a £220 psu from
a trusted industrial supplier I use
the difference being the one from the trusted supplier will be to a guaranteed spec

If I pay a bit more
I can buy power supplies that are guaranteed to work in parallel and share the load
if the spec does not expressly say they can be connected in parallel the odds are one power supply will try to supply all the current

having said that I expect 99% of the members here getaway with the £30 power supplies


John

PS

power supply design gets more interesting when you need to hot swop supplies or the boards they power

[dharmic]

Yep - my earlier post, sorry, I was thinking of voltages on the other side of the stepper driver. But the switchmode supplies have a similar issue and cheap ones can be utter rubbish at maintaining stability with pulsed loads.

[RCaffin]

In fact, many mfrs will recommend that you do NOT use a switch-mode PS on their drives. Maybe they know something?
MUCH safer to use a simple transformer, ring bridge rectifier and large cap, to make a voltage a bit less than the driver rating. Do NOT worry about regulation: the drivers are designed to handle this sort of supply.

Cheers
Roger

[john_m0ers]

Hi, thanks for all the replies, I'm not quite sure where to start. The driver is rated at 50vdc max at up to 5amps. It doesn't seem to matter which motor/driver combination I use (I have several), or if the psu is linear or switch mode, or just a transformer-rectifier-capacitor. The voltage rises as the motor speed increases. Fortunately the drivers have an overvoltage protection monitor that disconnects power if the voltage goes too high. I've tried this with several different makes of drivers and motors, and they all do it. The only consistent thing is that in each case the motors are running with no load. Could this be a factor? The only other thing I thought was that may be the motors were at their maximum speed rating. The only cheap thing I was using was a variable pulse generator pcb that generates the step pulse.
Thanks
john
p.s sorry forget to mention how and where the voltage is changing. Its at the dc input to the driver, but can be measured also at the psu terminals. I even tried it on a tti lab psu and sure enough the voltage rises until the ovp kicks in. I'm totally new to cnc, but I,ve been involved in most aspects of electronics for nearly fifty years, and this one has me baffled.

[RCaffin]

I am not sure what is going on here, but it sounds perilously close to a perpetual motion machine, That is, you draw power from a supply and the supply voltage goes UP. Very strange.
Cheers
Roger

[john-100]

Hi John

now we know the problem exists with different power supplies

I have a few questions

what is the inductance of the motor windings ?

using this rule of thumb from Geckodrive :-
The power supply voltage is determined by multiplying the square root of the motor’s inductance by 32,

using a 36V the motors inductance should be no more than 1.26 mH ( 32 x sqare root 1.26 = 36 )

any more inductance reduces the maximum speed you can step
(as the inductance is increased the longer it take the current to build up at the start of the step)

if you used a 50V supply the motors inductance can be increased to 2.44 mH


are you runing the motor without a mechanical load ?

you could find the problem goes once the motor is driving a mechanical load

John

[john_m0ers]

Hi, I've just found the inductance of the motor is 2.54mH per winding. I was testing the motors/drives with no load, which I guessed will have an impact. I was just curious how the power supply
could rise considerably above the p.s.u voltage regardless of type. Is it the motor behaving as a generator made worse with back emf? If I had any kind of batteries at 24v I suppose it should be rock steady.
The thing that concerns me is that these are not cheap ebay specials.
Thanks for all the thoughts
john

[dharmic]

I was just curious how the power supply could rise considerably above the p.s.u voltage regardless of type.

Short version: switching power supplies rectify and pulse mains voltage into a coil. So if the regulation is off, your output voltage can spike up to and even past mains voltage.

Long version: Switching power supplies are based on the property of an inductor that they like to maintain a constant current. If you put a voltage across an inductor it will try to stay zero current, then fight until it reaches some steady current through it. Turn the voltage off and the current will continue to flow (as the magnetic field collapses), eventually returning to zero. A big enough inductor means you can switch a voltage on and off quickly and the current through the inductor will stay pretty much the same.

Given a steady load like a light bulb etc, you can regulate the voltage output by changing how the input chopping is timed. A normal mains switcher converts the mains voltage to DC and then switches that into one side of the main inductor. The other side goes to your load and the voltage monitoring circuit. The voltage monitor watches the output voltage will change based on load and supplied current. It's all pretty quick but there is a definite lag between a load change and the PSU responding.

So let's say the PSU is humming along and happily delivering 36V or whatever to your stepper driver because the steppers are off. Now you get a step output. We've changed the load considerably so the voltage drops. The PSU reacts by increasing the width of the pulses being switched into inductor. The current through the coil increases and the voltage recovers. Now the step output stops. The current stays the same into what is now a much, much higher impedance load and so the voltage out of the coil spikes. Eventually the voltage monitor responds by dropping the pulse width into the coil but in the meantime, depending on how big and fast the load impedance switch is, you could see spikes of mains voltage on the output on the output. Even higher, if you are unlucky and there's a capacitative element in the load which just happens to tune to form a tank circuit.

Morals of the story:
1. There are a bazillion supplies in perforated shiny tin boxes out there. And the $$meanwell$$ looks identical to the cheapo eBay job. But a good switcher will have a much faster response time and often have some decent filtering on the output stage.
2. Even when you spend the big bucks on a decent switcher, there's a solid chance that you'll hit a combination of switcher operation frequency and driver (because they also often use a PWM output which is switched too) frequency which can resonate and cause all sorts of madness.

Big old toroid or iron core transformers with the traditional rectifier, capacitor and linear regulator are expensive, inefficient and can run hot. They are horrible to the mains too, because they introduce this whopping great phase offset onto the load. But they provide butter smooth power on the output that really doesn't care much about the load.

Switchers are compact, clean on the input side, cheap, run cool and for many applications are far, far better. But their output can be lumpy and they react strangely to nonlinear loads. Some switcher manufacturers go to some lengths to clean up the responsiveness but not all of them, it's hard to tell which ones do (price is a guide), and even then it won't always work.