G540 + EMC2 = missing steps

Arceye,

I found a power supply at 24v @ 15A 350 watt which will give me 5 amp supply per motor. In my case the stepper only draws 2.8 max. ea.

Thanks for your help. Hoping that this will solve my problem.

would it be okay if I change my power supply to 350 watt from 125. ( I am using xylotex (3) 269 oz stepper @ 2.8amp each.

I didn't know they made PSUs that small

I found a power supply at 24v @ 15A 350 watt which will give me 5 amp supply per motor. In my case the stepper only draws 2.8 max. ea.

You should certainly see an improvement, just make sure you keep your velocity and acceleration figures reasonable, they are still pretty small steppers after all

regards

Stepper coil current is not even remotely equal to power supply current. The stepper coil acts as a 'current storage'; it is charged to X amps and then only the losses (iron losses, copper losses, mechanical work done) have to be replenished. That does not take X amps times V volts, but only a fraction.
To make it even worse: when the current in the coil must rise power is drawn from the supply. When the current must fall, that energy is partially returned to the power supply.

It is this mechanism, together with the energy generated when the motor is decelerated, that causes so much trouble for switchmode power supplies. A normal supply is designed to supply power, not to absorb it. Of course they can be designed to cope with this situation, but almost noboby does that. Adding a large capacitor helps too, but all to often the PSU sees this as a short during startup and goes in overload protection. Grossly overdimensioning the PSU so it sees the load imposed by the stepper drives more as a nuisance than as work helps too.
.
On average a toroid, bridge rectifier and large capacitor is a far better power supply for motors that accelerate/decelerate much.

How much power do you really need? I have never seen more than 180 Watts going into 3x 4Nm NEMA34 motors fed from 75V running torture G-code which exercises all axes.
With NEMA23 ones and lower supply voltages it would be hard to pump 35 Watts into a single motor.

andypugh wrote:

I think that you premise is wrong but your conclusion may be correct.

Mainly oversimplified. I am quite comfortable with this stuff. I did design and build my own stepper drives and they are pretty decent. Full PI current loops at 60kHz, phase advance to improve high rpm torque, resonance avoidance through step interpolation, blabla. They have no problem spinning a NEMA34 motor at 2000+ rpm running off 75V (at 2000rpm there is hardly any torque left though).

In practice the current is limited by the drive PWM, rather than an actual voltage reduction, so it is the full V * I but only for a small fraction of the time.

Which is practically the same.

At full-speed the duty cycle will increase, but I am not clear on if it gets to 100% of time-on at the same time as 100% of back-emf voltage and 100% rated current.

I once thought that there must be one combination of RPM and load where the full motor voltage and full coil current flows into the stepper motor and we are indeed consuming Icoil x Vsupply. And not counting phase shifts due to inductance it happens, but only shortly. A few hundred microseconds later the drive software decides that fast decay mode is required to drop the current quickly enough, and then almost the same full current times supply voltage is returned to the supply capacitor so the average current is much lower. The input capacitor of the drive sees a lot of current though.

Can't find the scope images of coil current unfortunately; one image says more than many words.

A side note: it is always a good plan to use a low-inductance power path to the drive, so using a supply line per drive instead of daisy-chaining and twisting the supply lines is a good idea.

Interesting data-point about your setup, though. How (and where) is your power consumption measured?

Simply in the +75V line from supply to the three stepper drives. Anyone with a $20 multimeter can do the same.

Arceye,

I just dismantled my psu and stated 24v 6.5 amp so 156 watts which was supplied with kit that I bought with 3 269 oz motor drawing 2.8 amp ea. I just bought 24v @ 14A psu coming this wednesday. I will post the result
using 1/16 end mill running @ 20in/min and 32in/mm.

Thank you very for all the support that I rec'd from this forum.

Use both supplies, the big one for X and Y, the small one for Z. Unless you intend to sell one and recover some money, of course

Hi,

How is that done?

Thanks

How is that done?

Just use the output from the small PSU to the driver for the Z axis and use the output from the large PSU in parallel to both the X and Y drivers.

I do the same on my big mill, just larger motors and PSUs involved

regards