Hendrikus,
How would you drive a stepper that has a higher inductance, say 7mH? Also with a fan and 120VDC? Or, are you saying that your drive only works on very low inductance motors?
Hendrikus,
How would you drive a stepper that has a higher inductance, say 7mH? Also with a fan and 120VDC? Or, are you saying that your drive only works on very low inductance motors?
I would not like to drive such high inductance. Good Nema23 and Nema 34 steppers can have low induction, for example there is a Powerpac from PaSci, Nema34, doublestack, 10,83Nm, 10A current rating with only 1.4 milliH. (I need to look at the new CT-line from PaSci, as Danaher is making my beloved motors obsolete).
Such a motor can be driven with 160VDC and 10A, microstepsize dynamically depending on speed. Because one choose this motor because of its high torque, you accept a lower speed, max 1500 rpm. And with 1500 rpm the heating will not be too much. With low speed however, you construct machines with lower reduction in which the inherent step-elasticity (the phenomenon that Mariss described so nicely) might cause unacceptable positioning errors. Very often the Nema23 will have enough torque, and combined with high speed and high reduction, will deliver optimal precisions, much better then low-end servo.
Greetings from Amsterdam and good night.
Makes sense, less turns, more current, less inductance to drive. Water cooled steppers like high speed 3 phase spindles may be next.
Most people will then say to use servomotors. I would say to use a Nema34 stepper with a highres encoder and feed back the elasticity error to the drive. As this phase-lag error (see Mariss) is never more than about a quarter of a full step (unless you make a serious design error in your machine) and look-ahead software is already extensively built-in in every stepper-cnc, this makes a servo whose performance comes close to expensive high-end servo systems from Siemens (accelerations in G, precisions in micrometer). The only extra investment is a lot of embedded software development (I am already quite far with this) and an encoder. You also need a stepper drive based on highres microstepping and dynamic stepsize, as is the case with the Sensa drive.
Hey,
I recently came across this web page. Maybe someone will find it interesting and informative, I did.
Regards
Bart
With a good controller you will have slightly better torque with 64x microstepping than 1/2 stepping, each of them being roughly 8-10% less than full stepping.
At 64x microstepping your pulses are approximating a sine wave pretty well.
Lin Engineering has a white paper on this, and this is where I got that data from. See attached.
The attached document does not refer to how you achieve the same input powers per unit step/microstep, but I know with stepper drivers that close the loop with stepper motors via encoders, you can set current gains where you slightly overshoot the rated current for an extremely small time period. This has the possibility to squeeze more torque out of the motor than what it is actually rated for and exceed the torque/rpm curves.
I'm rather infatuated with the copley stepnet controllers as they will do exactly what I described above:
Copley Controls - Stepnet - Stepper Drive - CANopen - CAN bus
the stepnet plus goes a step further by allowing you to close a position loop with an external linear encoder.
I run M-drive Plus2 integrated stepper motors at 256x microstepping without any noticeable loss in torque, and certainly no resonances throughout a 1000 RPM range. Holding torque is claimed to take a huge hit with microstepping, but this certainly is not true with the mdrive motors.
hey i am new her and i build a cnc and now i hawe some litlle problems whit what pulse/rev and micro to set for my motors...i hope that i didn broken them:/
i use nema23 3A steper motor...
can anyone help me?
thanks
Hi to All
I'm new here and new with stepping motors and have a question about it.
I have two 5-phase stepping motors Berger Lahr VDRM 568LN and want to make a driver for it.
Specs says Current 1.5A / Winding.
I now in half-step mode alternative 4 or 5 phase are active.
Will that mean that in worst case I need 7.5A pro motor ?
If so, I can driving the motor with 10 wires so every phase with 1.5A.
If I use it in pentagon mode, 6 wires, I see in the cycles that some phase are used in serial, so will the torque be 1/2 ? (1.5A for 2 phase ?)
Hope someone can answer.
(sorry for my bad English)
No One ?
Wrong treat ?
Bit extreme, you shouldn't need encoders with good drivers. A dsp driver should be ones first resort. Accuracy with these things is much more predictable and in so it seems reliable correction can be made, I have these on an 8x4 diy build with 10x microstepping for direct drive motors and 2x microstepping on 20:1 planatary gearboxes and I'm seeing repeatable accuracy of 0.05 tested over entire bed and measurable in cut parts, So now in the realms of ballscrew and rack accuracy.
So unless you have c3 or at least c5 accuracy screws etc, stepper motors powered correctly with just 2x microstepping with 20:1 gearing or 10x direct drive are capable of matching c7 screw accuracy without encoders.
Go back and read my post where I responded to James. Microstepping at 64X will not reduce torque by 64X - that's just silly. In fact, with equivalent current RMS current output, 64X can produce more torque than 1/2 stepping at low and high RPMS (with near equivalency through mid-band), and it will be smoother, and you will operate nowhere near the resonance frequencies of the mechanical components.
A good driver will be able to use encoder feedback to it's advantage. The whole point of using steppers is to be accurate in an open loop configuration and to utilize high holding torque, but the capabilities of stepper motors have been expanded in drivers that utilize encoder feedback. Many drivers out there are servoing stepper motors now, you can squeeze out more torque out of a stepper motor than it is rated for, use less power (create much less heat) and more accurately position the stepper. Example: Copley Controls - Stepnet - Stepper Drive - CANopen - CAN busOriginally Posted by Jon.N.CNC
0.050" repeatability might be fine for a wood CNC router, but that would be pretty terrible for any commercial mill, or even what many of the DIY'ers want out of their system.
Another great use for encoders is that you can tolerate sloppier and less accurate mechanics and account for it in your control system.
Are you possitive about this? can you back it up with any evidence? because a simple search on googles image search for microstepping torque graph brings up a hell of a lot of graphs that seem to disagree. Here is a company for example with a nice explanation of it and a nice graph: Microstepping: Myths and Realities | MICROMO
quote: "The real compromise is that as you increase the number of microsteps per full step the INCREMENTAL torque per microstep drops off drastically. Resolution increases but accuracy will actually suffer."
I agree completely and i know this first hand.. my planetary gearbox motors struggle for torque at anything over 10x microstepping. I run them at half step. they run great, they match my R&P accuracy. And that is all that is required. if torque drop wasnt happening at just 10x then how would you explain this?
Even on my direct drive screws i see accuracy starts to drop at over 32x, there's another very true quote that explains this on that link:
"taking an infinite number of microsteps per full step results in two-phase synchronous permanent magnet ac motor operation, with speed a function of the frequency of the ac power supply. The rotor will lag behind the rotating magnetic field until sufficient torque is generated to accommodate the load."
If you increasing steps to reduce resonance, just need to be aware of torque drop and lag affecting accuracy. when it comes to electronics you never get anything for free, amps - volts, torque - smoothness, everything has its cost so its always a compromise.
so going by that graph and the many others floating around the net exactly the same, it looks as though at 64x steps its not just 64x less torque, its a massive 97x decrease.
No point in telling someone go for 64 microsteps and you will get a resolution of 0.005 for example if their screw or r&p is 0.05. Just wasting torque. And the same goes for encoders. That was my point.
I think you have really missed the meaning of INCREMENTAL
Increasing microsteps from 2 to 64 or even 1024 does _NOT_ reduce the motor stall torque
in fact it often improves performance by reducing resonances
The motor lag vs load does _NOT_ change with the amount of microstepping so increasing
microstepping will not decrease your accuracy. The torque vs displacement curve depends on the
number of poles, the motor current, and the motor construction, none of which vary appreciably
with the microstep ratio.
Last edited by PCW_MESA; 10-22-2015 at 03:42 PM.
yes i do understand what incremental means, ie holding torque, which is an important factor in regards to motor torque over high speed torque which is different, Maybe i should have specified but a legitimate comment none then less.
So what happens when say the bit is touching your stock and you want to incrementally step by one step? Incrementally the torque is reduced so tell me what is likely to happens in comparison to having the full range of torque available?.
If this is no concern as you prefer not to actually cut anything with the machine and just run it up and down at full velocity because you like how smooth it sounds, what about stopping, your gantry weighs say 1/4 of a ton and you have only 2% of possible torque available to stop it as the motor deceleration curve comes to an end where on the very last step it requires the motor to have enough torque to halt the inertia.
So by performance you mean, sounds smooth but lacks accuracy? larger amounts of microstepping only improves torque at speed. this is where all the heavy torque requirements of the machine have already happened. So again we are talking about another compromise that needs to be made, like ive already said, nothing is free. so if your asking me to tell you which i think is more important, then its incremental torque as this stall torque is pie in the sky if your machine wont move a single step because there is no bottom end torque or your forever playing with your backlash settings because you cant get it to stop in the same place consistently as you've not made the connection.
Last edited by Jon.N.CNC; 10-22-2015 at 04:37 PM.
Nope, you are still missing it.
What the MICROMO paper means by incremental torque is that if you have 64 microsteps per full step the
torque required to displace the shaft 1/64 of a full step is much less than the torque required to displace the shaft
1/2 or 1/4 of a full step, which is not terrible surprising.
The holding torque is _not_ affected by the microstep ratio (well 1/2 stepping is a special case and may have more static torque but you lose on resonance)